Endometrial biomarkers for the non‐invasive diagnosis of endometriosis

Abstract

Background

About 10% of reproductive‐aged women suffer from endometriosis, which is a costly, chronic disease that causes pelvic pain and subfertility. Laparoscopy is the gold standard diagnostic test for endometriosis, but it is expensive and carries surgical risks. Currently, there are no non‐invasive tests available in clinical practice that accurately diagnose endometriosis. This is the first diagnostic test accuracy review of endometrial biomarkers for endometriosis that utilises Cochrane methodologies, providing an update on the rapidly expanding literature in this field.

Objectives

To determine the diagnostic accuracy of the endometrial biomarkers for pelvic endometriosis, using a surgical diagnosis as the reference standard. We evaluated the tests as replacement tests for diagnostic surgery and as triage tests to inform decisions to undertake surgery for endometriosis.

Search methods

We did not restrict the searches to particular study designs, language or publication dates. To identify trials, we searched the following databases: CENTRAL (2015, July), MEDLINE (inception to May 2015), EMBASE (inception to May 2015), CINAHL (inception to April 2015), PsycINFO (inception to April 2015), Web of Science (inception to April 2015), LILACS (inception to April 2015), OAIster (inception to April 2015), TRIP (inception to April 2015) and ClinicalTrials.gov (inception to April 2015). We searched DARE and PubMed databases up to April 2015 to identify reviews and guidelines as sources of references to potentially relevant studies. We also performed searches for papers recently published and not yet indexed in the major databases. The search strategies incorporated words in the title, abstract, text words across the record and the medical subject headings (MeSH).

Selection criteria

We considered published peer‐reviewed, randomised controlled or cross‐sectional studies of any size that included prospectively collected samples from any population of reproductive‐aged women suspected of having one or more of the following target conditions: ovarian, peritoneal or deep infiltrating endometriosis (DIE).

Data collection and analysis

Two authors independently extracted data from each study and performed a quality assessment. For each endometrial diagnostic test, we classified the data as positive or negative for the surgical detection of endometriosis and calculated the estimates of sensitivity and specificity. We considered two or more tests evaluated in the same cohort as separate data sets. We used the bivariate model to obtain pooled estimates of sensitivity and specificity whenever sufficient data were available. The predetermined criteria for a clinically useful test to replace diagnostic surgery was one with a sensitivity of 94% and a specificity of 79%. The criteria for triage tests were set at sensitivity at or above 95% and specificity at or above 50%, which in case of negative results rules out the diagnosis (SnOUT test) or sensitivity at or above 50% with specificity at or above 95%, which in case of positive result rules in the diagnosis (SpIN test).

Main results

We included 54 studies involving 2729 participants, most of which were of poor methodological quality. The studies evaluated endometrial biomarkers either in specific phases of the menstrual cycle or outside of it, and the studies tested the biomarkers either in menstrual fluid, in whole endometrial tissue or in separate endometrial components. Twenty‐seven studies evaluated the diagnostic performance of 22 endometrial biomarkers for endometriosis. These were angiogenesis and growth factors (PROK‐1), cell‐adhesion molecules (integrins α3β1, α4β1, β1 and α6), DNA‐repair molecules (hTERT), endometrial and mitochondrial proteome, hormonal markers (CYP19, 17βHSD2, ER‐α, ER‐β), inflammatory markers (IL‐1R2), myogenic markers (caldesmon, CALD‐1), neural markers (PGP 9.5, VIP, CGRP, SP, NPY, NF) and tumour markers (CA‐125). Most of these biomarkers were assessed in single studies, whilst only data for PGP 9.5 and CYP19 were available for meta‐analysis. These two biomarkers demonstrated significant diversity for the diagnostic estimates between the studies; however, the data were too limited to reliably determine the sources of heterogeneity. The mean sensitivities and specificities of PGP 9.5 (7 studies, 361 women) were 0.96 (95% confidence interval (CI) 0.91 to 1.00) and 0.86 (95% CI 0.70 to 1.00), after excluding one outlier study, and for CYP19 (8 studies, 444 women), they were were 0.77 (95% CI 0.70 to 0.85) and 0.74 (95% CI 0.65 to 84), respectively. We could not statistically evaluate other biomarkers in a meaningful way. An additional 31 studies evaluated 77 biomarkers that showed no evidence of differences in expression levels between the groups of women with and without endometriosis.

Authors' conclusions

We could not statistically evaluate most of the biomarkers assessed in this review in a meaningful way. In view of the low quality of most of the included studies, the findings of this review should be interpreted with caution. Although PGP 9.5 met the criteria for a replacement test, it demonstrated considerable inter study heterogeneity in diagnostic estimates, the source of which could not be determined. Several endometrial biomarkers, such as endometrial proteome, 17βHSD2, IL‐1R2, caldesmon and other neural markers (VIP, CGRP, SP, NPY and combination of VIP, PGP 9.5 and SP) showed promising evidence of diagnostic accuracy, but there was insufficient or poor quality evidence for any clinical recommendations. Laparoscopy remains the gold standard for the diagnosis of endometriosis, and using any non‐invasive tests should only be undertaken in a research setting. We have also identified a number of biomarkers that demonstrated no diagnostic value for endometriosis. We recommend that researchers direct future studies towards biomarkers with high diagnostic potential in good quality diagnostic studies.

Plain language summary

Endometrial biomarkers for the non‐invasive diagnosis of endometriosis

Review question

Can physicians use biomarkers (distinctive molecules, genes or other characteristics that appear in certain conditions) to reduce the need to surgically diagnose endometriosis?

Background

The endometrium refers to the tissue that lines the womb and is shed during menstruation. Women with endometriosis have endometrial tissue growing outside the womb, within the pelvic cavity. This tissue responds to reproductive hormones causing painful periods, chronic lower abdominal pain and difficulty conceiving. Currently the only reliable way of diagnosing endometriosis is to perform keyhole surgery and visualise the endometriotic deposits inside the abdomen. Because surgery is risky and expensive, various tests within the endometrium that can be obtained during an in‐office womb sampling procedure have been assessed for their ability to detect endometriosis non‐invasively or with minimal invasion. An accurate test could lead to the diagnosis of endometriosis without the need for surgery, or it could reduce the need for diagnostic surgery so only women who were most likely to have endometriosis would require it. Review teams have also evaluated other non‐invasive ways of diagnosing endometriosis using blood, urine and imaging tests as well as a combination of several testing methods in separate Cochrane reviews within this series.

Study characteristics

The evidence in this review is current to April 2015. We included 54 studies involving 2729 participants. All studies evaluated reproductive‐aged women who were undertaking diagnostic surgery to investigate symptoms of endometriosis or for other indications. Twenty‐six studies evaluated the role of 22 different biomarkers in diagnosing endometriosis, and 31 studies identified 77 additional biomarkers that had no value in differentiating between women with and without the disease.

Key results and quality of evidence

Only two of the assessed biomarkers, a neural fibre marker PGP 9.5 and hormonal marker CYP19, were assessed in sufficient number of studies to obtain meaningful results. PGP 9.5 identified endometriosis with enough accuracy to replace surgical diagnosis. Several additional biomarkers (endometrial proteome, 17βHSD2, IL‐1R2, caldesmon and other neural markers) show promise in detecting endometriosis, but there are too few studies to be sure of their diagnostic value.

The studies differed in how they were conducted, which groups of women were studied and how the surgery was undertaken. The reports were of low methodological quality, which is why readers cannot consider these results to be reliable unless confirmed in large, high quality studies. Overall, there is not enough evidence to recommend any endometrial test for use in clinical practice for the diagnosis of endometriosis.

Future research

Further high quality research is necessary to accurately evaluate the diagnostic potential of the endometrial biomarkers for the diagnosis of endometriosis.

Summary of findings

Summary of findings'. 'Biomarkers evaluated as a diagnostic test for endometriosis.

Review question	What is the diagnostic accuracy of the endometrial biomarkers in detecting endometriosis?					Pelvic endometriosis (any site and depth of invasion)
						Ovarian endometriosis
Importance	A simple and reliable non‐invasive test for endometriosis with the potential to either replace laparoscopy or to triage patients in order to reduce surgery, would minimise surgical risk and reduce diagnostic delay
Patients	Reproductive‐aged women with suspected endometriosis or persistent ovarian mass, or women undergoing infertility work‐up
Settings	Hospitals (public or private of any level), outpatient clinics (general gynaecology, reproductive medicine, pelvic pain) or radiology departments
Reference standard	Visualisation of endometriosis at surgery (laparoscopy or laparotomy) with or without histological confirmation
Study design	Cross sectional of single gate design (N = 16) or two‐gate design (N = 11); prospective enrolment; one study could assess more than one test or more than one type of endometriosis
Risk of bias	Overall judgement		Poor quality of most of the studies (only 2 studies had 'low risk' assessment in all 4 domains)
	Patient selection bias		High risk: 20 studies; unclear risk: 4 studies; low risk ‐ 3 studies
	Index test interpretation bias		High risk: 18 studies; unclear risk: 5 studies; low risk ‐ 4 studies
	Reference standard interpretation bias		High risk: 0 studies; unclear risk: 10 studies; low risk ‐ 17 studies
	Flow and timing selection bias		High risk: 7 studies; unclear risk: 0 studies; low risk ‐ 20 studies
Applicability concerns	Concerns regarding patient selection		High concern ‐ 13 studies; unclear concern ‐ 4 studies; low concern ‐ 10 studies
	Concerns regarding index test		High concern ‐ 0 studies; unclear concern ‐ 1 study; low concern ‐ 26 studies
	Concerns regarding reference standard		High concern ‐ 0 studies; unclear concern ‐ 0 studies; low concern ‐ 27 studies
Diagnostic criteria	Replacement test: sensitivity ≥ 0.94 and specificity ≥ 0.79 SnOUT triage test: sensitivity ≥ 0.95 and specificity ≥ 0.50 SpIN triage test: sensitivity ≥ 0.50 and specificity ≥ 0.95
Test	N participants (studies)	Outcomes				Diagnostic estimates (95% CI)	Implications
		True positives (endometriosis)	False positives (incorrectly classified as endometriosis)	False negatives (incorrectly classified as disease‐free)	True negatives (disease‐free)
1. Angiogenesis and growth factors and their receptors
PROK‐1 mRNA (glandular) (prokineticin 1 gene evaluated in glandular epithelium)	24 (1)	8	2	4	10	Sens = 0.67 (0.35 to 0.90); spec = 0.83 (0.52 to 0.98)	Insufficient evidence to draw meaningful conclusions
2. Cell adhesion molecules and other matrix‐related proteins
Depolarised α‐6 integrin (glandular) (depolarised alpha‐6 integrin expression assessed in glandular epithelium)	49 (1)	20	3	10	16	Sens = 0.67 (0.47 to 0.83); spec = 0.84 (0.60 to 0.97)	Insufficient evidence to draw meaningful conclusions
α3β1 integrin (glandular) (alpha‐3 beta‐1 integrin chain evaluated in glandular epithelium)	32 (1)	17	11	0	4	Sens = 1.00 (0.08 to 1.00); spec = 0.27 (0.08 to 0.55)	Insufficient evidence to draw meaningful conclusions
α3β1 integrin (stroma) (alpha‐3 beta‐1 integrin chain evaluated in stroma cells)	32 (1)	9	11	8	4	Sens = 0.53 (0.28 to 0.77); spec = 0.27 (0.08 to 0.55)	Insufficient evidence to draw meaningful conclusions
α4β1 integrin (glandular) (alpha‐4 beta‐1 integrin chain evaluated in glandular epithelium)	32 (1)	11	9	6	6	Sens = 0.65 (0.38 to 0.86); spec = 0.40 (0.16 to 0.68)	Insufficient evidence to draw meaningful conclusions
α4β1 integrin (stroma) (alpha‐4 beta‐1 integrin chain evaluated in stroma cells)	32 (1)	10	12	7	3	Sens = 0.59 (0.33 to 0.82); spec = 0.20 (0.04 to 0.48)	Insufficient evidence to draw meaningful conclusions
β1 integrin (glandular) (beta‐1 integrin chain evaluated in glandular epithelium)	32 (1)	3	2	14	13	Sens = 0.18 (0.04 to 0.43); spec = 0.87 (0.60 to 0.98)	Insufficient evidence to draw meaningful conclusions
β1 integrin (stroma) (beta‐1 integrin chain evaluated in stroma cells)	32 (1)	13	15	4	0	Sens = 0.76 (0.50 to 0.93); spec = 0.00 (0.00 to 0.22)	Insufficient evidence to draw meaningful conclusions
3. DNA‐repair and telomer maintenance molecules
hTERT mRNA (human telomerase reverse transcriptase gene)	69 (1)	7	9	18	35	Sens = 0.28 (0.12 to 0.49); spec = 0.80 (0.65 to 0.90)	Insufficient evidence to draw meaningful conclusions
4. High throughput markers
Endometrial proteome by SELDI‐TOF‐MSa (high throughput surface enhanced laser desorption/ionisation time‐of‐flight mass spectrometry)	(5 peptide peaks of 2072 m/z; 2973 m/z; 3623 m/z; 3680 m/z and 21,133 m/z)						Insufficient evidence to draw meaningful conclusions
	27 (1)	15	2	2	8	Sens = 0.88 (0.64 to 0.99); spec = 0.80 (0.44 to 0.97)
	(5 protein peaks of 5385 m/z, 5425 m/z, 6898 m/z, 5891 m/z, 6448 m/z)						Approaches the criteria for a replacement, SnOUT and SpIN triage test; insufficient evidence to draw meaningful conclusions; promising marker for further investigation
	26 (1)	12	1	1	12	Sens = 0.92 (0.64 to 1.00); spec = 0.92 (0.64 to 1.00)
Mitichondrial proteome by SELDI‐TOF‐MS (high throughput surface enhanced laser desorption/ionisation time‐of‐flight mass spectrometry)	53 (1)	21	4	3	25	Sens = 0.88 (0.68 to 0.97); spec = 0.86 (0.68 to 0.96)	Insufficient evidence to draw meaningful conclusions
5. Hormonal markers
CYP19 (aromatase cytochrome P450)	444 (8)	171	60	50	163	Mean estimates: sens = 0.77 (0.70 to 0.85); spec = 0.74 (0.65 to 0.84)	Summary estimates do not meet the predetermined criteria for a replacement or triage test.
17βHSD2 mRNA (17‐beta hydroxysteroid dehydrogenase type 2 gene) The biomarker was evaluated for DIE only	53 (1)	16	2	14	24	Sens = 0.53 (0.34 to 0.72); spec = 0.91 (0.72 to 0.99)	Approaches the criteria for a SpIN triage test; Insufficient evidence to draw meaningful conclusions; Promising marker for further investigation
ER‐α (glandular) (oestrogen receptor ‐ alpha evaluated in glandular epithelium irrespective of menstrual cycle phase)	90 (1)	44	17	16	13	Sens = 0.73 (0.60 to 0.84); spec = 0.43 (0.25 to 0.63)	Insufficient evidence to draw meaningful conclusions
ER‐α (stroma) (oestrogen receptor ‐ alpha evaluated in stromal cells irrespective of menstrual cycle phase)	90 (1)	46	15	14	15	Sens = 0.77 (0.64 to 0.87); spec = 0.50 (0.31 to 0.69)	Insufficient evidence to draw meaningful conclusions
ER‐β (glandular) (oestrogen receptor ‐ beta evaluated in glandular epithelium irrespective of menstrual cycle phase)	90 (1)	40	11	20	19	Sens = 0.67 (0.53 to 0.78); spec = 0.63 (0.44 to 0.80)	Insufficient evidence to draw meaningful conclusions
ER‐β (stroma) (oestrogen receptor ‐ beta evaluated in stromal cells irrespective of menstrual cycle phase)	90 (1)	38	8	22	22	Sens = 0.63 (0.50 to 0.75) spec = 0.73 (0.54 to 0.88)	Insufficient evidence to draw meaningful conclusions
6. Immune system and inflammatory markers
IL‐1R2 mRNA (glandular) (interleukin‐1 receptor type II gene evaluated in glandular epithelium irrespective of menstrual cycle phase)	31 (1)	16	7	0	8	Sens = 1.00 (0.79 to 1.00); spec = 0.53 (0.27 to 0.79)	Meets the criteria for a SnOUT triage test; insufficient evidence to draw meaningful conclusions; promising marker for further investigation
IL‐1R2 mRNA (stroma) (Interleukin‐1 receptor type II gene evaluated in stromal endometrial cells irrespective of menstrual cycle phase)	32 (1)	14	4	1	13	Sens = 0.93 (0.68 to 1.00); spec = 0.76 (0.50 to 0.93)	Approaches the criteria for a replacement and SnOUT triage test; insufficient evidence to draw meaningful conclusions; promising marker for further investigation
IL‐1R2 mRNA (glandular secretory) (interleukin‐1 receptor type II gene assessed in glandular epithelium in secretory phase of menstrual cycle)	19 (1)	10	3	0	6	Sens = 1.00 (0.69 to 1.00); spec = 0.67 (0.30 to 0.93)	Meets the criteria for a SnOUT triage test; insufficient evidence to draw meaningful conclusions; promising marker for further investigation
IL‐1R2 mRNA (stroma secretory) (interleukin‐1 receptor type II gene assessed in stromal endometrial cells in secretory phase of menstrual cycle)	20 (1)	9	1	1	9	Sens = 0.90 (0.55 to 1.00); spec = 0.90 (0.55 to 1.00)	Approaches the criteria for a replacement, SnOUT and SpIN triage test; insufficient evidence to draw meaningful conclusions; promising marker for further investigation
7. Myogenic markers (markers of smooth muscle differentiation)
Caldesmon (proliferative) (calmodulin binding protein evaluated in proliferative phase of menstrual cycle)	35 (1)	19	0	1	15	Sens = 0.95 (0.75 to 1.00); spec = 1.00 (0.78 to 1.00)	Meets the criteria for a replacement, SnOUT and SpIN triage test; Insufficient evidence to draw meaningful conclusions; Promising marker for further investigation
Caldesmon (secretory) (calmodulin binding protein evaluated in secretory phase of menstrual cycle)	35 (1)	18	1	2	14	Sens = 0.90 (0.68 to 0.99); spec = 0.93 (0.68 to 1.00)	Approaches the criteria for a replacement, SnOUT and SpIN triage test; Insufficient evidence to draw meaningful conclusions; Promising marker for further investigation
CALD1 mRNA (proliferative) (gene encoding for caldesmon evaluated in proliferative phase of menstrual cycle)	35 (1)	12	2	8	13	Sens = 0.60 (0.36 to 0.81); spec = 0.87 (0.60 to 0.98)	Insufficient evidence to draw meaningful conclusions
CALD1 mRNA (secretory) (gene encoding for caldesmon evaluated in secretory phase of menstrual cycle)	35 (1)	15	5	5	10	Sens = 0.75 (0.51 to 0.91); spec = 0.67 (0.38 to 0.88)	Insufficient evidence to draw meaningful conclusions
8. Nerve sheath and nerve growth markers
PGP 9.5 (protein gene product 9.5)	429 (8) 361 (7)b	192	33	49	155	Mean estimatesb: Sens = 0.96 (0.91 to 1.00); spec = 0.86 (0.70 to 1.00)	Summary estimates meet the predetermined criteria for a replacement and SnOUT triage test
VIP (vasoactive intestinal polypeptide)	40 (1)	19	4	1	16	Sens = 0.95 (0.75 to 1.00); spec = 0.80 (0.56 to 0.94)	Meets the criteria for a replacement and SnOUT triage test; insufficient evidence to draw meaningful conclusions; promising marker for further investigation
CGRP (calcitonin gene‐related protein)	40 (1)	18	3	2	17	Sens = 0.90 (0.68 to 0.99); spec = 0.85 (0.62 to 0.97)	Approaches the criteria for a replacement and SnOUT triage test; insufficient evidence to draw meaningful conclusions; promising marker for further investigation
SP (substance P)	40 (1)	19	4	1	16	Sens = 0.95 (0.75 to 1.00); spec = 0.80 (0.56 to 0.94)	Meets the criteria for a replacement and SnOUT triage test; insufficient evidence to draw meaningful conclusions; promising marker for further investigation
NPY (neuropeptide Y)	40 (1)	19	7	1	13	Sens = 0.95 (0.75 to 1.00); spec = 0.65 (0.41 to 0.85)	Meets the criteria for a SnOUT triage test; insufficient evidence to draw meaningful conclusions; promising marker for further investigation
NF (neurofilament)	40 (1)	19	18	1	2	Sens = 0.95 (0.75 to 1.00); spec = 0.10 (0.01 to 0.32)	Insufficient evidence to draw meaningful conclusions
Combined test (VIP, PGP 9.5, SP) (Combination of 3 neural markers (vasoactive intestinal polypeptide, protein gene product 9.5 and substance P)	40 (1)	19	0	1	20	Sens = 0.95 (0.75 to 1.00); spec = 1.00 (0.83 to 1.00)	Meets the criteria for a replacement, SnOUT and SpIN triage test; insufficient evidence to draw meaningful conclusions; promising marker for further investigation
9. Tumour markers
CA‐125 (menstrual fluid) (cancer antigen 125 evaluated in menstrual fluid)	104 (1)	27	7	13	59	Sens = 0.66 (0.49 to 0.80); spec = 0.89 (0.79 to 0.96)	Insufficient evidence to draw meaningful conclusions

^aDifferent groups of proteins were discovered and evaluated in each study, hence the data were not combined in meta‐analysis.
 ^bMean estimates of 7 studies in 361 women, excluding the outlier study Leslie 2013.

Background

Target condition being diagnosed

Endometriosis

Endometriosis is an inflammatory condition associated with pelvic pain and infertility, characterised by lesions of endometrial‐like tissue outside of the uterus (Johnson 2013). Endometriotic lesions can occur at different locations, including the pelvic peritoneum and the ovary, or penetrate pelvic structures below the surface of the peritoneum as deeply infiltrating endometriosis. Each of these types of endometriosis is thought to represent a separate clinical entity but can also coexist in the same patient. Rarely, endometriotic implants can be found at more distant sites, including lung, liver, pancreas and operative scars, with consequent variations in presenting symptoms.

Endometriosis afflicts 10% of reproductive‐aged women, causing dysmenorrhoea (painful periods), dyspareunia (painful intercourse), chronic pelvic pain and infertility (Vigano 2004). The clinical presentations may vary from asymptomatic and unexplained infertility to severe dysmenorrhoea and chronic pain. These symptoms can occur with bowel or urinary symptoms, an abnormal pelvic examination or the presence of a pelvic mass; however, no symptom is specific to endometriosis. The prevalence of endometriosis in the symptomatic population is reported as 50% to 60% in women and teenage girls with pelvic pain, and in up to 50% of women with infertility (Eskenazi 1997; Goldstein 1980).

Women with endometriosis are also at an increased risk of developing several cancers and autoimmune disorders (Sinaii 2002; Somigliana 2006). The presence of disease is associated with changes in the immune response, vascularisation, neural function, the peritoneal environment and the eutopic endometrium (tissue lining the uterine cavity), suggesting that endometriosis is a systemic rather than localised condition (Giudice 2004). Endometriosis has a profound effect on psychological and social well‐being and imposes a substantial economic burden on society. Women with endometriosis may incur significant direct medical expenses from diagnostic and therapeutic surgeries, hospital admissions and fertility treatments, while indirect costs, including absenteeism and loss of productivity, compound the economic impact (Gao 2006;Simoens 2012). In the United States, the financial burden of endometriosis is about USD 12,419 per woman (Simoens 2012).

Although research has not been able to fully elucidate the pathogenesis of endometriosis, specialists commonly believe that it occurs when endometrial tissue contained within the menstrual fluid implants at an ectopic site within the pelvic cavity through retrograde flow (Sampson 1927). However, this theory does not explain the fact that only 10% of women develop endometriosis while retrograde menstruation occurs in up to 90% of women (Halme 1984). There is evidence that a variety of environmental, immunological and hormonal factors are associated with endometriosis and genetic loci that confer a risk of endometriosis, but the relative contribution of these and other causal factors is still unclear (Nyholt 2012; Vigano 2004).

Although it is impossible to time the onset of disease, on average, women have a 6‐ to 12‐year history of symptoms before obtaining a surgical diagnosis, indicative of considerable diagnostic delay (Matsuzaki 2006). Untreated endometriosis is associated with reduced quality of life and contributes to outcomes such as depression, inability to work, sexual dysfunction and missed opportunity for motherhood (Gao 2006).

Treatment of endometriosis

There is no cure for endometriosis. Treatment options include expectant management, pharmacological (hormonal) therapy and surgery (Johnson 2013). Treatment is individualised, taking into consideration a therapeutic goal (pain relief or conception) and the location of the disease. Current pharmacological therapies such as the combined oral contraceptive pill, progestogens, weak androgens and GnRH agonists and antagonists act to reduce the effect of oestrogen on endometrial tissues and suppress menstruation. These drugs can ameliorate the symptoms of dysmenorrhoea and chronic pelvic pain but are associated with side effects such as breast discomfort, irritability, androgenic symptoms and bone loss. Surgical excision of endometriotic lesions can reduce pain symptoms, but it is associated with high recurrence rates of 40% to 50% at five years postsurgery (Guo 2009). Early treatment of endometriosis improves pain levels as well as physical and psychological functioning. Furthermore, improvements in menstrual management (the use of the intrauterine system (hormonal coil) and the continuous use of the combined contraceptive pill) and fertility preservation (oocyte vitrification) raise the possibility of suppressing the progression of endometriosis and prospectively managing subfertility in endometriosis sufferers. The potential success of these preventive strategies depends on an accurate and early diagnosis. A major impediment to earlier and more efficacious treatment of this disease is diagnostic delay, due to the invasive nature of standard diagnostic tests (Dmowski 1997).

Diagnosis of endometriosis

Clinical history and pelvic examination can raise the possibility of a diagnosis of endometriosis, but the heterogeneity in clinical presentation, the high prevalence of asymptomatic endometriosis (2% to 50%) and the poor association between presenting symptoms and severity of the disease contribute to the difficulty in obtaining a reliable diagnosis based solely on presenting symptoms (Ballard 2008; Fauconnier 2005; Spaczynski 2003). Although an abnormal pelvic examination correlates with the presence of endometriosis on laparoscopy in 70% to 90% of cases (Ling 1999), there is a wide differential diagnosis for most positive physical findings. Furthermore, a normal clinical examination does not exclude endometriosis, as laparoscopically proven disease has been diagnosed in more than 50% of women with a clinically normal pelvic examination (Eskenazi 2001). A variety of tests utilising pelvic imaging, blood markers, eutopic endometrium characteristics, urinary markers or peritoneal fluid components have been suggested as diagnostic measures for endometriosis. Although large numbers of the reported markers distinguish women with and without endometriosis in small pilot studies, many do not show convincing potential as a diagnostic test when they are evaluated in larger studies by different research groups. The diagnostic value of these tests has not previously been fully systematically evaluated and summarised using Cochrane methods. Currently, there is no simple non‐invasive test for the diagnosis of endometriosis that is routinely implemented in clinical practice.

Surgical diagnostic procedures for endometriosis include laparoscopy (minimal access, or keyhole surgery) or laparotomy (open surgery via an abdominal incision). In the last several decades, laparoscopy has become an increasingly common procedure and has largely replaced traditional open surgery in patients suspected of having endometriosis (Yeung 2009). Laparoscopy has significant advantages over laparotomy, including fewer complications and shorter recovery times. Furthermore, a magnified view at laparoscopy allows better visualisation of the peritoneal cavity. Despite continuing controversy in the literature with regard to the superiority of one surgical modality over another in treating pelvic pathology, laparoscopy is the preferred technique to evaluate the pelvis and abdomen and to treat benign conditions such as ovarian endometriomas (Medeiros 2009). Surgery is currently also the only acceptable method of determining the extent and severity of endometriosis. There are several different classification systems for endometriosis (Adamson 2008; Batt 2003; Chapron 2003a; Martin 2006), but most researchers and clinicians use the revised American Society for Reproductive Medicine (rASRM) classification, which is internationally accepted as a respected tool for the objective assessment of the disease (ASRM 1997). The rASRM classification system considers the appearance, size and depth of peritoneal or ovarian implants and adhesions that are visualised during laparoscopy (Table 2) and allows uniform documentation of the extent of disease. Unfortunately, this classification system has little value in clinical practice due to the lack of correlation between laparoscopic staging, the severity of symptoms and response to treatment (Chapron 2003b; Guzick 1997; Vercellini 1996). The World Endometriosis Society has recently undertaken an endeavour to attain consensus around the optimal classification for endometriosis (Johnson 2015).

1. Staging of endometriosis, rASRM classification.

Location of endometriosis	Extent	Depth
		< 1 cm	1‐3 cm	> 3 cm
Peritoneum	Superficial	1	2	4
	Deep	2	4	6
Ovary	R Superficial	1	2	4
	Deep	4	16	20
	L Superficial	1	2	4
	Deep	4	16	20
	Posterior cul‐de‐sac obliteration	Partial	Complete
		4	40
	Adhesions	< 1/3 Enclosure	1/3‐2/3 Enclosure	> 2/3 Enclosure
Ovary	R Filmy	1	2	4
	Dense	4	8	16
	L Filmy	1	2	4
	Dense	4	8	16
Tube	R Filmy	1	2	4
	Dense	4a	8a	16
	L Filmy	1	2	4
	Dense	4a	8a	16

Stage ·1 (Minimal) ‐ score 1‐5; Stage II (Mild) ‐ score 6‐15; Stage III (Moderate) ‐ score 16‐40; Stage IV (Severe) ‐ score >40

^aIf the fimbriated end of the fallopian tube is completely enclosed, change the point assignment to 16 (ASRM 1997)

The European Society of Human Reproduction and Embryology (ESHRE) Special Interest Group for Endometriosis stated in their diagnostic and treatment guidelines that for most forms of endometriosis, women presenting with symptoms cannot obtain a definitive diagnosis without visual inspection of the pelvis at laparoscopy as the gold standard investigation (Kennedy 2005). Currently the visual or histological identification of endometriotic tissue in the pelvic cavity during surgery is not just the best available but the only diagnostic test for endometriosis that is used routinely in clinical practice.

The disadvantages of laparoscopic surgery include (but are not limited to) the high cost, the need for general anaesthesia and the potential for adhesion formation post procedure. Laparoscopy has been associated with a 2% risk of injury to pelvic organs, a 0.001% risk of damaging a major blood vessel and a mortality rate of 0.0001% (Chapron 2003c). Only a third of women who undertake a laparoscopic procedure will receive a diagnosis of endometriosis; therefore many disease‐free women are unnecessarily exposed to surgical risk (Frishman 2006).

The validity of laparoscopy as a reference test for endometriosis has is highly dependent on the skills of the surgeon. The diagnostic accuracy of laparoscopic visualisation has been compared with histological confirmation in a sole systematic review, and it was estimated as having a 94% sensitivity and 79% specificity (Wykes 2004). Subsequent studies suggested that incorporating histological verification in the diagnosis of endometriosis may improve diagnostic accuracy (Almeida Filho 2008; Marchino 2005; Stegmann 2008), but these papers have not been systematically reviewed. The clinical significance of histological verification remains debatable, and a diagnosis based on visual findings is generally reliable as long as properly trained and experienced surgeons perform an appropriate inspection of the abdominal cavity (Redwine 2003). Furthermore, excised potential endometriotic tissues are rarely serially sectioned in clinical practice, and pathologists can miss small lesions in mild disease. Thus sampling inconsistencies are also likely to influence the accuracy of histological reporting.

Summary

A diagnostic test without the need for surgery would reduce the associated surgical risks, increase accessibility to a diagnostic test and improve treatment outcomes. The need for an accurate non‐invasive diagnostic test for endometriosis continues to encourage extensive research in the field and was endorsed at the international consensus workshop at the 10^th World Congress of Endometriosis in 2008 (Rogers 2009). Although multiple markers and imaging techniques have been explored as diagnostic tests for endometriosis, none of them have been implemented routinely in clinical practice, and many have not been subject to a systematic review.

Index test(s)

This review is part of the review series on the non‐invasive diagnostic tests for endometriosis and looks at endometrial biomarkers that have been proposed for the diagnosis of endometriosis. The other reviews from this series are 'Blood biomarkers for the non‐invasive diagnosis of endometriosis', 'Urinary biomarkers for the non‐invasive diagnosis of endometriosis', 'Imaging modalities for the non‐invasive diagnosis of endometriosis' and 'Combination of the non‐invasive tests for the diagnosis of endometriosis'.

The definition of 'non‐invasive' varies between medical dictionaries but refers to a procedure that does not involve penetration of the skin or physical entrance to the body (McGraw‐Hill Dictionary of Medicine 2006; The Gale Encyclopedia of Medicine 2011). Although the endometrial tests are associated with intrauterine tissue sampling and therefore are invasive by this definition, these tests are generally considered to be non‐invasive or minimally invasive compared with surgery. For the purposes of our series, we define all tests that do not involve anaesthesia and surgery as non‐invasive. This review concentrated on studies that investigated eutopic endometrial and menstrual fluid biomarkers.

The potential advantages of using endometrial tissue/menstrual fluid samples for the diagnosis of endometriosis include their non‐ or minimally invasive nature, lower cost and increased availability when compared to surgery. These tests are more acceptable to patients and usually provide a rapid result. However, the testing is dependent on the skills of the surgeon performing the biopsy, the type of instrument used for the procedure, the time of the menstrual cycle, the time to process the sample, the reliability of laboratory techniques and the quality control protocols in place.

Researchers have identified the cellular and molecular processes to characterise ectopic endometrium and peritoneal fluid in human and animal models (D'Hooghe 2001; Hull 2008; Kao 2003). Animal (baboon) and human studies demonstrate clear differences in the eutopic endometrium in subsets of individuals with endometriosis versus normal controls, suggesting that endometriosis induces characteristic changes in eutopic endometrial tissues (Akoum 1995; Fedele 1990; Jones 2006) or that the eutopic endometrium of women who develop endometriosis is basically different from the endometrium of women who do not develop the condition (Al Jefout 2009b). The identification of the proteins and transcripts that probably account for these changes could form the basis of a diagnostic test utilising endometrial biopsy tissue. A growing body of literature on aberrant expression of genes in the endometrium of women with endometriosis supports this assumption (Brosens 2003). Proteolytic enzymes and immune cell populations have displayed a differential expression in eutopic endometrium of women with and without endometriosis (Chung 2001; Cox 2001; Klentzeris 1995). Additional studies have evaluated and eliminated glycodelin A, CYR61, annexin 1, osteopontin and aromatase P450 as potential endometrial biomarkers (Absenger 2004; Dheenadayalu 2002; Kao 2003; Kitawaki 1999b; Wei 2009). A promising biomarker for endometriosis is the immunohistochemical identification of small nerve fibres in the functional layer of the endometrium using an antibody against PGP 9.5 (protein gene product 9.5) (Al Jefout 2009a). Endometrial fluid, aspirated from the uterine cavity, may be another possible diagnostic biomarker for endometriosis (Ametzazurra 2009).

To date, a limited number of small studies with varying methodologies, laboratory techniques and types of assays have assessed endometrial tests. One large systematic review from 2011 studied endometrial differences in women with endometriosis (May 2011). The review included 32 eligible papers, 9 of which were of high quality. Six papers of high quality examined nerve fibre growth or cell cycle control in endometrial biopsies. These biomarkers showed promise as a minimally invasive form of diagnosis for endometriosis. An updated review in 2015 reviewed the literature on plasma, urine and endometrial biomarkers (Fassbender 2015). They were unable to confidently identify any modality of testing with high sensitivity or specificity. There is a current need to re‐evaluate the diagnotic test accuracy of the endometrial tests using Cochrane methods.

Clinical pathway

Women presenting with symptoms of endometriosis (dysmenorrhoea, dyspareunia, chronic pelvic pain or difficulty conceiving) are generally investigated with a pelvic ultrasound scan to exclude other pathologies, which is in line with international guidelines (ACOG 2010; Dunselman 2014; SOGC 2010). There are no other standard investigative tests, and MRI is used conservatively because of its cost. If patients seek pain management rather than conception, physicians generally initiate empirical treatment with progestogens or the combined oral contraceptive pill. Diagnostic laparoscopy is considered if empirical treatment fails or if women decline or do not tolerate empirical treatment. In women who have difficulty conceiving, laparoscopy can be undertaken before fertility treatment (particularly if severe pelvic pain or endometrioma are present) or after failed assisted reproductive technology (ART) treatments. Physicians may also diagnosis endometriosis during fertility investigations in women who have minimal or no pain symptomatology.

On average there is a delay of 6 to 12 years from onset of symptoms to definitive diagnosis at surgery. Early referral to a gynaecologist with the capability to perform diagnostic surgery is associated with a shorter time to diagnosis. Collectively, young women, women in remote and rural locations and women of lower socioeconomic status have reduced access to surgery and are less likely to obtain a prompt diagnosis of endometriosis.

Prior test(s)

Most women presenting with symptoms suggestive of endometriosis have a full history, examination and a routine gynaecological ultrasound before a decision is made to have diagnostic surgery. However, there is no consensus on whether or not a routine ultrasound or any other test should be part of a standardised approach.

Role of index test(s)

A new diagnostic test can fulfil one of three roles:

1. Replacement: replacing an existing test due to better accuracy or a similar accuracy with other advantages.

2. Triage: used as an initial step in a diagnostic pathway to identify the group of patients who need further testing with an existing test. Although ideally a triage test has a high sensitivity and specificity, it may have a lower sensitivity but higher specificity than the current test or vice versa. The triage test does not aim to improve the diagnostic accuracy of the existing test but rather to reduce the number of individuals having an unnecessary diagnostic test.

3. Add‐on: used in addition to existing testing to improve diagnostic performance (Bossuyt 2008).

Ideally a diagnostic test is expected to correctly identify all patients with a disease and to exclude all patients without that disease; in other words it should have a sensitivity and specificity of 100%. A high sensitivity indicates that there are a low number of patients who have a negative test and do have the disease (i.e. a low number of false negative results). High specificity corresponds to a low number of patients who have a positive test but do not have the disease (i.e. low false positive results). In practice, however, it is extremely rare to find a test with equally high sensitivity and specificity. An acceptable replacement test would need to have a similar or higher sensitivity and specificity than the current gold standard. In the case of laparoscopy for diagnosis of endometriosis, the only systematic review reported a sensitivity of 94% and a specificity of 79%, and we have taken this as a cut‐off for a replacement test (Wykes 2004).

The purpose of triage tests can vary depending on the clinical context and patients' priorities. One reasonable approach is to exclude the diagnosis to avoid further unnecessary and expensive diagnostic investigations. High sensitivity tests have few false negative results and act to rule conditions out (SnOUT). A negative result from a test with high sensitivity will exclude the disease with high certainty independent of the specificity. As women without disease would be assured of having a negative test, unnecessary invasive interventions can be avoided. However, a positive result has less diagnostic value, particularly when the specificity is low. We predetermined that a clinically useful 'SnOUT' triage test should have a sensitivity of 95% or more and a specificity of 50% and above. The sensitivity cut‐off for a 'SnOUT' triage test was set at 95% and above, assuming that a 5% false negative rate is statistically and clinically acceptable. The specificity cut‐off was set at 50% and above, to avoid diagnostic uncertainty in more than 50% of the population with a positive result.

An alternative approach would be to avoid a missed diagnosis. High specificity tests have few false positive results and act to rule conditions 'in' (SpIN). A positive result for a highly specific triage test indicates a high likelihood of having endometriosis. This information could be used to prioritise these patients for surgical treatment. A positive 'SpIN' test could also provide a clinical rationale to start targeted disease‐specific medical management in a patient without a surgical diagnosis, under the assumption that disease is present. Surgical management could then be reserved for cases when conservative treatment fails. This is particularly relevant in some populations where the therapeutic benefits of surgery for endometriosis have to be carefully balanced with the disadvantages (e.g. young women, women with medical conditions or pain‐free patients with a history of infertility). In this scenario we considered a sensitivity of 50% and above and a specificity of 95% and higher as suitable cut‐offs for a 'SpIN' triage test.

We evaluated combinations of tests for their potential to replace surgery (replacement test) or to improve the selection of patients for surgery (triage test) that can either rule out (SnOUT) or rule in (SpIN) the disease. Both types of triage test are clinically useful, minimising the number of unnecessary interventions. Sequential implementation of SnOUT and SpIN tests can also optimise a diagnostic algorithm (Figure 1). We did not assess any test as an add‐on test, as we sought tests that reduce the need for surgery and not tests that improve the accuracy of the currently available surgical diagnosis.

1.

Sequential approach to non‐invasive testing of endometriosis

Alternative test(s)

There are no alternative tests for the diagnosis of endometriosis that are available in routine clinical practice.

Rationale

Many women with endometriosis suffer longstanding pelvic pain and infertility prior to a diagnosis. Surgery is the only current method of diagnosing endometriosis, but it is associated with high costs and surgical risks. A simple and reliable non‐invasive test for endometriosis with the potential to either replace laparoscopy or to triage patients in order to reduce surgery, would minimise surgical risk and reduce diagnostic delay. Endometriosis could then be detected at a less advanced stage and earlier interventions instituted. This would provide the opportunity for a preventive approach for this debilitating disease. Healthcare‐related costs of endometriosis could drop with early diagnosis and more cost‐effective and efficient treatments. Furthermore, identifying endometrial biomarkers that do not pertain to endometriotic disease would help clinicians and researchers focus on clinically relevant biomarker detection.

Objectives

Primary objectives

To determine the diagnostic accuracy of the endometrial biomarkers for pelvic endometriosis, using a surgical diagnosis as the reference standard. We evaluated the tests as replacement tests for diagnostic surgery and as triage tests to inform decisions to undertake surgery for endometriosis.

Secondary objectives

1. To investigate the influence of heterogeneity on the diagnostic accuracy of endometrial biomarkers for endometriosis. Potential sources of heterogeneity include:

   1. participant characteristics: age (adolescents versus later reproductive years), clinical presentation (subfertility, pelvic pain, ovarian mass, asymptomatic women), stage of disease (rASRM classification system), geographic location of study;

   2. histological confirmation in conjunction with laparoscopic visualisation compared to laparoscopic visualisation alone;

   3. changes in technology over time: year of publication, modifications applied to conventional laboratory techniques;

   4. methodological quality: differences in the revised Quality Assessment of Diagnostic Accuracy Studies (QUADAS‐2) evaluation (Table 3), including low versus unclear or high risk; consecutive versus non‐consecutive enrolment; and blinding of surgeons to the results of index tests;

   5. study design (single‐gate design versus two‐gate design studies).

2. To assess the biomarkers that were not affected by endometriosis and hence are unlikely to discriminate between patients with and without the disease.

2. QUADAS 2‐ risk of bias assessment tool: methodological items and applicability judgement.

Domain 1 ‐ Patient selection
Description	Describe methods of patient selection; describe included patients (previous testing, presentation, intended use of index test, and setting)
Type of bias assessed	Selection bias, spectrum bias
Review question	Women of reproductive age with clinically suspected endometriosis (symptoms, clinical examination ± presence of pelvic mass), scheduled for surgical exploration of pelvic or abdominal cavity for confirmation of the diagnosis ± treatment
Informaton collected	Study objectives, study population, selection (inclusion and exclusion criteria), study design, clinical presentation, age, number of enrolled and number of available for analysis, setting, place and period of the study
Signalling question 1	Was a consecutive or random sample of patients enrolled?
Yes	If a consecutive sample or a random sample of the eligible patients was included in the study
No	If a consecutive sample or a random sample of the eligible patients was not included in the study
Unclear	All the studies that did not specify enrolment as consecutive or random sample of patients were classified as 'no'; therefore none of the included studies were classified as 'unclear'
Signalling question 2	Did the study avoid inappropriate exclusions?
Yes	If inclusion and exclusion criteria were presented and all patients with suspected endometriosis were included, with an exception for those who either had a history of medical conditions or were on medical therapy that would have potentially interfered with interpretation of index test (e.g. malignancy, pregnancy, autoimmune disorders, infectious diseases, treatment with hormonal or immunomodulator substances); refused to participate in the study; or were unfit for surgery
No	If the study excluded the patients based on education level, psychosocial factors, genetic testing or phenotype or excluded patients with any comorbidities commonly present in general population, including a population that could have undergone a testing for endometriosis in clinical setting (hypertension, asthma, obesity, benign gastrointestinal or renal disease, etc.)
Unclear	If the study did not provide clear definition of the selection (inclusion/exclusion) criteria and 'no' judgement was not applicable
Signalling question 3	Was a 'two‐gate' design avoided?
Yes	If the study had a single set of inclusion criteria, defined by the clinical presentation (i.e. only participants in whom the target condition is suspected) ‐ a single‐gate design
No	If the study had more than one set of inclusion criteria in respect to clinical presentation (i.e. participants suspected of target condition and participants with alternative diagnosis in whom the target condition would not be suspected in clinical practice) ‐ a 'two‐gate' study design
Unclear	If it was unclear whether a 'two‐gate deign' was avoided or not
Risk of bias	Could the selection of patients have introduced bias?
Low	If 'yes' classification for all the above 3 questions
High	If 'no' classification for any of the above 3 questions
Unclear	If 'unclear' classification for any of the above 3 questions and 'high risk' judgement was not applicable
Concerns about applicability	Are there concerns that the included patients do not match the review question?
Low	If the study includes only clinically relevant population that would have undergone index test in real practice and includes representative form of target condition
High	If the study population differed from the population defined in the review question in terms of demographic features and comorbidity (e.g. studies with multiple sets of inclusion criteria with respect to clinical presentation including either healthy controls or alternative diagnosis controls that would not have undergone index test in real practice). Further, if target condition diagnosed in the study population was not representative of the entire spectrum of disease, such as limited spectrum of severity (e.g. only mild forms) or limited type of endometriosis (e.g. only DIE)
Unclear	If this information was unclear (e.g. severity of endometriosis was not reported)
Domain 2 ‐ Index test
Description	Describe the index test, how it was conducted and interpreted
Type of bias assessed	Test review bias, clinical review bias, interobserver variation bias
Review question	Any test of endometrial tissue or menstrual fluid sample
Informaton collected	Index test name, description of positive case definition by index test as reported, threshold for positive result, examiners (number, level of expertise, blinding), interobserver variability
Signalling question 1	Were the index test results interpreted without knowledge of the results of the reference standard?
Yes	If the operators performing/interpreting index test were unaware of the results of reference standard
No	If the operators performing/interpreting index test were not blinded to the results of reference standard
Unclear	If this information was unclear
Signalling question 2	If a threshold was used, was it pre‐specified?
Yes	If study clearly provided a threshold for positive result, which was defined before execution/interpretation of index test
No	If a threshold for positive result was not provided or not defined prior to test execution
Unclear	If it was unclear whether a threshold was pre‐specified or not
Signalling question 3	Was a menstrual cycle phase considered in interpreting the index test?
Yes	If all the included participants were in the same phase of menstrual cycle, if the study reported subgroup analyses per cycle phase, or if study reported the pooled estimates after impact of the cycle phase on biomarker expression was not detected
No	If study included participants in different phases of menstrual cycle, but effect of cycle phase on index test was not assessed
Unclear	If the cycle phase was not reported
Risk of bias	Could the conduct or interpretation of the index test have introduced bias?
Low	If 'yes' classification for all the above 3 questions
High	If 'no' classification for any of the above 3 questions
Unclear	If 'unclear' classification for any of the above 3 questions and 'high risk' judgement was not applicable
Concerns about applicability	Are there concerns that the index test, its conduct, or interpretation differ from the review question?
Low	We considered all types of eutopic endometrial biomarkers as eligible, therefore all the included studies were classified as 'low concern', unless 'unclear' judgement was applicable
High	We did not consider the studies where index tests other than eutopic endometrial biomarkers were included (or excluded information on other index tests reported in addition to endometrial tests) or where index test looked at other target conditions not specified in the review (e.g. studies aimed at classifying pelvic masses as benign and malignant); therefore none of the included studies was classified as 'high concern'
Unclear	If study did not present sufficient information on at least one of the following: laboratory method, sample handling, reagents used or experience of the test operators
Domain 3 ‐ Reference standard
Description	Describe the reference standard, how it was conducted and interpreted
Type of bias assessed	Verification bias, bias in estimation of diagnostic accuracy due to inadequate reference standard
Review question	Target condition ‐ pelvic endometriosis, ovarian endometriosis, DIE. Reference standard ‐ visualisation of endometriosis at surgery (laparoscopy or laparotomy) with or without histological confirmation
Informaton collected	Target condition, prevalence of target condition in the sample, reference standard, description of positive case definition by reference test as reported, examiners (number, level of expertise, blinding)
Signalling question 1	Is the reference standard likely to correctly classify the target condition?
Yes	If the study reported at least one of the following: surgical procedure was described in sufficient detail; criteria for positive reference standard were stated; diagnosis was confirmed by histopathology; or the procedure was performed by the team with high level of expertise in diagnosis/surgical treatment of target condition, including tertiary referral centres for endometriosis
No	If reference standard did not classify target condition correctly; considering the inclusion criteria, none of the studies were classified as 'no' for this item
Unclear	If information on execution of the reference standard, its interpretation or operators was unclear
Signalling question 2	Were the reference standard results interpreted without knowledge of the results of the index tests?
Yes	If operators performing the reference test were unaware of the results of index test
No	If operators performing the reference test were aware of the results of index test
Unclear	If this information was unclear
Risk of bias	Could the reference standard, its conduct, or its interpretation have introduced bias?
Low	If 'yes' classification for all the above 2 questions
High	If 'no' classification for any of the above 2 questions
Unclear	If 'unclear' classification for any of the above 2 questions and 'high risk' judgement was not applicable
Concerns about applicability	Are there concerns that the target condition as defined by the reference standard does not match the question?
Low	Considering the inclusion criteria, all the studies were classified as 'low concern', as anticipated
High	We excluded the studies where participants did not undergo surgery for diagnosis of endometriosis, therefore none of the included studies were classified as 'high concern'
Unclear	Only studies were laparoscopy/laparotomy served as a reference test were included; therefore none of the included studies was classified as 'unclear concern'
Domain 4 ‐ Flow and timing
Description	Describe any patients who did not receive the index tests or reference standard or who were excluded from the 2 × 2 table; describe the interval and any interventions between index tests (sample collection) and the reference standard
Type of bias assessed	Disease progression bias, bias of diagnostic performance due to missing data
Review question	Less than 12‐month interval between index test (sample collection) and reference standard ‐ endometriosis may progress over the time, so we had chosen an arbitrary time interval of 12 months as an acceptable time interval between the sample collection and surgical confirmation of diagnosis
Informaton collected	Time interval between index test (sample collection) and reference standard, withdrawals (overall number of reported and if were explained)
Signalling question 1	Was there an appropriate interval between index test (sample collection) and reference standard?
Yes	If time interval was reported and was less than 12 months
No	We excluded all the studies where time interval was longer than 12 months; therefore none of the included studies were classified as 'no' for this item
Unclear	If time interval was not stated clearly, but authors description allowed to assume that the interval was reasonably short
Signalling question 2	Did all patients receive the same reference standard?
Yes	If all participants underwent laparoscopy/laparotomy as a reference standard. Considering the inclusion criteria, all the studies were classified as 'yes' for this item, as anticipated
No	If all participants did not undergo surgery or had alternative reference standard or if only a subset of participants had surgery as reference standard, but the information on this population was not available in isolation. Considering the inclusion criteria, none of the included studies were classified as 'no' for this item
Unclear	If this information was unclear. Considering the inclusion criteria, none of the included studies were classified as 'unclear' for this item
Signalling question 3	Were all patients included in the analysis?
Yes	If all the patients were included in the analysis or if the patients were excluded because they did not meet inclusion criteria prior to execution of index test or if the withdrawals were less than 5% of the enrolled population (arbitrary selected cut‐off)
No	If any patients were excluded from the analysis because of un interpretable results, inability to undergo either index test or reference standard, or if withdrawals were more than 5% of the enrolled population
Unclear	If this information was unclear
Risk of bias	Could the patient flow have introduced bias?
Low	If 'yes' classification for all the above 3 questions
High	If 'no' classification for any of the above 3 questions
Unclear	If 'unclear' classification for any of the above 3 questions and 'high risk' judgement was not applicable

Methods

Criteria for considering studies for this review

Types of studies

Published peer‐reviewed studies that compared the results of one or several types of eutopic endometrial biomarker tests with the results obtained by surgical visualisation of lesions for the diagnosis of endometriosis.

We included the following types of studies.

1. Randomised controlled trials.

2. Observational studies with the following designs.

   1. 'Single‐gate design' (studies with a single set of inclusion criteria defined by clinical presentation). All participants had clinically suspected endometriosis.

   2. 'Two‐gate design' (studies where participants are sampled from distinct populations with respect to clinical presentation). The same study includes participants with a clinical suspicion of having the target condition (e.g. women with pelvic pain) and also participants in whom the target condition is not suspected (e.g. women admitted for tubal ligation). Two‐gate studies were eligible only where all cases and controls belonged to the same population with respect to the reference standard (i.e. all the participants were scheduled for laparoscopy) (Rutjes 2005).

3. Studies performed on prospectively collected samples, irrespective of the actual time of the test assay. The timing of sample collection relative to surgery is important because the surgical excision of endometriotic lesions could influence endometrial biomarker expression and hence bias the results. Therefore, we only included studies where the biological sample was collected before the surgical procedure, i.e. 'prospectively collected'. We considered to be eligible the studies performed on tissue bank samples collected from prospectively recruited, well‐defined populations, which prevented the omission of valuable data from adequately designed studies. The time interval between sample collection and laboratory testing may influence test outcomes, which could be dependent on sample storage conditions and the stability of each individual biomarker during storage and freeze‐thawing. This information was not readily available for most molecules, and we did not address it in this review, but we will consider it in future updates if more evidence emerges.

We did not impose limits on eligibility related to the healthcare settings where the study took place, the language of publication, the number of participants in the included studies or the number of studies that evaluated each index test.

We excluded the following types of studies.

1. Narrative or systematic reviews.

2. Studies of retrospective design where investigators collected samples after execution of the reference test.

3. Studies of retrospective design where investigators selected participants from retrospective review of the case notes/archived samples and where information on recruitment methods or study population was not available.

4. Case reports or case series.

5. Studies reported only in abstract form or in conference proceedings where the full text was not available. We applied this limitation after facing substantial difficulty in obtaining the information from the abstracts, which precluded a reliable assessment of eligibility and methodological quality.

Participants

Study participants included reproductive‐aged women (puberty to menopause) with suspected endometriosis based on clinical symptoms, pelvic examination or both, who undertook the index test as well as the reference standard.

Participants came from populations of women undergoing abdominal surgery for the following indications.

• Clinically suspected endometriosis (pelvic pain, infertility, abnormal pelvic examination, or a combination of the above).

• Ovarian mass, regardless of symptoms.

• A mixed group consisting of women with suspected endometriosis/ovarian mass or women with other benign gynaecological conditions (e.g. surgical sterilisation, fibroid uterus, etc).

• Asymptomatic women who have an incidental finding of endometriosis at surgery performed for another indication.

Articles that included participants of postmenopausal age were eligible when the data for the reproductive age group was available in isolation. We excluded studies with participants that clearly would not undergo the index test in the relevant clinical situation or would not benefit from the test (e.g. women with ectopic pregnancies or acute pelvic inflammatory disease). We also excluded publications that only analysed participants with a positive index test or reference standard and did not provide data for the whole cohort.

Index tests

Any type of eutopic endometrial biomarker (including biomarkers in menstrual fluid), which were assessed either separately or in combination with other endometrial tests. We have classified the assessed index tests according to the type of biomarker, presented these categories in Table 4. To assist readers in the search for a specific biomarker, we include an index of all biomarkers with biological annotation in Appendix 1.

3. Endometrial biomarkers evaluated in this review.

Biomarker		Diagnostic potential for endometriosis
1	Angiogenesis and growth factors and their receptors
A	EGF (epidermal growth factor)	Expression not altered in endometriosis
B	FGF‐2 (fibroblast growth factor‐2)	Expression not altered in endometriosis
C	Glycodelin A (PP14 or PAEP) (placental protein 14 or progestogen‐associated endometrial protein)	Expression not altered in endometriosis
D	PDGF (platelet derived growth factor)	Expression not altered in endometriosis
E	PIGF (placental growth factor)	Expression not altered in endometriosis
F	PKR1 (prokineticin receptor 1), EG‐VEGF receptor	Expression not altered in endometriosis
G	PKR2 (prokineticin receptor 2), EG‐VEGF receptor	Expression not altered in endometriosis
H	PROK‐1 (prokineticin 1)	Diagnostic accuracy assessed
I	TSP‐1 (thrombospondin‐1)	Expression not altered in endometriosis
J	TYMP (thymidine phosphorylase)	Expression not altered in endometriosis
K	VEGF (vascular endothelial growth factor)	Expression not altered in endometriosis
2	Apoptosis markers and regulators
A	Bax (BCL2‐associated X protein)	Expression not altered in endometriosis
B	Bcl‐xL (B‐cell lymphoma‐extra large, or BCL2‐like 1 isoform 1)	Expression not altered in endometriosis
C	Bcl‐xL:Bcl‐xS ratio (ratio B‐cell lymphoma‐extra large/B‐cell lymphoma‐extra small)	Expression not altered in endometriosis
3	Cell adhesion molecules and other matrix‐related proteins
A	α2β1 integrin	Expression not altered in endometriosis
B	α3β1 integrin	Diagnostic accuracy assessed; expression not altered in endometriosis in some studies
C	α4β1 integrin	Diagnostic accuracy assessed; expression not altered in endometriosis in some studies
D	α5β1 integrin	Expression not altered in endometriosis
E	α6β1 integrin	Expression not altered in endometriosis
F	αVβ3 integrin	Expression not altered in endometriosis
G	αVβ5 integrins	Expression not altered in endometriosis
H	αVβ6 integrins	Expression not altered in endometriosis
I	β1 integrin	Diagnostic accuracy assessed
J	Depolarised α6 integrin	Diagnostic accuracy assessed
K	ICAM‐1 (intercellular adhesion molecule‐1) or sICAM‐1 (soluble form of intercellular adhesion molecule‐1)	Expression not altered in endometriosis
L	E‐cadherin	Expression not altered in endometriosis
M	LAMA5 (laminin subunit alpha‐5)	Expression not altered in endometriosis
N	LFA‐3 (CD58) (leukocyte function associated molecule‐3)	Expression not altered in endometriosis
O	MMP‐1 (matrix metalloproteinase‐1)	Expression not altered in endometriosis
P	MMP‐9 (matrix metalloproteinase‐9)	Expression not altered in endometriosis
Q	OPN (osteopontin)	Expression not altered in endometriosis
R	PAI‐1/‐2/‐3 (plasminogen activator inhibitors 1/2/3)	Expression not altered in endometriosis
S	PAs: tPA, uPA (plasminogen activators: tissue‐type PA, urokinase‐type PA)	Expression not altered in endometriosis
T	TIMP‐1 (tissue inhibitor of metalloproteinases)	Expression not altered in endometriosis
U	VCAM‐1 (CD106) (vascular cell adhesion molecule‐1)	Expression not altered in endometriosis
4	Cell cycle regulatory molecules
A	Cyclin B1	Expression not altered in endometriosis
B	Cdc2 (cyclin dependent kinase‐2)	Expression not altered in endometriosis
3	CPlk1 (polo‐like kinase‐1)	Expression not altered in endometriosis
5	Cell proliferation markers
A	Ki‐67 (antigen KI‐67 or MKI67, marker of cellular proliferation)	Expression not altered in endometriosis
B	BW 495/36, endometrial epithelial marker	Expression not altered in endometriosis
6	Cytoskeleton molecules
A	Cytokeratin 18	Expression not altered in endometriosis
B	CK19 or CYFRA 21‐1 (cytokeratin 19)	Expression not altered in endometriosis
C	Vimentin	Expression not altered in endometriosis
7	DNA‐repair and telomer maintenance molecules
A	hTERT (human telomerase reverse transcriptase)	Diagnostic accuracy assessed
B	Telomerase activity	Expression not altered in endometriosis
8	High throughput markers
A	Endometrial proteome	Diagnostic accuracy assessed
B	Mitochondrial proteome	Diagnostic accuracy assessed
C	mRNAome (mRNA micro‐array)	Expression not altered in endometriosis
9	Hormonal markers
A	17βHSD2 (17‐β hydroxysteroid dehydrogenase type 2)	Diagnostic accuracy assessed
B	CYP19 (aromatase cytochrome P450)	Diagnostic accuracy assessed
C	ER‐α (oestrogen receptor‐alpha)	Diagnostic accuracy assessed
D	ER‐β (oestrogen receptor‐beta)	Diagnostic accuracy assessed
E	EST (oestrogen sulphotransferase)	Expression not altered in endometriosis
F	LGR7 (leucine‐rich G protein‐coupled receptor 7), relaxin receptor	Expression not altered in endometriosis
G	Relaxin	Expression not altered in endometriosis
10	Immune system and inflammatory markers
A	Cytokines
i	LIF (leukaemia‐inhibitory factor)	Expression not altered in endometriosis
ii	TNF‐α (tumour necrosis factor alpha)	Expression not altered in endometriosis
B	Immune cells: peripheral blood mononuclear cells (PBMC)
i	Lymphocytes	Expression not altered in endometriosis
ii	B‐lymphocytes	Expression not altered in endometriosis
iii	Monocytes/macrophages	Expression not altered in endometriosis
iv	NK (natural killer cells)	Expression not altered in endometriosis
v	T‐lymphocytes	Expression not altered in endometriosis
C	Interleukins
i	IL‐1β	Expression not altered in endometriosis
ii	IL‐11	Expression not altered in endometriosis
iii	IL‐1R1 (interleukin‐1 receptor type II)	Expression not altered in endometriosis
iv	IL‐1R2 (interleukin‐1 receptor type II)	Diagnostic accuracy assessed
D	Other immune/inflammatory markers
i	MPO (myeloperoxidase)	Expression not altered in endometriosis
ii	NAG (N‐acetyl‐b‐D‐Glucosaminidase)	Expression not altered in endometriosis
11	Mediators of prostaglandin biosynthesis
A	Akr1B1 mRNA (aldoketoreductase ‐1B1, PGF2a synthase)	Expression not altered in endometriosis
B	Akr1C3 mRNA (aldoketoreductase ‐1C3, PGF2a synthase)	Expression not altered in endometriosis
C	Cox‐1 mRNA (cyclo‐oxygenase‐1)	Expression not altered in endometriosis
D	15‐PGDH mRNA (15‐hydroxyprostaglandin dehydrogenase)	Expression not altered in endometriosis
E	cPGES mRNA (cytosolic PGE2 synthase)	Expression not altered in endometriosis
12	Myogenic markers (markers of smooth muscle differentiation)
A	Caldesmon (calmodulin binding protein)	Diagnostic accuracy assessed
B	CALD1 (gene encoding for caldesmon)	Diagnostic accuracy assessed
13	Nerve sheath and nerve growth markers
A	CGRP (calcitonin gene‐related protein)	Diagnostic accuracy assessed
B	NF (neurofilament)	Diagnostic accuracy assessed, but expression not altered in endometriosis
C	NPY (neuropeptide Y)	Diagnostic accuracy assessed
D	PGP 9.5 (protein gene product 9.5)	Diagnostic accuracy assessed, expression not altered in endometriosis in some studies
E	SP (substance P)	Diagnostic accuracy assessed
F	VIP (vasoactive intestinal polypeptide)	Diagnostic accuracy assessed
14	Other peptides and proteins
A	hBD‐2 (human b‐defensin‐2)	Expression not altered in endometriosis
15	Transcription factors and signalling molecules
A	AKT1 (RAC‐alpha serine/threonine‐protein kinase)	Expression not altered in endometriosis
B	JAG1 (jagged‐1 protein)	Expression not altered in endometriosis
16	Tumour markers
A	CA‐125 (cancer antigen 125) in menstrual fluid	Diagnostic accuracy assessed

We included index tests performed on the whole tissue sample or separate endometrial compartments and reported them the same way as presented by the authors (e.g. separate testing of glandular epithelium, stromal cells or mixed cell sample). We included tests performed in one or several phases of the menstrual cycle.

The combined evaluations of endometrial biomarkers with other methods for diagnosing endometriosis (e.g. pelvic examination or blood tests) are beyond the scope of this review and are presented separately in another review: 'Combined tests for the non‐invasive diagnosis of endometriosis'. We excluded the studies that solely assessed specific technical aspects, qualitative descriptions of lesion appearance or interobserver variability of the index tests without reporting the data on diagnostic performance. We only considered studies in which the evaluated biomarker(s) showed differential expression between the groups of women with and without endometriosis if the data were reported in sufficient detail for the construction of 2 × 2 contingency tables. We included studies in which the expression levels of the index test did not significantly differ between the groups and where contingency tables were not available, as long as the inclusion criteria were met otherwise. We considered these to be studies reporting unchanged biomarker expression in the presence of endometriosis, and we presented them in the descriptive portion of the review. Thus, we evaluated the adequately designed studies that identified biomarkers without diagnostic value, as they provide information that is likely to guide future research towards other more clinically useful biomarkers. This methodology also identified biomarkers that presented conflicting findings associated with endometriosis in some but not other publications.

We considered the diagnostic performance of an index test to be high when the test reached the criteria for a replacement test (sensitivity at or above 94% with specificity at or above 79%) or triage test (sensitivity at or above 95% with specificity at or above 50% or vice versa), or approached these criteria (diagnostic estimates within 5% of the set thresholds). We considered all other diagnostic estimates to be low.

Target conditions

Pelvic endometriosis, defined as endometrial tissue located in the pelvic cavity: involving any of the pelvic organs, peritoneum and pouch of Douglas.

We assessed three types of pelvic endometriosis.

1. Peritoneal endometriosis, defined as endometrial deposits detected on peritoneum covering pelvic organs, pelvic side walls or pouch of Douglas.

2. Ovarian endometriosis (endometrioma), defined as an ovarian cyst lined by endometrial tissue, appearing as an ovarian mass of varying size.

3. Deep infiltrating endometriosis (DIE), defined as subperitoneal infiltration of endometrial implants, i.e. when the endometriotic implants penetrate the retroperitoneal space at a distance of 5 mm or more (Koninckx 1991). DIE may be present in multiple locations, involving either the anterior or posterior pelvic compartments, or both.

We did not include certain rare types of endometriosis such as extrapelvic, bladder and ureteric endometriosis because the majority were reported in case reports or case series, and laparoscopy or laparotomy are not reliable reference standards for these conditions.

We excluded the studies where diagnosis of endometriosis was not the primary outcome (e.g. malignant versus benign masses or normal versus abnormal pelvis) and separate data for endometriosis was not available.

We did include studies that recruited selected populations of women with endometriosis (i.e. those with specific rASRM stages), because there is a poor correlation between the rASRM classification and infertility or pain symptoms. Exclusion of these studies could result in the loss of potentially important diagnostic information from otherwise eligible publications. Where possible, we addressed the impact of these studies in the assessment of heterogeneity. When a study analysed a large population with a wide spectrum of endometriosis and additionally reported a subgroup analysis of the different stages of disease severity, we only considered estimates for the entire population. This is because a subgroup analysis would not directly address the review question regarding the clinical utility of the biomarker in disease detection.

Reference standards

The reference standard was visualisation of endometriosis at surgery (laparoscopy or laparotomy) with or without histological confirmation, as this is currently the best available test for endometriosis. If reported, we reviewed information regarding the inter‐ and intraobserver correlation of the reference standard.

We only included studies in which the reference test was performed within 12 months of the sample collection, on the assumption that disease status could change within a period of one year or longer, either naturally or as a result of treatment. We excluded studies in which the participants did not undergo the reference standard or where the findings of the index test formed the basis of selection for undertaking the reference standard, as this was likely to distort an assessment of the diagnostic value of the index test.

Summary of inclusion and exclusion criteria

Inclusion criteria

1. Types of studies

   1. Published and peer‐reviewed

   2. RCTs

   3. Observational designs, including:

      1. single‐gate design (single set of inclusion criteria defined by clinical presentation): all the participants had clinically suspected endometriosis;

      2. two‐gate design (two sets of inclusion criteria with respect to clinical presentation and one set of inclusion criteria with respect to reference standard): the participants with or without a clinical suspicion of endometriosis scheduled for abdominal surgery;

   4. Published in any language

   5. Performed in any healthcare setting

   6. Any sample size

   7. Prospectively recruited participants (consecutive or non‐consecutive enrolment)

2. Participants

   1. Reproductive‐aged women

   2. Clinically suspected endometriosis, including:

      1. women who underwent abdominal surgery for other benign gynaecological conditions and had a surgical assessment for presence/absence of endometriosis;

      2. asymptomatic women who have an incidental finding of endometriosis at surgery performed for another indication;

   3. Undertook both the index test and reference standard

3. Index tests

   1. One or several types of eutopic endometrial biomarkers (including biomarkers in menstrual fluid)

   2. Data reported in sufficient detail for the construction of 2 × 2 tables for the tests that showed differential expression between the groups

4. Target condition

   1. Pelvic endometriosis: peritoneal endometriosis, ovarian endometrioma, DIE or combinations of the above

5. Reference standard

   1. Surgical visualisation of lesions for the diagnosis of endometriosis (laparoscopy or laparotomy) with or without histological verification

   2. Performed within 12 months of the endometrial sample collection

Exclusion criteria

1. Types of studies

   1. Narrative or systematic reviews

   2. Retrospective design where the index test was performed after execution of reference test or participants were selected from retrospective review of the case notes

   3. Prospectively collected samples that were selected from the archived material, but where information on the study population or the selection process was unclear

   4. Case reports or case series

   5. Conference proceedings

2. Participants

   1. Included cohort was not representative of the target population that would benefit from the test (e.g. women with known genital tract malignancy, ectopic pregnancies or acute pelvic inflammatory disease)

   2. Study included participants of postmenopausal age, where data for the reproductive age group were not available in isolation

   3. Only participants with positive index test or positive reference standard were included in the analysis;

3. Index tests

   1. Endometrial biomarkers presented in combination with other diagnostic tests for endometriosis and where separate information for endometrial biomarkers was not available

   2. Study presented only specific technical aspects of an index test or data on interobserver variability

   3. Study presented only qualitative description of the tissue samples or focused on the biological events, rather than diagnostic performance of the test

4. Target condition

   1. Endometriosis was not the primary outcome of the trial (e.g. malignant versus benign masses or normal versus abnormal pelvis)

   2. Atypical, rare sites of endometriosis

5. Reference standard

   1. Endometriosis was not the primary outcome of the trial (e.g. malignant versus benign masses or normal versus abnormal pelvis)

   2. Reference standard performed only in a subset of the study or control group

   3. Findings of the index test formed the basis of selection for the reference standard

Search methods for identification of studies

We developed the search strategy in collaboration with the Trials Search Coordinator of the Gynaecology and Fertility Review Group, following recommendations of the Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy (De Vet 2008). We did not limit the searches to particular types of study design or impose language or publication date restrictions. The search strategy incorporated words in the title, abstract, text words across the record and the medical subject headings (MeSH). We initially created the search for one broad review looking at all diagnostic markers for endometriosis, but due to complexity, the review team split the originally planned review into five separate reviews. We designed two separate search strategies: one for all the biomarkers‐based tests, and another for the imaging tests; we used the former in this review. We performed all searches from database inception to April ‐ July 2015. We present the search strategies for each database and the number of hits per search in Appendix 2; Appendix 3; Appendix 4; Appendix 5; Appendix 6. The summary of the results is presented in Results of the search.

Electronic searches

We searched the following databases to identify the published articles that assessed the diagnostic value of endometrial biomarkers for endometriosis.

• CENTRAL (2015, July).

• MEDLINE (inception to May 2015).

• EMBASE (inception to May 2015).

• CINAHL (inception to April 2015).

• PsycINFO (inception to April 2015).

• Web of Science (inception to April 2015).

• LILACS (inception to April 2015).

• OAIster (inception to April 2015).

• TRIP (inception to April 2015).

• Databases of the trial registers.

  • ClinicalTrials.gov (inception to April 2015).

  • World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (inception to April 2015).

• Databases to identify reviews and guidelines as sources of references to potentially relevant studies.

  • MEDION (inception to January 2014, the last available date).

  • DARE (inception to April 2015).

  • PubMed, a 'Systematic Review' search under the 'Clinical Queries' link (inception to April 2015).

• Searches for papers recently published and not yet indexed in the major databases.

  • PubMed (simple search for the 6 months to April 2015).

Searching other resources

We handsearched the reference list of all relevant publications (retrieved full texts of the key articles and identified reviews).

We abandoned an initial attempt to locate the grey literature (unpublished studies and conference proceedings), as we faced substantial difficulty in obtaining full‐text publications or further details of studies reported in an abstract form.

Data collection and analysis

Selection of studies

Two authors of this review (DG, VN) and four other authors from the other reviews in this series (Deepika Arora, Emily Liu, Lucy Prentice and Rabia Shaikh) scanned the titles of studies identified by our search to remove any clearly irrelevant articles. We reviewed the titles and abstracts of the remaining studies to select potentially relevant publications. We then divided the relevant articles into four categories of endometriosis biomarkers: serum, endometrial, urinary and combined tests. Two out of three review authors (DG, LM or VN) independently reviewed the full‐text versions of each article selected by title and abstract and assessed them for eligibility based on the considerations listed in Criteria for considering studies for this review. A single failed eligibility criterion was sufficient to exclude a study from the review.

The review authors who assessed the relevance of the studies and eligibility for inclusion were not blind to the information about each article, including the publishing journal, the names of authors, the institution and the results. We resolved any disagreements by discussion and, if necessary, with an additional reviewer (the late Professor Ali Akoum), an expert in the field and in methodological aspects of systematic reviews.

When papers updated previous publications and were performed on the same study population at different recruitment points, we used the most complete data set that superseded previous publications to avoid double counting participants or studies. We directly contacted authors to retrieve missing data needed to clarify study eligibility. When we found potentially relevant studies in languages other than English, we had it translated. For excluded studies, we documented the reasons for exclusion and details of which criteria were not met. We present the characteristics of included and excluded studies in Characteristics of included studies and Characteristics of excluded studies, respectively.

Data extraction and management

Two out of three review authors (DG, LM or VN) independently extracted data from each eligible study, resolving disagreements through an additional reviewer (the late Professor Ali Akoum). If required, we contacted the study investigators to resolve questions regarding the data.

To collect details from included studies, we designed a specific data extraction form for this review and piloted it on three studies of diagnostic accuracy tests for endometriosis. The following information was recorded for each study.

1. General information and study design: first author, year of publication, country, language, setting, objectives, inclusion/exclusion criteria, type of enrolment.

2. Characteristics of the study participants: age, symptoms/history/previous tests, type of target condition and its prevalence in the study population, number of participants enrolled and available for analysis, reasons for withdrawal.

3. Features of the index test and reference standard: type; diagnostic criteria; number and experience of the operators; blinding of the operators to other tests, clinical data or both; interobserver variability; time interval between index test and reference standard.

4. The reported number of true positives (TP), false negatives (FN), true negatives (TN) and false positives (FP) was used to construct a 2 × 2 table for each index test. If studies did not report these values, we attempted to reconstruct the 2 × 2 tables from the diagnostic estimates presented in the article.

We extracted data into Review Manager (RevMan) software, which was used to graphically display the quality assessment, the diagnostic estimates data and the descriptive analyses (RevMan 2014).

Assessment of methodological quality

We used QUADAS‐2, a modified version of the QUADAS tool, to assess the quality of each included study (Whiting 2011).

We present the review‐specific QUADAS‐2 tool and explanatory document in Table 3. For each paper, we assigned a 'low', 'high' or 'unclear' risk of bias for four domains and assessed concerns about applicability in three domains. We considered studies as having low methodological quality when classified as having high or unclear risk of bias and raising high concerns on applicability at least in one domain. Two of the three reviewers (DG, LM or VN) independently assessed each included study for quality, settling disagreements by consensus. Two review authors (DG, LM) independently piloted the topic‐specific tool to rate four of the included studies with a high level of agreement. Modifications specific to the endometrial biomarkers review were made to the signalling questions of the original QUADAS‐2 tool.

1. Domain 1: We rephrased the original signalling question 'Was a case‐control design avoided?' as 'Was a two‐gate design avoided?'. The diagnostic studies are cross‐sectional in nature, aiming to compare the result of an index test with the result of the reference standard in the same group of participants. In these studies the parameters are measured at a single point of time and the groups are classified by the outcome of the reference standard, albeit the analysis is performed retrospectively. Therefore, the terminology 'cohort' and 'case‐control' is less informative for diagnostic test trials than for epidemiological studies, and we substituted it for 'single‐gate' and 'two‐gate' designs. We included this question because a two‐gate design has more potential to introduce selection bias.

2. Domain 2: We introduced an additional signalling question, 'Was the phase of the menstrual cycle considered in interpreting the index test?' to assess bias in the interpretation of the test results. Endometrium is a menstrual cycle‐dependent tissue and is a sensitive target for steroid sex hormones, which can result in differential expression of biomarkers at different cycle phases.

The assessment of methodological quality was undertaken for each domain, but we did not calculate a summary score to estimate the overall quality of studies (Whiting 2005).

Statistical analysis and data synthesis

We entered the extracted data into RevMan 2014 to produce forest plots of sensitivity and specificity, and to plot study‐specific estimates of sensitivity and specificity in the receiver operating characteristic (ROC) space for each index test. We investigated the diagnostic performance of each test and visually explored between‐study variation in performance of each index test and its relationship to patient characteristics, study design and study quality factors. In case of two or more tests evaluated in the same cohort, we included them as separate data sets, since the unit of analysis was the test result, not the patient.

For studies that reported the subgroup analyses per phase of menstrual cycle, we applied judgement to what was relevant to present. For instance, we presented only pooled estimates where available when there was no statistically significant difference in biomarker expression between the cycle phases. Alternatively, where putative biomarkers demonstrated cycle‐dependent expression or were noted to be modulated by ovarian hormones, we reported the test performance either for several time points across the menstrual cycle or for the most distinctive phase.

We estimated the expected operating point (mean sensitivity and specificity) and corresponding 95% confidence region using the bivariate logistic normal random‐effects model for all meta‐analyses with four studies or more. When the number of studies was fewer than five, we did not attempt to estimate the covariance and reported a zero. To estimate the performance of the other tests in meta‐analyses with a small number of studies (two or three), we performed fixed effect meta‐analysis of sensitivity and specificity (summary sensitivity and specificity), in the absence of substantial heterogeneity. We performed the meta‐analyses using SAS NLMIXED software (Cary, NC: SAS Institute Inc). We entered results from SAS into RevMan to provide plots of the mean or summary point(s) and confidence region(s), superimposed on the study specific estimates of sensitivity and specificity.

We assessed the comparative accuracy of index tests in two ways. In direct, fully paired comparisons where all the study participants received more than one index test as well as the reference standard, we plotted the estimates in RevMan. If meta‐analysis was possible, we used test‐level covariates in the bivariate logit normal model to identify statistically significant differences. Otherwise we reported the available comparative data in a narrative way and illustrated it using forest and ROC plots.

When judging test performance against the predetermined diagnostic criteria, we considered the point estimates of sensitivity and specificity as the most informative presentation of test performance. We acknowledge that tests with point estimates that did not reach the predetermined criteria, but with confidence intervals (CIs) that contained values above the threshold, could have diagnostic value. Furthermore, tests with point estimates that reached the criteria but with CIs containing values below the threshold could have an overestimated diagnostic value. If we use the range of the CIs rather than the point estimates of the data, the predetermined cut‐off becomes meaningless. Therefore we did not consider CIs in qualifying the test performance but utilised this information in interpreting the reliability of the obtained data.

Dealing with missing data

We defined missing data as any information on the study population, index tests or reference standard that were not available from the publication and that were required to determine the eligibility of the study for inclusion, assess the methodological quality, or construct the results table. If we identified missing data, we contacted the authors in an attempt to obtain them. If missing data prevented a clear judgment regarding applicability for inclusion or the construction of accurate 2 × 2 tables and the data were unavailable from the primary investigators (for example we were unable to locate the contact details of the authors, there was no reply from the authors or the authors replied that the requested information was unavailable), we excluded the study from the review.

Investigations of heterogeneity

We initially assessed heterogeneity by visually examining the forest plots of sensitivities and specificities and the ROC plots for each index test. We describe the potential sources of heterogeneity in the Secondary objectives. For diagnostic tests where there were more than 10 eligible studies, we initially planned to formally explore heterogeneity by using study level covariates, but we were unable to do so because of the small numbers of studies in each group.

Sensitivity analyses

We planned to conduct sensitivity analyses to assess the impact of methodological quality of included studies on the results of meta‐analysis if sufficient data were available. We defined low quality studies as having a high risk of bias for one or more QUADAS‐2 domains. We also planned to assess the sensitivity of results to the inclusion and exclusion of outlying studies in all analyses and planned to use the 'leave‐one‐out' procedure (Higgins 2008) to assess the impact of each study on the meta‐analysis results (leading study effect). However, we refrained from doing so because of the small number of studies for most analyses. It is important to use caution when interpreting small meta‐analyses (few studies) with a limited total sample size.

Assessment of reporting bias

A comprehensive search of multiple sources for eligible studies, a search of trial registers and no language restrictions minimised the risk of reporting bias. However, publication bias generally arises when studies have a higher chance of being published if their results are positive. Therefore, we initially searched and evaluated unpublished and published study databases and conference proceedings. During the process of qualifying the studies for inclusion in this review, we faced substantial difficulty in obtaining full‐text publications or further details of studies published in an abstract form. This precluded a reliable assessment of eligibility and methodological quality, and we decided not to include these publication sources in the review.

Results

Results of the search

The literature search identified 33,438 references from: CENTRAL (N = 226), MEDLINE (n = 10,328), EMBASE (n = 10,313), CINAHL (n = 1,131), PsycINFO (n = 174), Web of Science (n = 7,425), LILACS (n = 420), OAIster (n = 446), Trip (n = 1,648), Trial registers for ongoing and registered trials (n = 523), MEDION (n = 2), DARE (n = 99), PubMed, a ‘Systematic Review’ search (n = 418) and simple search PubMed (n = 267). These databases were searched from inception to 20 April ‐ 31 July 2015. We present the flow of the selection process in Figure 2. We screened titles to exclude duplicates (N = 9312) and clearly irrelevant studies (N = 21,534), and we eliminated another 2224 references after reading the abstracts because they did not address the research question or clearly did not meet the inclusion criteria. We retrieved the full texts of the remaining 368 references and assessed them for eligibility. We needed author clarification on data from 35 studies, and we had 33 non‐English publications translated. Ultimately, 54 studies were eligible according to the inclusion criteria and provided data for the review, while 314 studies were ineligible and excluded.

2.

Flow of the studies identified in literature search for systematic review on endometrial biomarkers for a non‐invasive diagnosis of endometriosis.

Basic features of included studies

We present the details of the included studies in Characteristics of included studies. The 54 eligible studies included 2729 participants, with a median of 49 women per study (range 16 to 104). Twenty‐six studies took place in Europe, 15 in Asia, 6 in South America, 3 in North America, 3 in Australia, and 1 in Africa. Fifty‐three of the included studies took place at university hospitals, of which nine were tertiary referral centres for endometriosis. The earliest article was published in 1990; 50 articles were published after 2000, and 17 studies were published after 2010. All the included studies evaluated women of reproductive age.

While a number of research groups explored more than one marker, some authors reported several estimates for the same biomarker at different menstrual cycle phases or in different endometrial compartments, with each estimation considered a separate test. Twenty‐seven studies evaluated the diagnostic accuracy of endometrial biomarkers. An additional 31 methodologically eligible studies examined biomarkers that demonstrated no differential expression in endometriosis, so we did not evaluate the diagnostic test accuracy of these biomarkers. Four studies assessed diagnostic accuracy in only some of the investigated biomarkers.

Twenty‐five studies had a 'single‐gate design', 27 studies had a 'two‐gate design', and 2 studies presented insufficient information to determine which study design they used. Laparoscopy was the predominant surgical modality, whereas laparotomy was co‐utilised in 52% (28/54) of the studies. Fifty‐four per cent (29/54) of the included studies used histopathology to confirm the surgical diagnosis. The reported prevalence of endometriosis varied, ranging from 30% to 77%. Seven studies included only participants with minimal‐mild endometriosis (rASRM stage I‐II), eight studies included participants with moderate‐severe endometriosis (rASRM stage III‐IV), and in nine studies the information on the severity of the disease was not available.

Basic features of excluded studies

We describe the excluded studies in Characteristics of excluded studies. After a full‐text assessment, we excluded 314 publications. Of these 314, 165 studies contained insufficient diagnostic accuracy information to construct 2 × 2 contingency tables. We excluded an additional 84 studies that not adequately describe methods or population; 10 of these studies used archived samples with poor population definition. In 32 studies, the index test was outside the inclusion criteria. Many articles in this excluded group were comparisons between ectopic endometrium in cases versus endometrium in controls (N = 16), or they presented data on physiological events of endometrium (N = 22) without direct measurement or comparison of biomarker expression between the groups. We excluded seven papers because they enrolled a wide age group, pregnant women or women with genital tract malignancies, and we could not independently assess reproductive‐aged women without these conditions. Another eight studies failed to meet eligibility criteria because they had an overlapping cohort with another updated included paper. Fourteen excluded studies used a reference standard other than abdominal surgery or did not provide data on surgical diagnosis. In four of the excluded studies, the target condition was outside inclusion criteria, reporting data for benign versus malignant masses or normal versus abnormal pelvis, with no separate data for endometriosis. We excluded one study because the time interval between index test and surgery exceeded 12 months and another because we were not able to locate the full text. One publication was retracted, one paper was a review article and one was a case report.

Methodological quality of included studies

We show the quality of the included studies in the QUADAS‐2 results summary (Figure 3; Figure 4). Overall, the studies were of poor methodological quality, and there were only two studies (Al‐Jefout 2009 and Leslie 2013) with a low risk of bias in every domain assessed.

3.

Risk of bias and applicability concerns graph: review authors' judgements about each domain presented as percentages across included studies

4.

Risk of bias and applicability concerns summary: review authors' judgements about each domain for each included study

Methodological quality assessment for the patient selection

Forty‐five studies presented a high risk of patient selection bias (Al‐Jefout 2007; Bokor 2009; Bourlev 2006; Casals 2012; Cetin 2013; Chen 2004; Chen 2013; Cho 2012; Da Silva 2014;Dheenadayalu 2002; Ding 2010; Elgafor el Sharkwy 2013; Fassbender 2012; Gilabert‐Estelles 2003; Gilabert‐Estelles 2007; Hudelist 2007; Johnson 2004; Kim 2007; Klentzeris 1995; Laudanski 2014; Lawson 2008; Lee 2010; Li 2006; Liu 2008; Mafra 2014; Makari 2012; Matsuzaki 2006a; Meibody 2011; Meola 2009; Meola 2013; Mikolajczyk 2006; Morelli 2010; Pino 2009; Prefumo 2002; Puy 2002; Rakhila 2013; Ramon 2011; Szymanowski 2003; Tang 2009; Tiberi 2010; Van der Linden 1994; Van der Linden 1995; Vernet‐Tomas 2006a; Visnovsky 2008; Wang 2010a), and six studies demonstrated an unclear risk (Hatok 2011; Lee 2007; Wolfler 2005; Yadav 2013; Zeng 2005; Zubor 2009). Non‐consecutive or non‐random patient selection, utilisation of a two‐gate design for patient selection, the absence of a clear definition of inclusion/exclusion criteria and use of a highly selected group of women were the main reasons for a high risk assessment of bias. Three studies demonstrated a low risk of bias (Al‐Jefout 2009; Leslie 2013; Takahashi 1990), as they enrolled a consecutive group of women, had clearly defined selection criteria, utilised a single‐gate design and included a wide spectrum of endometriosis.

Thirty‐six studies presented a high concern for patient selection applicability (Bourlev 2006; Casals 2012; Cetin 2013; Chen 2004; Chen 2013; Cho 2012; Da Silva 2014; Dheenadayalu 2002; Ding 2010; Elgafor el Sharkwy 2013; Gilabert‐Estelles 2003; Gilabert‐Estelles 2007; Hudelist 2007; Johnson 2004; Klentzeris 1995; Laudanski 2014; Lawson 2008; Lee 2007; Li 2006; Liu 2008; Mafra 2014; Matsuzaki 2006a; Meola 2009; Meola 2013; Mikolajczyk 2006; Morelli 2010; Pino 2009; Puy 2002; Rakhila 2013; Ramon 2011; Tang 2009; Tiberi 2010; Van der Linden 1994; Van der Linden 1995; Vernet‐Tomas 2006a; Wang 2010a), and seven demonstrated an unclear concern (Kim 2007; Meibody 2011; Prefumo 2002; Szymanowski 2003; Visnovsky 2008; Yadav 2013; Zubor 2009). We assigned a high concern in patient selection applicability if the study utilised a two‐gate selection of women as cases and controls or if the study only evaluated a limited spectrum of disease. These types of study design introduce additional uncertainty to the accuracy of the index test, which usually varies across patient subgroups. In our view, any sampling deviation from a representative group of the entire clinically relevant population could skew the estimates of diagnostic accuracy in any direction, which raises concerns for applicability of the test in practice. The remained 11 studies were of low concern (Al‐Jefout 2007; Al‐Jefout 2009; Bokor 2009; Fassbender 2012; Hatok 2011; Lee 2010; Leslie 2013; Makari 2012; Takahashi 1990; Wolfler 2005; Zeng 2005). These studies included only a clinically relevant population that would have undergone the index test in clinical practice and displayed a representative form of the target condition.

Methodological quality assessment for the index test

Forty‐two studies carried a high risk of index test interpretation bias (Al‐Jefout 2007; Bokor 2009; Bourlev 2006; Cetin 2013; Chen 2004; Chen 2013; Da Silva 2014; Ding 2010; Fassbender 2012;Gilabert‐Estelles 2003; Gilabert‐Estelles 2007; Hatok 2011; Hudelist 2007; Johnson 2004; Kim 2007; Klentzeris 1995; Laudanski 2014; Lawson 2008; Lee 2007; Lee 2010; Li 2006; Liu 2008; Matsuzaki 2006a; Meola 2009; Meola 2013; Mikolajczyk 2006; Morelli 2010; Pino 2009; Prefumo 2002; Puy 2002; Rakhila 2013; Ramon 2011; Szymanowski 2003; Takahashi 1990; Tang 2009; Visnovsky 2008; Wang 2010a; Wolfler 2005; Yadav 2013; Zeng 2005; Zubor 2009), and eight studies demonstrated an unclear risk (Casals 2012; Dheenadayalu 2002; Elgafor el Sharkwy 2013; Mafra 2014; .Makari 2012; Tiberi 2010; Van der Linden 1994; Van der Linden 1995). A lack of clear pre‐specified criteria for a positive diagnosis and failure to blind index test operators to the results of the reference standard were the main reasons for a high risk assessment. Few studies reported the skill level of a test operator or interobserver variability, both of which directly relate to the performance of tests. Four studies presented a low risk of bias and clearly mentioned the use of a predetermined cut‐off threshold, an impact of the cycle phase on biomarker expression and blinding of the test operators to the results of the reference standard (Al‐Jefout 2009; Leslie 2013; Meibody 2011; Vernet‐Tomas 2006a). Overall, the interobserver variability was rarely mentioned across the included studies.

We did not consider the studies including index tests other than eutopic endometrial biomarkers (or we excluded information on other index tests reported in addition to endometrial tests). Likewise, we did not consider studies where the index test assessed other target conditions not specified in the review (e.g. studies aimed at classifying pelvic masses as benign and malignant); therefore we did not classify any of the included studies as 'high concern'. One study presented an unclear concern of index test applicability (Zeng 2005). We assigned an unclear concern when the study did not present sufficient information regarding the conduct of the tests, such as the laboratory methods or reagents used or the level of expertise of the test operators. The remaining 53 studies were of low concern and presented adequate information to conclude that performance or interpretation of the test did not differ from the review question.

Methodological quality assessment for the reference standard

We addressed the risk of bias and concerns about applicability related to the reference standard via some key questions. These included whether the reference standard was interpreted without the knowledge of the index test results and whether this was an appropriate investigation for the classification of the target condition (endometriosis).

There were no studies that demonstrated a high risk of bias for the reference standard domain, while twenty studies demonstrated an unclear risk (Casals 2012; Chen 2004; Da Silva 2014; Dheenadayalu 2002; Fassbender 2012; Gilabert‐Estelles 2007; Hudelist 2007; Johnson 2004; Kim 2007; Klentzeris 1995; Lawson 2008; Lee 2010; Li 2006; Liu 2008; Pino 2009; Rakhila 2013; Takahashi 1990; Tiberi 2010; Vernet‐Tomas 2006a; Zeng 2005). We assigned an unclear risk of bias if there was not enough information to determine how likely the reference standard was to have correctly classified the target condition. Specifically, this might occur if studies did not adequately describe surgical procedures or the criteria for a positive reference standard, or if it was unclear if histology was used to confirm surgical diagnosis or there was no information regarding the experience of the surgeons or the pathologists involved. Thirty‐four studies were at low risk of bias in the 'reference standard' domain and reported at least one of the following adequately: surgical procedure, the criteria for positive reference standard, histopathological confirmation of diagnosis, and high level of expertise in team performing the procedure to diagnose the target condition (Al‐Jefout 2007; Al‐Jefout 2009; Bokor 2009; Bourlev 2006; Cetin 2013; Chen 2013; Cho 2012; Ding 2010; Elgafor el Sharkwy 2013; Gilabert‐Estelles 2003; Hatok 2011; Laudanski 2014; Lee 2007; Leslie 2013; Mafra 2014; Makari 2012; Matsuzaki 2006a; Meibody 2011; Meola 2009; Meola 2013; Mikolajczyk 2006; Morelli 2010; Prefumo 2002; Puy 2002; Ramon 2011; Szymanowski 2003; Tang 2009; Van der Linden 1994; Van der Linden 1995; Visnovsky 2008; Wang 2010a; Wolfler 2005; Yadav 2013; Zubor 2009).

None of the studies had high or unclear concerns for applicability in regards to the reference standard. All 54 studies were of low concern and implemented pelvic surgery (laparoscopy or laparotomy) as a reference standard, which could be relied upon to match the review question.

Methodological quality assessment for flow and timing

Fifteen studies presented a high risk of bias in the flow and timing domain (Casals 2012; Cetin 2013; Chen 2004; Cho 2012; Elgafor el Sharkwy 2013; Fassbender 2012; Gilabert‐Estelles 2003; Gilabert‐Estelles 2007; Johnson 2004; Laudanski 2014; Lawson 2008; Morelli 2010; Tang 2009; Tiberi 2010; Wolfler 2005), and two studies demonstrated an unclear risk (Da Silva 2014; Lee 2010). All participants in every study received the same reference standard. The time interval between the index test (sample collection) and the reference standard was established as 12 months or less, and the most commonly reported time interval was immediately before surgery. We assigned a high risk of bias if any patients were excluded from the analysis because of uninterpretable results, inability to undergo either index test or reference standard or if withdrawals were more than 5% of the enrolled population (arbitrary selected cut‐off). Where the time interval was not stated clearly, but the authors' description suggested that the interval was reasonably short, we classified this study as having unclear risk. Thirty‐seven studies presented a low risk of bias in the 'flow and timing' domain (Al‐Jefout 2007; Al‐Jefout 2009; Bokor 2009; Bourlev 2006; Chen 2013; Dheenadayalu 2002; Ding 2010; Hatok 2011; Hudelist 2007; Kim 2007; Klentzeris 1995; Lee 2007; Leslie 2013; Li 2006; Liu 2008; Mafra 2014; Makari 2012; Matsuzaki 2006a; Meibody 2011; Meola 2009; Meola 2013; Mikolajczyk 2006; Pino 2009; Prefumo 2002; Puy 2002; Rakhila 2013; Ramon 2011; Szymanowski 2003; Takahashi 1990; Van der Linden 1994; Van der Linden 1995; Vernet‐Tomas 2006a; Visnovsky 2008; Wang 2010a; Yadav 2013; Zeng 2005; Zubor 2009). We assigned a low risk of bias in this domain where all the patients were included in the analysis, were excluded because they did not meet inclusion criteria prior to execution of index test, or if the withdrawals were less than 5% of the enrolled population and the time interval between the sample collection and reference standard was clearly stated.

Findings

We evaluated a total of 95 endometrial biomarkers in the 54 included studies. Of these, 22 biomarkers had a diagnostic evaluation (Table 1; Figure 5), 77 biomarkers were not altered by the presence of endometriosis (Appendix 7), and four biomarkers exhibited conflicting results (α3β1 and α4β1 integrins, PGP 9.5 and NF). Endometrial biomarkers were evaluated in either whole endometrial tissue or in separate endometrial compartments (stromal cells or glandular epithelium), or in menstrual fluid. The tests were performed in specific phases of the menstrual cycle or irrespective of the cycle phase. Most studies assessed any type of pelvic endometriosis, one study solely focused on DIE (17βHSD2 in Matsuzaki 2006a) and another study compared ovarian endometriosis with other types of non‐malignant ovarian cysts (relaxin and its receptor LGR7 in Morelli 2010).

5.

Forest plot of all endometrial biomarkers evaluated for detection of pelvic endometriosis. Plot shows study‐specific estimates of sensitivity and specificity (squares) with 95% CI (black line) along with country in which the study was conducted, stage of menstrual cycle when tissue was collected, method of sample collection and stage of endometriosis assessed. The studies for each test are ordered according to the year of publication. FN: false negative; FP: false positive; TN: true negative; TP: true positive.

The endometrial biomarkers that were evaluated as a diagnostic test for endometriosis comprise the main findings of this review. The endometrial biomarkers that do not distinguish between women with and without endometriosis represent biomarkers with limited diagnostic potential and have been reported to add to the completeness of the presented topic.

Biomarkers evaluated as a diagnostic test for endometriosis

1. Angiogenesis and growth factors

1.1 PROK‐1 (prokineticin 1)

One European study involving 24 women assessed the value of PROK‐1 in detecting pelvic endometriosis (rASRM I‐IV) (Tiberi 2010). Endometrial samples were collected via the sharp curette technique in the secretory phase of the menstrual cycle. The authors utilised RT‐PCR to assess PROK‐1 levels and demonstrated reduced expression in women affected by endometriosis as compared to disease‐free controls. Using a negative PROK‐1 expression to define endometriosis, the sensitivity was 0.67 (95% CI 0.35 to 0.90), and the specificity was 0.83 (95% CI 0.52 to 0.98), which did not meet the criteria for either replacement or triage tests. Further testing in larger studies including participants at different phases of the menstrual cycle is needed to confirm the role of PROK‐1 in detecting endometriosis.

2. Cell adhesion molecules

Two European studies assessed the diagnostic performance of different integrins, focusing on different sets of molecules (Szymanowski 2003; Vernet‐Tomas 2006a).

2.1 Depolarised α6 (alpha‐6) integrin

One study involving 49 women evaluated the accuracy of depolarised α6 integrin within the glandular epithelium in detecting endometriosis (rASRM I‐IV) (Vernet‐Tomas 2006a). The authors performed the biopsies in the proliferative cycle phase; however, they did not specify the biopsy type. Using immunohistochemistry (IHC), investigators evaluated the percentage of positive glandular cells and the localisation of expression in each section, considering α6 to be polarised when expression was exhibited only on the basal side of the cell and depolarised when expression was observed at any side of the cell. Using depolarised expression as a positive test result, the sensitivity was 0.67 (95 CI 0.47 to 0.83) and specificity was 0.84 (95% CI 0.60 to 0.97). It did not meet the criteria for either replacement or triage test (Figure 6).

6.

Summary ROC plot of tests depolarised α‐6 Integrin and of α3β1, α4β1 and β1 Integrins in glandular and stromal compartments of endometrium for the diagnosis of endometriosis. Each point represents the pair of sensitivity and specificity from a study, the bars correspond to 95% CIs of each individual study. The size of each point is proportional to the study sample size, each shape corresponds to different type of test.

2.2 Integrins α3β1 (alpha‐3 beta‐1), α4β1 (alpha‐4 beta‐1) and β1 (beta‐1)

Another study (32 women, rASRM stage not reported) evaluated the role of three integrins in diagnosing endometriosis (α3β1, α4β1 and β1) (Szymanowski 2003). The investigators performed a sharp curette in the secretory cycle phase and subdivided the specimens into glandular and stromal epithelium. They used IHC to assess integrin expression and defined a positive test as positive staining of any intensity. In glandular epithelium, α3β1 integrin had a sensitivity of 1.00 (95% CI 0.80 to 1.00) and a specificity of 0.27 (95% CI 0.08 to 0.55), and in stromal epithelium, it had a sensitivity of 0.53 (95% CI 0.28 to 0.77) and a specificity of 0.27 (95% CI 0.08 to 0.55). For its part, α4β1 integrin in glandular epithelium demonstrated a sensitivity of 0.65 (95% CI 0.38 to 0.86) and specificity of 0.40 (95% CI 0.16 to 0.68), and in stromal epithelium it exhibited a sensitivity of 0.59 (95% CI 0.33 to 0.82) and a specificity of 0.20 (95% CI 0.4 to 0.48). In glandular epithelium, β1 integrin demonstrated a sensitivity of 0.18 (95% CI 0.04 to 0.43) and a specificity of 0.87 (95% CI 0.60 to 0.98) and in stromal epithelium, it had a sensitivity of 0.76 (95% CI 0.50 to 0.93) and a specificity of 0.00 (95% CI 0.00 to 0.22) (Figure 6). One additional study (16 women, rASRM I, early proliferative cycle phase) showed that α3β1 and α4β1 integrins remained unchanged in endometriosis (Van der Linden 1994). Collectively, the results for all the above integrins were discouraging, but not sufficient to draw conclusions regarding the role of these biomarkers in the detection of endometriosis. These findings require further validation in larger studies with well‐defined populations at different cycle phases across a wide spectrum of the disease.

3. DNA repair and telomere maintenance molecules

3.1 hTERT (human telomerase reverse transcriptase)

One study conducted in South America with a total of 69 women assessed the performance of endometrial hTERT in diagnosing endometriosis (rASRM I‐IV) (Mafra 2014). The authors performed aspiration endometrial biopsies in the secretory phase of the menstrual cycle, using qRT‐PCR to assess hTERT mRNA, which was found to have a higher expression in endometriosis compared to controls. Using positive hTERT mRNA expression as the definition of a positive test, hTERT had a sensitivity of 0.28 (95% CI 0.12 to 0.49) and specificity of 0.80 (95% CI 0.65 to 0.90). The test did not meet the predefined criteria for a replacement or triage test, but further work is required to support this observation.

4. High‐throughput molecular markers

4.1 Endometrial proteome

Two small studies assessed the accuracy of endometrial proteome in detection of endometriosis. Fassbender 2012 took place in Europe (27 women, rASRM I‐IV, secretory cycle phase), and Wang 2010a was in East Asia (26 women, rASRM I‐IV, cycle phase not reported). Both studies performed aspiration biopsies for endometrial sampling and utilised SELDI‐TOF‐MS (surface enhanced laser desorption ionisation‐time of flight mass spectrometry) protein chip array technology for biomarker detection. Each group implemented varying approaches to the data analysis and the construction of a diagnostic model. Each study identified different sets of proteins that were used in a diagnostic model to discriminate the women with and without endometriosis, therefore these two studies were not combined into a meta‐analysis.

Fassbender 2012 identified five peptide peaks of 2072 m/z (mass over charge); 2973 m/z; 3623 m/z; 3680 m/z and 21,133 m/z, reporting a sensitivity of 0.88 (95% CI 0.64 to 0.99) and a specificity of 0.80 (95% CI 0.44 to 0.97), which did not meet the criteria for a replacement or triage test (Figure 7). Wang 2010a observed five protein peaks with different spectral intensities of 5385 m/z, 5425 m/z, 6898 m/z, 5891 m/z and 6448 m/z., with a sensitivity of 0.92 (95% CI 0.64 to 1.00) and a specificity of 0.92 (95% CI 0.64 to 1.00) (Figure 7). The test approached the criteria for a replacement as well as SnOUT and SpIN triage tests. Although promising, this observation provides too few data to draw conclusions regarding the diagnostic role of these endometrial peptides in endometriosis. These results require further validation in large, independent, well‐defined cohorts, using standardised and reproducible methodologies.

7.

Summary ROC plot of tests endometrial proteome and mitochondrial proteome for the diagnosis of endometriosis. Each point represents the pair of sensitivity and specificity from a study, the bars correspond to 95% CIs of each individual study. The size of each point is proportional to the study sample size, each shape corresponds to the type of tested proteome.

4.2 Mitochondrial proteome

One study conducted in East Asia comprising 53 women evaluated diagnostic performance of mitochondrial proteome for pelvic endometriosis (rASRM I‐IV) (Ding 2010). The authors performed aspiration biopsies in unspecified phases of the menstrual cycle and isolated three potential biomarkers for endometriosis within the endometrium. Authors reported mitochondrial proteome to have a sensitivity of 0.88 (95% CI 0.68 to 0.97) and a specificity of 0.86 (95% CI 0.68 to 0.96) (Figure 7) and did not qualify as a replacement or a triage test, although these data are not sufficient to draw a meaningful conclusion.

5. Hormonal markers

5.1 CYP19 (aromatase cytochrome P450)

Eight studies with a total of 444 women determined the accuracy of CYP19 in detecting pelvic endometriosis (Dheenadayalu 2002; Hatok 2011; Hudelist 2007; Johnson 2004; Matsuzaki 2006a; Visnovsky 2008; Wolfler 2005; Zeng 2005). Six studies took place in Europe, one in South America and one in East Asia. Five studies performed the biopsy using a sharp curette, two studies utilised aspiration techniques, and one study did not report the method of sample collection. Five studies assessed endometrium in the proliferative or secretory cycle phase; one study, only in proliferative phase; the menstrual cycle phase was not reported in two studies. Five studies included women with stages I‐IV endometriosis, one study evaluated only stages I‐II, and two studies did not specify severity of the disease. Four studies were of a single‐gate design, of which three studies included women presenting either with pelvic pain or infertility (Visnovsky 2008; Wolfler 2005; Zeng 2005), and one study included only women with pelvic pain (Hatok 2011). The remaining four studies had a two‐gate design and included a symptomatic group as well as symptom‐free women undergoing tubal ligation or gynaecological surgery for other indications (Dheenadayalu 2002; Hudelist 2007; Johnson 2004; Matsuzaki 2006a). Five studies utilised RT‐PCR method for CYP19 detection, and three studies utilised IHC, all demonstrating increased expression of CYP19 in endometriosis. The estimates for sensitivity ranged from 0.60 to 0.90 and for specificity from 0.55 to 0.86. The mean sensitivity and specificity of all these evaluations were 0.78 (95% CI 0.70 to 0.85) and 0.74 (95% CI 0.65 to 0.84), which did not meet the criteria for either a replacement or triage test. Forest plots (Figure 8) and the ROC plot (Figure 9) showed a high degree of heterogeneity for estimates of both sensitivity and specificity, which could be attributed to the differences in patient selection, cycle phase, spectrum of the disease and laboratory methods. None of the individual studies presented diagnostic estimates that met the predetermined diagnostic criteria for a replacement or triage test. These data suggest that CYP19 is either not sensitive or specific enough to be clinically useful in diagnosing endometriosis.

8.

Forest plot of CYP19 for detection of pelvic endometriosis. Plot shows study‐specific estimates of sensitivity and specificity (squares) with 95% CI (black line) along with country in which the study was conducted, stage of menstrual cycle when tissue was collected, method of sample collection and stage of endometriosis assessed. The studies are ordered according to the year of publication. FN: false negative; FP: false positive; TN: true negative; TP: true positive.

9.

Summary ROC plot of CYP19 for the diagnosis of endometriosis. Each point represents the pair of sensitivity and specificity from a study. The size of each point is proportional to the study sample size. The solid black circle represents the mean sensitivity and specificity which is surrounded by a 95% confidence region (dotted line) and 95% prediction region (dashed line).

5.2 17bßHSD2 (17‐β hydroxysteroid dehydrogenase type 2)

One study conducted in Europe (53 women, rASRM stage not reported) evaluated the diagnostic role of 17bßHSD2 in DIE (Matsuzaki 2006a). Aspiration technique was utilised for endometrial biopsy, which was performed in varying stages of the menstrual cycle (proliferative, early‐ mid‐ and late‐secretory). By using qRT‐PCR, the authors demonstrated higher expressions of 17bßHSD2 mRNA in endometriosis in stromal endometrial cells, and there was no statistically significant difference between the study and control groups for glandular epithelium. Interestingly, there was a differential expression of 17bßHSD2 only in secretory cycle phase, but no statistically significant difference in the proliferative phase. Considering both cycle phases, 17HSD2 mRNA in stromal cells had a sensitivity of 0.53 (95% CI 0.34 to 0.72) with a specificity of 0.91 (95% CI 0.72 to 0.99). The test approached the criteria for a SpIN triage test, and possibly could perform better if testing were limited to the secretory cycle phase. However, separate data for only the secretory phase was not available from this paper. Further data from additional studies that include a broad spectrum of disease, clearly defined populations and sub‐analysis per cycle phase is necessary to draw meaningful conclusions.

5.3 ER‐α (oestrogen receptor alpha) and 5.4 ER‐β (oestrogen receptor ‐ beta)

One study conducted in Asia (90 women, rASRM stage not reported) examined endometrial expression of ER‐α and ER‐β in endometriosis. The authors did not specify a method of endometrial sampling and performed the test in either the proliferative or secretory cycle phase in glandular epithelium and in stromal cells. Considering IHC staining of any intensity as a positive test, the diagnostic estimates for ER‐α in either glandular cells (sensitivity 0.73, 95% CI 0.60 to 0.84; specificity 0.43, 95% CI 0.25 to 0.63) or in stromal component (sensitivity 0.77, 95% CI 0.64 to 0.87; specificity 0.50, 95% CI 0.31 to 0.69) did not meet the predetermined criteria for either replacement or triage test and displayed wide confidence intervals (Figure 10). Likewise, testing for ER‐β exhibited unsatisfactory diagnostic performance in glandular cells (sensitivity 0.67, 95% CI 0.53 to 0.78; specificity 0.63, 95% CI 0.44 to 0.80) and in stromal cells (sensitivity 0.63, 95% CI 0.50 to 0.75; specificity 0.73, 95% CI 0.54, 0.88) (Figure 10). The authors also presented separate data for each phase of the menstrual cycle, but the cycle‐specific evaluations did not meet or approach the threshold for a replacement or triage test, and none were superior to the estimates of the tests performed irrespective of the cycle phase (data not shown).

10.

Summary ROC plot of ER‐α and ER‐β in glandular and stromal compartments of endometrium for the diagnosis of endometriosis. Each point represents the pair of sensitivity and specificity from a study, the bars correspond to 95% CIs of each individual study. The size of each point is proportional to the study sample size, each shape corresponds to different type of test.

6. Immune system and inflammatory markers

6.1 IL‐1R2 (interleukin‐1 receptor type II)

One study conducted in North America (71 women, rASRM I‐II) examined the value of IL‐1R2 in detecting endometriosis (Lawson 2008). The authors reported levels of IL‐1R2 in glandular and stromal epithelium in both proliferative and secretory phases of the menstrual cycle. Authors did not specify the method of endometrial sampling. They evaluated IL‐1R2 expression by using fluorescence in situ hybridisation (FISH) and noted a decrease in endometriosis in both types of cells. The cut‐off value of hybridisation score (HS) > 3 demonstrated a comparable diagnostic estimate in glandular epithelium at any cycle stage (sensitivity 1.00, 95% CI 0.79 to 1.00; specificity 0.53, 95% CI 0.27 to 0.79) and when only considering the secretory stage (sensitivity 1.00, 95% CI 0.69 to 1.00; specificity 0.67, 95% CI 0.30 to 0.93). Both tests met the criteria for a SnOUT triage test, although there were wide confidence intervals, especially for specificity values (Figure 11). The cut‐off value of HS > 2 for stromal cells demonstrated a sensitivity of 0.93 (95% CI 0.68 to 1.00) and a specificity of 0.76 (95% CI 0.76 to 0.93) for testing irrespective of cycle phase, which approached the criteria for both replacement and SnOUT triage tests (Figure 11). When considering only the secretory cycle phase and using the same threshold, testing for IL‐1R2 in stromal cells revealed a sensitivity of 0.90 (95% CI 0.55 to 1.00) and a specificity of 0.90 (95% CI 0.55 to 1.00), which approached the criteria for replacement, SnOUT and SpIN triage tests, albeit with wide confidence intervals that undermine the reliability of the observed estimates. Larger diagnostic accuracy studies need to validate these data in a well‐characterised population.

11.

Summary ROC plot of tests IL‐1R2 mRNA (glandular) and IL‐1R2 mRNA (stroma) for the diagnosis of endometriosis. Each point represents the pair of sensitivity and specificity from a study, the bars correspond to 95% CIs of each individual study, different shapes represent different endometrial compartments tested. The size of each point is proportional to the study sample size.

7. Myogenic markers

7.1 Caldesmon (calmodulin binding protein) and 7.2 CALD1 (caldesmon gene)

One study conducted in South America (35 women, rASRM I‐IV) assessed the diagnostic accuracy of caldesmon protein and gene in detecting endometriosis (Meola 2013). The authors sampled the endometrium via a sharp curette in both the proliferative and secretory phases of the menstrual cycle. CALD1 gene was detected by RT‐PCR and caldesmon protein by IHC stain intensity. The phase of menstrual cycle did not influence the gene expression or the protein levels in endometrial tissue. The CALD1 in the proliferative phase had a sensitivity of 0.60 (95 CI 0.36 to 0.81) and a specificity of 0.87 (95% CI 0.60 to 0.98), and in the secretory phase a sensitivity of 0.75 (95% CI 0.51 to 0.91) and a specificity of 0.67 (95% CI 0.38 to 0.88). The test did not meet the diagnostic criteria for a replacement or triage test (Figure 12). Caldesmon protein in the proliferative phase had a sensitivity of 0.95 (95% CI 0.75 to 1.00) with a specificity of 1.00 (95% CI 0.78 to 1.00), meeting the criteria for replacement, SnOUT and SpIN triage tests. In the secretory phase the sensitivity and specificity of caldesmon protein were slightly lower at 0.90 (95% CI 0.68 to 0.99) and 0.93 (95% CI 0.68 to 1.00), respectively, approaching the criteria for replacement, SnOUT and SpIN triage tests (Figure 12). Further work to confirm or refute this observation and determine the value of caldesmon endometrial testing to diagnose endometriosis is warranted.

12.

Summary ROC plot of tests Caldesmon (proliferative), caldesmon (secretory), CALD1 mRNA (proliferative) and CALD1 mRNA (secretory) for the diagnosis of endometriosis. Each point represents the pair of sensitivity and specificity from a study, the bars correspond to 95% CIs of each individual study, different shapes represent different test (protein or mRNA in different endometrial compartments). The size of each point is proportional to the study sample size.

8. Nerve sheath and nerve growth markers

8.1 PGP 9.5 (protein gene product 9.5)

Eight studies involving 429 women evaluated PGP 9.5 for the diagnosis of pelvic endometriosis (Al‐Jefout 2007; Al‐Jefout 2009; Bokor 2009; Elgafor el Sharkwy 2013; Leslie 2013; Makari 2012; Meibody 2011; Yadav 2013). Three studies took place in Australia, two in Europe and one each in Africa, South Asia and Western Asia. Three studies performed endometrial biopsy using different aspiration techniques, and three studies utilised sharp curette. The other two studies did not report the method of endometrial sampling. Three studies assessed endometrium in the proliferative phase of the menstrual cycle; one study, in secretory; and three studies, in any cycle phase. One study did not report the cycle phase. Two studies evaluated only women with minimal to mild endometriosis (rASRM I‐II), four studies assessed a wide range of disease (rASRM I‐IV), and two studies did not report the severity of the disease. All the included studies had a single‐gate design and included women with suspected endometriosis who presented with pelvic pain, infertility or an adnexal mass.

The biopsy specimen underwent IHC staining for PGP 9.5 followed by nerve fibre identification under a digital camera (photomicrography) in all but Leslie 2013, which used conventional light microscopy. The number of small nerve fibres present for the diagnosis varied between 2 and 50 nerve fibres per mm². The estimates for sensitivity ranged from 0.19 to 1.00 and for specificity from 0.50 to 1.00, with substantial heterogeneity between the studies (Figure 13; Figure 14).

13.

Forest plot of PGP 9.5 for detection of pelvic endometriosis. Plot shows study‐specific estimates of sensitivity and specificity (squares) with 95% CI (black line) along with country in which the study was conducted, stage of menstrual cycle when tissue was collected, method of sample collection and stage of endometriosis assessed. The studies are ordered according to the year of publication. FN: false negative; FP: false positive; TN: true negative; TP: true positive.

14.

Summary ROC plot of PGP 9.5 for the diagnosis of endometriosis ( the data is presented for 7 studies, excluding one outlier study (Leslie 2013)). Each point represents the pair of sensitivity and specificity from a study. The size of each point is proportional to the study sample size. The solid black circle represents the mean sensitivity and specificity which is surrounded by a 95% prediction region (dashed line). The outlier study is presented on the graph for descriptive purposes, but not included in the meta‐analysis.

Leslie 2013 (68 women, rASRM I‐IV, proliferative or secretory cycle phase, sample collected with sharp curettage or Pipelle aspiration biopsy) found a markedly lower sensitivity of 0.19 (95% CI 0.09 to 0.33) with a specificity of 0.71 (95% CI 0.48 to 0.89). There are a couple of reasons as to why these authors may have reported a lower sensitivity in comparison to the other studies. The included patients had been and were on hormonal remedies, and 7 of the 68 patients had other pathologies present at the time of laparoscopy. Importantly, the method of endometrial biopsy was not full thickness nor as adequate as reported by other studies. The authors obtained biopsies with a Pipelle and a metal curette, each of which provided only small tissue fragments that could not be specifically orientated for microscopic examination. This study also utilised conventional light microscopy compared to photomicrography, which investigators used in the studies that demonstrated a higher sensitivity. When we excluded this outlier study, the mean sensitivity and specificity (7 studies, 361 women) were 0.96 (95% CI 0.91 to 1.00) and 0.86 (95% CI 0.70 to 1.00), which met the criteria for replacement and SnOUT triage tests.

Alternatively, Al‐Jefout 2007 (37 women, rASRM I‐IV, cycle phase not specified) demonstrated the highest diagnostic estimates (sensitivity 1.00, 95% CI 0.83 to 1.00; specificity 1.00, 95% CI 0.80 to 1.00). A subsequent larger study from the same group (99 women, rASRM I‐IV, any cycle phase) demonstrated a high diagnostic performance (sensitivity 0.98, 95% CI 0.92 to 1.00; specificity 0.83, 95% CI 0.66 to 0.93) (Al‐Jefout 2009). Possible reasons for this high diagnostic performance include meticulous attention to endometrial sampling using the endo sampler, which ensured that operators obtained a narrow but deep sample of endometrium, with adequate amounts of stroma for assessment. This technique appears to be a necessary prerequisite for a successful diagnosis, as nerve fibres are only found in stroma, especially in the deeper portions of the functional layer. The Pipelle biopsy is good for demonstrating structures in the epithelium and superficial stroma but may not be as good for deeper stroma and nerve fibres.

Notably, Cetin 2013 (60 women, rASRM I‐IV, any cycle phase, sample collected with aspiration biopsy) detected no expression of PGP 9.5 in endometrium in either study (N = 31) or control (N = 29) groups, including those presenting with pelvic pain. The main point of difference from all the other studies assessing PGP 9.5 is that Cetin 2013 utilised ready‐to‐use immunohistochemical markers, which can explain the difference in PGP 9.5 expression in this study. The authors observed positive PGP 9.5 staining in the muscle layers of intestinal tissue, which they used as a positive control on each slide, suggesting that the markers used could not be applicable to endometrial tissue. Adequate discrimination of PGP 9.5 seems to be very dependent on good optimisation of only one or two of the commercially available markers. Fine nerve fibres may be difficult to distinguish, so using a bright coloured chromogen, such as 'Fast Red' may help them to stand out. Therefore, in addition to the required technical precision, laboratory technique may hinder/influence the applicability of the test to routine clinical practice. Overall the diagnostic studies for PGP 9.5 are encouraging, but we recommend more comprehensive assessment of this diagnostic tool in large high quality studies performing standardised endometrial sampling and laboratory methods.

8.1 Other neural markers

One European study evaluated the role of other endometrial neural markers, including VIP (vasoactive intestinal polypeptide), CGRP (calcitonin gene related protein), SP (substance P), NPY (neuropeptide Y), NF (neurofilament) and combinations of the above (40 women, rASRM I‐II) (Bokor 2009). While PGP 9.5 stains all nerve fibres, these nerve markers stain for specific nerve fibres and hence can improve the diagnostic accuracy of the endometrial test. The investigators performed aspiration biopsies in the secretory phase of menstrual cycle, evaluating nerve fibre density in the biopsy specimen following IHC. The reported diagnostic estimates for either VIP or SP (sensitivity 0.95, 95% CI 0.75 to 1.00; specificity 0.80, 95% CI 0.56 to 0.94) met the criteria for replacement and SnOUT triage tests. For CGRP, they approached these criteria (sensitivity 0.90, 95% CI 0.68 to 0.99; specificity 0.85, 95% CI 0.62 to 0.97), and for NPY, they met the criteria for a SnOUT triage test (sensitivity 0.95, 95% CI 0.75 to 1.00; specificity 0.65, 95% CI 0.41 to 0.85). The combined test for the panel of neural markers (VIP + PGP 9.5 + SP) showed the highest diagnostic estimates (sensitivity 0.95, 95% CI 0.75 to 1.00; specificity 1.00, 95% CI 0.83 to 1.00) and could qualify for either replacement, SnOUT or SpIN triage tests (Figure 15). In this study, the levels of NF were not significantly different in women with and without endometriosis. Investigators calculated the diagnostic test estimates and observed poor test performance, with a good sensitivity of 0.95 (95% CI 0.75 to 1.00) but unacceptably low specificity of 0.10 (95% CI 0.01 to 0.32). Another study demonstrated no staining with NF in endometrium from women with and without endometriosis (Cetin 2013). The authors reported similar findings for PGP 9.5 and, as presented above, it is quite possible that the assay selected in this study is not sensitive enough for endometrial tissue, hence these findings cannot confirm or refute the previous observations on endometrial expression of PGP 9.5 and NF. The reported alterations in VIP, CGRP, SP and NPY expression in endometriosis indicate that these biomarkers are suitable for detection of the disease, whereas endometrial NF is of little diagnostic value; however, there is insufficient data to draw meaningful conclusions on the findings from a single small study. These results require further validation in large, well‐defined populations across a wide spectrum of disease utilising a standardised, reproducible methodology.

15.

Summary ROC plot of tests VIP, CGRP, SP, NPY, NF and combined test (VIP, PGP9.5, SP) for the diagnosis of endometriosis. Each point represents the pair of sensitivity and specificity from a study, the bars correspond to 95% CIs of each individual study. The size of each point is proportional to the study sample size.

9. Tumour markers

9.1 CA‐125 (cancer antigen‐125) in menstrual fluid

One study conducted in East Asia involving 104 women explored the role of CA‐125 in pelvic endometriosis (rASRM I‐IV) (Takahashi 1990). Investigators assessed the biomarker within the menstrual fluid using radioimmunoassay with commercial kits, demonstrating a sensitivity of 0.66 (95% CI 0.49 to 0.80) and a specificity of 0.89 (95% CI 0.79 to 0.96). Further data is required before commenting on its diagnostic role.

The eligible studies did not assess any other biomarkers in menstrual fluid.

Biomarkers that did not distinguish between women with and without endometriosis

Seventy‐seven biomarkers showed no difference in expression levels between the groups of women with and without endometriosis, as presented in Appendix 7. Most of these biomarkers were assessed in small individual studies, and their association with endometriosis remains unclear. Six biomarkers were assessed in two studies or more and included:

1. CD68 + cells (macrophages) (Cetin 2013;Klentzeris 1995);

2. PAI‐1 (plasminogen activator inhibitor‐1) protein or mRNA (Gilabert‐Estelles 2003; Gilabert‐Estelles 2007);

3. TIMP‐1 (tissue inhibitor of metalloproteinases‐1) protein or mRNA (Chen 2004; Gilabert‐Estelles 2003; Gilabert‐Estelles 2007; Laudanski 2014; Li 2006);

4. TNF‐α (tumour necrosis factor‐alpha) protein or mRNA (Chen 2013; Da Silva 2014);

5. TSP‐1 (thrombospondin‐1) protein or mRNA (Gilabert‐Estelles 2007; Ramon 2011); and

6. VEGF (vascular endothelial growth factor) protein or mRNA (Cho 2012; Da Silva 2014; Lee 2010).

All six biomarkers showed a similar trend of no differential expression between the groups, which indicates their limited value in detection of endometriosis.

Investigations of heterogeneity and sensitivity analysis

We outline potential sources of heterogeneity in Secondary objectives. Although we attempted to assess these sources of heterogeneity, there were not enough studies evaluating each test to make this a meaningful analysis. Furthermore, the sensitivity analyses were not possible due to the small number of studies.

Discussion

Summary of main results

The review presents data from 2729 reproductive‐aged women who had presented with symptoms of endometriosis and underwent eutopic endometrial biopsies during or prior to diagnostic surgery. We evaluated diagnostic performance for 22 endometrial biomarkers and identified 71 biomarkers that were not altered by endometriosis. Most of the biomarkers were assessed in a limited number studies, and only two biomarkers (PGP 9.5 and CYP19) provided sufficient data for a meta‐analysis.

PGP 9.5, the marker of small unmyelinated sensory C nerve fibres, was the most studied biomarker. The meta‐analysis revealed that evaluation of nerve fibre density with PGP 9.5 staining could qualify as a replacement test (mean sensitivity of 0.96, 95% CI 0.91 to 1.00; specificity 0.86, 95% CI 0.70 to 1.00), after exclusion of one outlier study. However, there was a notable variability in reporting of the diagnostic estimates, which could be attributed to a difference in the endometrial sampler used, the method of curettage and the laboratory technique.

Estimates of aromatase cytochrome P450 (CYP19), the key enzyme in conversion of C19 steroids into oestrogen, did not meet the diagnostic criteria to qualify for either a replacement or triage test (mean sensitivity 0.77, 95% CI 0.70 to 0.85; specificity 0.74, 95% CI 0.65 to 0.84).

The findings of the meta‐analyses presented in this review need to be interpreted with caution. Considering both the level of heterogeneity and the high or unclear risk of bias of the included studies, the results do not seem to have sufficient reliability to direct clinical practice.

The remaining biomarkers were classified as follows.

1. Tests to be validated for their diagnostic potential. This group included:

   1. those with an adequate diagnostic performance, but insufficient data to confidently comment on their diagnostic role (fewer than three studies with the diagnostic estimates meeting the criteria for either a replacement or triage test); and

   2. tests where the diagnostic estimates approached the criteria for replacement or triage tests in a small number of studies, with a possibility that they would reach these criteria if further studies were performed (fewer than three studies with the diagnostic estimates within 5% of the criteria for either replacement or triage tests). We present these tests in Table 5.

2. Tests of limited diagnostic value (at least three studies demonstrating low diagnostic estimates that do not meet or approach the criteria for either replacement or triage test, or demonstrating no differential expression in endometriosis). These tests are presented in Appendix 8 and include:

   1. TIMP‐1 (tissue inhibitor of metalloproteinases‐1) protein or mRNA (Chen 2004; Gilabert‐Estelles 2003; Gilabert‐Estelles 2007; Laudanski 2014; Li 2006);

   2. VEGF (vascular endothelial growth factor) protein or mRNA (Cho 2012; Da Silva 2014; Lee 2010). We advise against further evaluation of these biomarkers for the diagnosis of endometriosis;

   3. aromatase cytochrome P450 (CYP19) as suggested by the meta‐analysis presented above.

3. Tests that possibly have limited diagnostic value, but where there are insufficient data to confidently comment on their diagnostic role (fewer than three studies with low diagnostic estimates or demonstrating no differential expression in endometriosis). The full list of tests from this group is presented in Appendix 9. We advise considering further investigation of these biomarkers with a focus on specific phases of the menstrual cycle, specific phenotypes of endometriosis, by implementation of different cut‐off values or by utilising different laboratory methods.

4. Endometrial biomarkers to be validated for their diagnostic potential in endometriosis.

Endometrial biomarkers	Replacement test	SnOUT triage test	SpIN triage test
1. High throughput molecular markers
Endometrial proteome by SELDI‐TOF‐MS (5 protein peaks with molecular weights of 5,385 m/z, 5,425 m/z, 6,898 m/z, 5,891 m/z, 6,448 m/z)	±	±	±
2. Hormonal markers
17βHSD2 mRNA (17‐β hydroxysteroid dehydrogenase type 2 gene)	—	—	±
3. Immune system and inflammatory markers
IL‐1R2 mRNA (interleukin‐1 receptor type II gene) (glandular epithelium)	—	+	—
IL‐1R2 mRNA (interleukin‐1 receptor type II gene) (stromal endometrial cells)	±	±	—
IL‐1R2 mRNA (interleukin‐1 receptor type II gene) (glandular epithelium, secretory phase of menstrual cycle)		+	—
IL‐1R2 mRNA (stromal endometrial cells, secretory phase of menstrual cycle)	±	±	±
4. Myogenic markers (markers of smooth muscle differentiation)
Caldesmon (proliferative phase of menstrual cycle)	+	+	+
Caldesmon (secretory phase of menstrual cycle)	±	±	±
5. Neural and nerve sheath markers
PGP 9.5 (protein gene product 9.5)	+	+	—
VIP (vasoactive intestinal polypeptide)	+	+	—
CGRP (calcitonin gene‐related protein)	±	±	—
SP (substance P)	+	+	—
NPY (neuropeptide Y)	—	+	—
Combined test (VIP + PGP 9.5 + SP)	+	+	+
Notes: + meets the criteria Replacement test: sensitivity ≥ 94% and specificity ≥ 79% SnOUT triage test: sensitivity ≥ 95% and specificity ≥ 50% SpIN triage test: sensitivity ≥ 50% and specificity ≥ 95% ± approaches the criteria (within 5% of the pre‐defined criteria) — does not meet criteria

Strengths and weaknesses of the review

This review is part of a comprehensive review series of minimally invasive biomarkers for the diagnosis of endometriosis. The main strength of this review is its attempt to systematically review the vast number of studies with an aim to present the most accurate picture of diagnostic test accuracy of endometrial tests for endometriosis.

The main strengths of the review are the following.

1. We carried out a thorough search of the current literature, including for studies written in languages other than English.

2. Two or three independent reviewers extracted data and used the modified QUADAS‐2 tool to perform quality assessments.

3. Stringent selection criteria ensured that eligible studies utilised prospectively collected samples and only included women of reproductive age, which minimised the risk of bias in interpreting the reference standard and index test.

4. We approached the authors of the studies in an attempt to obtain any missing information that was required to assess eligibility and critically appraise the studies.

5. We included studies that demonstrated biomarker levels that did not significantly differ in endometriosis, providing a more comprehensive evaluation of the diagnostic role of the biomarkers and identifying the tests of limited value in diagnosing the disease. These biomarkers are evaluated in adequately designed studies and provide information that is likely to focus future research on other more clinically useful tests. Although this information has little influence on the conclusions of this review due to the paucity of the available data, the relevance of these studies will increase in future updates with a growing body of evidence.

The limitations of this review are the following.

1. There were only individual small studies for most evaluated index tests, and a meta‐analysis was possible only for two endometrial biomarkers.

2. The low number of studies for the most studied biomarkers, most of which presented a high or unclear risk of bias, contributed to the low quality of evidence presented in this review and undermined the reliability of the estimates from the meta‐analyses. There was variation between studies with respect to the included populations, the severity of endometriosis, the menstrual cycle phase at testing, endometrial sampling and laboratory methods. We were unable to formally explore sources of heterogeneity due to the low number of studies for each biomarker. There was a loss of patients from the analyses due to poor quality of eutopic endometrium samples. This excluded potentially viable patients from the analysis and further reduced the population size. We now have a standardised methodology available for tissue biospecimen collection, processing and storage, and we recommend adhering to these standards in future diagnostic studies (Fassbender 2014).

3. Thresholds for a positive diagnostic test were not always pre‐specified, and often the definition of a positive test varied for the same test between the studies. Threshold definition was less likely when the primary design of the study was to determine differential regulation in endometriosis rather than diagnostic accuracy.

4. We cannot rule out the presence of publication bias, since we were unable to include unpublished studies due to difficulties in obtaining the data.

5. We cannot rule out the risk of patient selection bias, since most of the studies took place in academic institutions and some in referral centres for endometriosis. The reported prevalence of endometriosis in this review (30% to 77%) was generally higher than previously reported prevalence for endometriosis (6% to 10% in the general female population and 35% to 50% in symptomatic women) (Giudice 2004). This may reflect a high level of surgical diagnostic expertise but could be due to a tertiary referral pattern with possible pre‐selection of more challenging cases. Selection bias appeared to be reduced but not completely eliminated by consecutively enrolling participants. However, the information on method of enrolment was missing in most of the included studies.

6. There was variation in the selection of the case and control groups with inclusion of participants that may not reflect a clinically representative population. In this review 27 studies (50%) had a two‐gate design that included a wide group of participants who underwent surgery for various indications, including either healthy asymptomatic individuals or participants with other pathological conditions within the control group. This could have biased the test outcomes, since such a control group is broader than that normally seen in clinical practice and includes the women that would not be tested in the relevant clinical situation. In this review 15 studies (28%) included women with a limited spectrum of endometriosis, and 9 studies did not report the spectrum of the evaluated condition. We included these studies to avoid omission of potentially valuable diagnostic information, but considering only mild or only severe forms of the disease could skew the diagnostic estimates in either direction and subsequently interfere with the interpretation of the index test results. It was not possible to evaluate population and disease spectrum effects on the data because there were too few reports for most of the evaluated tests.

7. We could not rule out inappropriate assignment to the endometriosis and control groups. Surgical misdiagnosis is a potential cause of bias, as the number and experience of the surgical team, the surgical diagnostic criteria and the surgical methods were poorly described in most of the included studies. We now have a standardised technique for performing laparoscopy, and we recommend that any future studies use this standardised method for undertaking laparoscopy (Becker 2014). Additionally, we did not confine the studies included in this review to those that reported histological confirmation of endometriotic lesions. Although a recent ESHRE guideline stated that evidence is lacking to support laparoscopy without histology to confirm endometriosis (Dunselman 2014), the clinical significance of histological verification remains debatable. Diagnosis by surgical visualisation alone remains a common clinical practice and can be considered reliable when an accurate inspection of the abdominal cavity is performed by experienced surgeons. We chose to include the 25 studies (46%) that reported surgical visualisation as the only reference standard, as we did not wish to lose potentially valuable information by excluding these studies that did not confirm the diagnosis histologically; however, this could impact on the accuracy of assignment to the case and control groups.

8. There are no well‐established criteria for replacement or triage diagnostic tests, therefore we chose the criteria that were both realistic and clinically applicable to assist in the interpretation of the complex results. For a replacement test, we considered the threshold reported by the only systematic review on accuracy of the reference standard (laparoscopy) in the detection of endometriosis to be the most objective (Wykes 2004). The meta‐analysis was published in 2004 and included four eligible studies comprising 433 women. We acknowledge the limitations associated with emphasising a single review, particularly if it does not present the latest and possibly most accurate data that reflect advances in surgical expertise and technology. Several studies on accuracy of laparoscopy in detecting endometriosis have been published in the last decade; however, no reviews have assessed their results in a systematic way. A further systematic analysis to determine the accuracy of laparoscopy was beyond the scope of this review. The criteria for triage tests were based on the common concepts of SnOUT and SpIN in medical statistics, and the cut‐offs were set at levels we considered to be clinically relevant (see Role of index test(s)). We encourage readers to apply independent interpretations of the presented diagnostic estimates using thresholds that may be more applicable to specific populations and clinical circumstances.

Applicability of findings to the review question

In terms of applicability, there was a high concern regarding patient selection in 36 reports, unclear concern in 7 articles and low concern in only 11 studies (20%). We assigned high or unclear concern when the set of participants in the study was broader than that seen in clinical practice or when the spectrum of the target condition was limited, possibly making the findings inapplicable to the review question and clinical practice. Only one study presented an unclear concern with respect to index test applicability. The remaining 53 studies were of low concern regarding implementation of the index test and presented adequate information to conclude that conduct or interpretation of the test did not differ from the review question. We judged applicability of the reference standard to be satisfactory for all studies. However, the majority of included studies took place in academic institutions with a high level of expertise in laboratory techniques, and the index test outcome measures may not be reproducible in all institutions or extrapolated to general practice.

We excluded some potentially relevant, well‐designed studies, as they did not directly address the review question. For example, we excluded studies that reported on biomarkers with differential expression in endometriosis, but without providing enough information to assess the diagnostic performance of the biomarker. We also excluded some forms of endometriosis, such as bladder, ureteric or the forms involving extra‐pelvic sites (e.g. umbilicus, hernia sacs, abdominal wall, lung, kidney, etc.), as they were predominantly reported in case reports or small case series, and diagnostic laparoscopy is not an applicable reference test for these conditions. Although these target conditions are rare, from a clinical perspective the diagnostic options for these forms of endometriosis remain unclear.

Authors' conclusions

Implications for practice.

The ability to diagnose endometriosis with minimally invasive or non‐invasive tests would avoid invasive diagnostic surgery and reduce the potential delay in diagnosis. The ability to perform an endometrial biopsy in a clinical setting would allow earlier diagnosis, with fewer follow‐up visits, speedier institution of effective treatment and expected cost savings benefits for the patient and health service.

The review demonstrated that PGP 9.5, a marker of small nerve fibres, met the criteria for a replacement diagnostic test, but the included studies displayed substantial heterogeneity of diagnostic estimates. The expression of PGP 9.5 appears to be highly dependent on the type of sample collection and laboratory assays. The test requires further validation via utilisation of reproducible universal methodology before meriting a recommendation for its use in clinical practice.

Endometrial proteome (of which five peaks emerged as potential biomarkers for endometriosis) also remains a long‐term promising diagnostic test. Of note, separate research groups identified different sets of biomarkers by utilising a similar method of proteome detection, implying that a standardised analytical approach needs to be established along with further testing in several independent cohorts before making any conclusions.

Additional biomarkers identified in this review, such as 17βHSD2, IL‐1R2, caldesmon, and neural markers VIP, CGRP, SP, NPY displayed diagnostic estimates that qualified for either replacement or triage test, but these results came from individual studies, and hence further research is necessary.

In light of the above, currently there is not enough evidence to recommend any endometrial biomarker for use in clinical practice as a substitute for laparoscopy.

In this review we identified a list of biomarkers that have no value in detection of endometriosis and hence are not recommended for evaluation in future diagnostic studies. This is important for appropriate allocation of research resources and to guide clinically relevant experimental work in the field. These biomarkers include: CYP19, TIMP‐1 and VEGF.

As there is an absence of well‐established criteria for an adequate diagnostic test, we determined the diagnostic criteria for replacement and triage tests in a way that we believe will aid the interpretation for clinically active readers. However, we encourage readers to apply a different and relevant criteria according to each clinical population and situation.

Implications for research.

Currently randomised controlled treatment trials require women with and without endometriosis to have had diagnostic surgery for accurate group allocation. For ethical reasons, therapeutic surgery is usually performed at the same time, potentially biasing treatment trial outcomes. Thus, our current inability to diagnose and assess the progression of endometriosis in a non‐invasive way is a significant limitation in the advancement of clinical research in endometriosis.

Critical appraisal of the studies included in this review has identified several weaknesses in study design that can affect an objective evaluation of the presented findings, and in general, we rated the available evidence as low quality. Although several biomarkers appeared to have high diagnostic estimates, further well‐designed diagnostic studies are required to establish the diagnostic test accuracy and utility of these endometrial tests. We recommend that future authors consider:

1. recruiting a larger pool of participants following power calculations for sample size determination (Liu 2005b);

2. focusing on a single‐gate design that only includes a clinically relevant population (Rutjes 2005);

3. using a diagnostic test accuracy study design that adheres to the STARD reporting standards and includes the 25 items on the checklist and the QUADAS checklist (Bossuyt 2003; Whiting 2011);

4. working on improving techniques of endometrial biopsy, including endocervical analgesia, making further improvements on the biopsy cannulas and possibly developing endometrial brushes for superficial sampling of the endometrium (although this might not be suited for markers that are only expressed within the stroma);

5. developing a simplified and improved detection technique for neuronal immunohistochemical biomarkers, such as PGP 9.5, by utilising digitally enhanced assessment of differences in IHC appearances;

6. establishing universally acceptable laboratory methodologies and diagnostic criteria for a positive test, and formally assessing the inter‐ and intraobserver variability of the laboratory methods (Fassbender 2014);

7. utilising universally acceptable methods of performing laparoscopy as the reference standard test (Becker 2014);

8. undertaking direct comparisons of promising tests in conjunction with a cost‐effectiveness analysis;

9. applying testing to different clinical phenotypes rather than to women classified according to rASRM staging (Vitonis 2014);

10. assessing the long‐term outcomes and lifetime healthcare costs of women that have participated in diagnostic test accuracy trials of specific diagnostic tests.

Notes

The Cochrane review group split the initially planned single review on the non‐invasive tests for diagnosis of endometriosis into several smaller reviews in order to facilitate data handling and interpretation, due to abundance and diversity of the suggested tests. All of the reviews in the series drew from a common protocol designed for the purpose. The other reviews from the series include 'Blood biomarkers for the non‐invasive diagnosis of endometriosis', 'Urinary biomarkers for the non‐invasive diagnosis of endometriosis', 'Imaging modalities for the non‐invasive diagnosis of endometriosis' and 'Combined biomarkers for the non‐invasive diagnosis of endometriosis', which summarises the review findings of the series.

Appendices

Appendix 1. Alphabetical index of endometrial biomarkers

	Biomarker	Biological group	Biological sub‐group
1	15‐PGDH mRNA (15‐hydroxyprostaglandin dehydrogenase)	Mediators of prostaglandin biosynthesis	—
2	17βHSD2 (17‐β hydroxysteroid dehydrogenase type 2)	Hormonal markers	—
3	Akr1B1 mRNA (aldoketoreductase ‐1B1, PGF2a synthase)	Mediators of prostaglandin biosynthesis	—
4	Akr1C3 mRNA (aldoketoreductase ‐1C3, PGF2a synthase)	Mediators of prostaglandin biosynthesis	—
5	AKT1 (RAC‐alpha serine/threonine‐protein kinase)	Transcription factors and signalling molecules	—
6	Bax (BCL2‐associated X protein)	Apoptosis markers and regulators	—
7	Bcl‐xL (B‐cell lymphoma‐extra large, or BCL2‐like 1 isoform 1)	Apoptosis markers and regulators	—
8	Bcl‐xL:Bcl‐xS ratio (ratio B‐cell lymphoma‐extra large/B‐cell lymphoma‐extra small)	Apoptosis markers and regulators	—
9	BW 495/36, endometrial epithelial marker	Cell proliferation markers	—
10	CD 4+ (T helper cells)	Immune system and inflammatory markers	Immune cells: peripheral blood mononuclear cells (PBMC)
11	CD 8+ (T suppressor cells)	Immune system and inflammatory markers	Immune cells: PBMC
12	CD 16+ (natural killer cells)	Immune system and inflammatory markers	Immune cells: PBMC
13	CD 22+ (B‐lymphocytes)	Immune system and inflammatory markers	Immune cells: PBMC
14	CD 38+ (granulated lymphocytes)	Immune system and inflammatory markers	Immune cells: PBMC
15	CD 56+ (granulated lymphocytes)	Immune system and inflammatory markers	Immune cells: PBMC
16	CD 68+ (macrophages)	Immune system and inflammatory markers	Immune cells: PBMC
17	CA‐125 (cancer antigen‐125) in menstrual fluid	Tumour markers	—
18	CALD1 (gene encoding for caldesmon)	Myogenic markers	—
19	Caldesmon (calmodulin binding protein)	Myogenic markers	—
20	Cdc2 (cyclin dependent kinase‐2)	Cell cycle regulatory molecules	—
21	CGRP (calcitonin gene‐related protein)	Nerve sheath and nerve growth markers	—
22	CK19 or CYFRA 21‐1 (cytokeratin 19)	Cytoskeleton molecules	—
23	Cox‐1 mRNA (cyclo‐oxygenase‐1)	Mediators of prostaglandin biosynthesis	—
24	cPGES mRNA (cytosolic PGE2 synthase)	Mediators of prostaglandin biosynthesis	—
25	Cyclin B1	Cell cycle regulatory molecules	—
26	CYP19 (aromatase cytochrome P450)	Hormonal markers	—
27	Cytokeratin 18	Cytoskeleton molecules
28	Depolarised α6 Integrin	Cell adhesion molecules and other matrix‐related proteins	—
29	E‐cadherin	Cell adhesion molecules and other matrix‐related proteins	—
30	EGF (epidermal growth factor)	Angiogenesis and growth factors and their receptors	—
31	Endometrial proteome	High throughput markers	—
32	EST (oestrogen sulphotransferase)	Hormonal markers	—
33	ER ‐ α (oestrogen receptor ‐ alpha)		—
34	ER ‐ β (oestrogen receptor ‐ beta)		—
35	FGF‐2 (fibroblast growth factor‐2)	Angiogenesis and growth factors and their receptors	—
36	glycodelin A (PP14 or PAEP) ((placental protein 14 or progestogen‐associated endometrial protein))	Angiogenesis and growth factors and their receptors	—
37	hBD‐2 (human b‐defensin‐2)	Other peptides and proteins	—
38	hTERT (human telomerase reverse transcriptase)	DNA‐repair and Telomer maintenance molecules	—
39	ICAM‐1 (intercellular adhesion molecule‐1) or sICAM‐1 (soluble form of intercellular adhesion molecule‐1)	Cell adhesion molecules and other matrix‐related proteins	—
40	IL‐11	Immune system and inflammatory markers	Interleukins
41	IL‐1R1 (interleukin‐1 receptor type II)	Immune system and inflammatory markers	Interleukins
42	IL‐1R2 (interleukin‐1 receptor type II)	Immune system and inflammatory markers	Interleukins
43	IL‐1β	Immune system and inflammatory markers	Interleukins
44	JAG1 (jagged‐1 protein)	Transcription factors and signalling molecules	—
45	Ki‐67 (antigen KI‐67 or MKI67, marker of cellular proliferation)	Cell proliferation markers	—
46	LAMA5 (laminin subunit alpha‐5)	Cell adhesion molecules and other matrix‐related proteins	—
47	LFA‐3 (CD58) (leukocyte function associated molecule‐3)	Cell adhesion molecules and other matrix‐related proteins	—
48	LGR7 (leucine‐rich G protein‐coupled receptor 7), relaxin receptor	Hormonal markers	—
49	LIF (leukaemia‐inhibitory factor)	Immune system and inflammatory markers	Cytokines
50	Mitochondrial proteome	High throughput markers	—
51	MMP‐1 (matrix metalloproteinase‐1)	Cell adhesion molecules and other matrix‐related proteins	—
52	MMP‐9 (matrix metalloproteinase‐9)	Cell adhesion molecules and other matrix‐related proteins	—
53	MPO (myeloperoxidase)	Immune system and inflammatory markers	other immune/inflammatory markers
54	mRNAome (mRNA micro‐array)	High throughput markers	—
55	NAG (N‐acetyl‐b‐D‐Glucosaminidase)	Immune system and inflammatory markers	other immune/inflammatory markers
56	NF (neurofilament)	Nerve sheath and nerve growth markers	—
57	NPY (neuropeptide Y)	Nerve sheath and nerve growth markers	—
58	OPN (osteopontin)	Cell adhesion molecules and other matrix‐related proteins	—
59‐61	PAI‐1/‐2/‐3 (plasminogen activator inhibitors 1/2/3)	Cell adhesion molecules and other matrix‐related proteins	—
62	PDGF (platelet derived growth factor)	Angiogenesis and growth factors and their receptors	—
63	PGP 9.5 (protein gene product 9.5)	Nerve sheath and nerve growth markers	—
64	PIGF (Placental growth factor)	Angiogenesis and growth factors and their receptors	—
65	PKR1 (prokineticin receptor 1), EG‐VEGF receptor	Angiogenesis and growth factors and their receptors	—
66	PKR2 (prokineticin receptor 2), EG‐VEGF receptor	Angiogenesis and growth factors and their receptors	—
67	Plk1 (polo‐like kinase‐1)	Cell cycle regulatory molecules	—
68	PROK‐1 (prokineticin 1)	Angiogenesis and growth factors and their receptors	—
69	Relaxin	Hormonal markers	—
70	SP (substance P)	Nerve sheath and nerve growth markers	—
71	Telomerase activity	DNA‐repair and Telomer maintenance molecules	—
72	TIMP‐1 (tissue inhibitor of metalloproteinases)	Cell adhesion molecules and other matrix‐related proteins	—
73	TNF‐α (tumour necrosis factor alpha)	Immune system and inflammatory markers	Cytokines
74	tPA (tissue‐type plasminogen activator)	Cell adhesion molecules and other matrix‐related proteins	—
75	TSP‐1 (thrombospondin‐1)	Angiogenesis and growth factors and their receptors	—
76	TYMP (thymidine phosphorylase)	Angiogenesis and growth factors and their receptors	—
77	uPA (urokinase‐type plasminogen activator)	Cell adhesion molecules and other matrix‐related proteins	—
78	VCAM‐1 (CD106) (vascular cell adhesion molecule‐1)	Cell adhesion molecules and other matrix‐related proteins	—
79	VEGF (vascular endothelial growth factor)	Angiogenesis and growth factors and their receptors	—
80	Vimentin	Cytoskeleton molecules	—
81	VIP (vasoactive intestinal polypeptide)	Nerve sheath and nerve growth markers	—
82	α2β1 integrin	Cell adhesion molecules and other matrix‐related proteins	—
83	α3β1 integrin	Cell adhesion molecules and other matrix‐related proteins	—
84	α4β1 integrin	Cell adhesion molecules and other matrix‐related proteins	—
85	α5β1 integrin	Cell adhesion molecules and other matrix‐related proteins	—
86	α6 β1 integrin	Cell adhesion molecules and other matrix‐related proteins	—
87	αVβ3 integrin	Cell adhesion molecules and other matrix‐related proteins	—
88	αVβ5 integrins	Cell adhesion molecules and other matrix‐related proteins	—
89	αVβ6 integrins	Cell adhesion molecules and other matrix‐related proteins	—
90	β1 integrin	Cell adhesion molecules and other matrix‐related proteins	—

Appendix 2. Search strategy for CENTRAL (OVID platform)

Database: EBM Reviews ‐ Cochrane Central Register of Controlled Trials <July 2015 (3.09.2015)>

Search Strategy:

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

1 (biomarker$ or marker$).tw. (23692)

2 Laboratory Test$.tw. (2793)

3 growth factor$.tw. (5448)

4 scatter factor$.tw. (8)

5 cytokine$.tw. (6264)

6 hepatocyte growth factor.tw. (111)

7 (FGF or fibroblast growth factor$).tw. (433)

8 (PDGF or platelet derived growth factor$).tw. (250)

9 (EGF or epidermal growth factor$).tw. (1077)

10 (IGF‐I or insulin‐like growth factor$ or IGF1).tw. (2132)

11 (TGF‐a or transforming growth factor alfa or TGFa).tw. (519)

12 (TGF‐b or transforming growth factor beta or TGFb).tw. (236)

13 (EGFR or epidermal growth factor receptor$).tw. (1905)

14 (VEGF or vascular endothelial growth factor$).tw. (1532)

15 exp Luteinizing Hormone/bl [Blood] (151)

16 leptin$.tw. (1399)

17 exp Progesterone/bl [Blood] (58)

18 Proteolytic enzyme$.tw. (136)

19 exp matrix metalloproteinase 1/ or exp matrix metalloproteinase 2/ or exp matrix metalloproteinase 3/ or exp matrix metalloproteinase 9/ (292)

20 matrix metalloproteinase$.tw. (676)

21 MMP$.tw. (905)

22 TIMP$.tw. (229)

23 exp "tissue inhibitor of metalloproteinase‐1"/ or exp "tissue inhibitor of metalloproteinase‐2"/ (101)

24 exp Glycoproteins/ (10108)

25 (Ca‐125 or Ca125 or cancer antigen 125).tw. (305)

26 (Ca‐19‐9 or Ca19‐9 or cancer antigen 19‐9).tw. (71)

27 (PP 14 or PP14).tw. (23)

28 serum placental protein$.tw. (6)

29 exp Follistatin/ (13)

30 Osteopontin$.tw. (80)

31 exp intercellular adhesion molecule‐1/ or exp selectins/ (929)

32 soluble intercellular adhesion.tw. (256)

33 Soluble adhesion molecule$.tw. (89)

34 sICAM.tw. (319)

35 sVCAM$.tw. (223)

36 (sEcadherin or soluble E‐cadherin).tw. (4)

37 (sEselectin or soluble E‐selectin).tw. (99)

38 exp t‐lymphocytes/ or exp natural killer t‐cells/ (2645)

39 Immune cells alteration$.tw. (1)

40 (T helper$ or T supressor$ or T helper$ T supressor$ ratio).tw. (445)

41 Total complement level$.tw. (0)

42 Autoantibodies.tw. (428)

43 exp Antibodies, Antiphospholipid/ (85)

44 Anti‐endometrial.tw. (0)

45 Antiphospholipid$.tw. (152)

46 exp hla antigens/ or exp hla‐a1 antigen/ or exp hla‐a2 antigen/ (563)

47 (HLA or human leucocyte antigen$).tw. (1724)

48 Anti‐laminin‐1.tw. (0)

49 Anti‐thyroid.tw. (49)

50 Anti‐Thomsen Friedenreich antigen$.tw. (0)

51 Anti‐transferrin.tw. (0)

52 Anti‐LDL.tw. (3)

53 (Anti‐2HSG or Heremans‐Schmidt glycoprotein).tw. (0)

54 interleukin$.tw. (7276)

55 (MCP‐I or monocyte chemoattractant protein‐I).tw. (0)

56 (MIF or migration inhibitory factor$).tw. (75)

57 (TNF‐a or tumour necrosis factor$ alfa).tw. (3923)

58 Fas ligand$.tw. (47)

59 Endometrial marker$.tw. (2)

60 CAMs.tw. (53)

61 cell adhesion molecule$.tw. (568)

62 exp Integrins/ (781)

63 Integrin$.tw. (248)

64 Selectin$.tw. (2183)

65 Cadherin$.tw. (71)

66 Aromatase P450.tw. (3)

67 estrogen receptor$.tw. (1252)

68 progesterone receptor$.tw. (531)

69 MTMMP$.tw. (0)

70 cyr61.tw. (1)

71 exp Cysteine‐Rich Protein 61/ (1)

72 cysteine‐rich heparin‐binding protein$.tw. (0)

73 (ANXA 1 or ANXA1).tw. (3)

74 (Annexin 1 or Annexin1).tw. (2)

75 (PGP 9?5 or PGP9?5 or protein gene product$).tw. (18)

76 serum marker$.tw. (411)

77 neural marker$.tw. (9)

78 cell surface marker$.tw. (46)

79 inflammatory marker$.tw. (1739)

80 microarray$.tw. (501)

81 microRNA$.tw. (103)

82 proteomic$.tw. (176)

83 genomic$.tw. (526)

84 (endometri$ adj2 biops$).tw. (464)

85 Follistatin$.tw. (26)

86 Vascular Endothelial Growth Factor A/ (560)

87 Vitamin D‐Binding Protein/ (18)

88 exp Cytokines/ (13960)

89 exp interleukins/ or exp interleukin‐1/ or exp interleukin‐6/ or exp interleukin‐8/ or exp interleukin‐12/ or exp interleukin‐13/ (4413)

90 exp Epidermal Growth Factor/ (91)

91 exp Fibroblast Growth Factors/ (197)

92 Platelet‐Derived Growth Factor/ (99)

93 Keratin‐19/ (19)

94 exp Clinical Laboratory Techniques/ (35164)

95 (Luteinizing Hormone$ or LH).tw. (2935)

96 cytokeratin‐19.tw. (25)

97 (VDBP or vitamin D‐binding protein$).tw. (44)

98 urinary peptide$.tw. (8)

99 VDBP‐Cr.tw. (0)

100 urinary VDBP corrected for creatinine expression.tw. (0)

101 urinary marker$.tw. (67)

102 or/1‐101 (90390)

103 Endometriosis/di [Diagnosis] (6)

104 102 or 103 (90394)

105 exp Endometriosis/ (469)

106 Endometrio$.tw. (1026)

107 105 or 106 (1067)

108 104 and 107 (226)

109 (animals not (humans and animals)).sh. (1)

110 108 not 109 (226)

Appendix 3. Search strategy for MEDLINE (OVID platform)

Database: MEDLINE (Ovid) <1946 to February, week 2 2015 (16.2.2015)> >

Ovid MEDLINE(R) In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE(R) Daily and Ovid MEDLINE(R)

Search Strategy:

1 (biomarker$ or marker$).tw. (605002)

2 Laboratory Test$.tw. (29839)

3 growth factor$.tw. (272049)

4 scatter factor$.tw. (1287)

5 cytokine$.tw. (250618)

6 hepatocyte growth factor.tw. (8053)

7 (FGF or fibroblast growth factor$).tw. (31798)

8 (PDGF or platelet derived growth factor$).tw. (19864)

9 (EGF or epidermal growth factor$).tw. (58069)

10 (IGF‐I or insulin‐like growth factor$ or IGF1).tw. (43539)

11 (TGF‐a or transforming growth factor alfa or TGFa).tw. (281)

12 (TGF‐b or transforming growth factor beta or TGFb).tw. (28842)

13 (EGFR or epidermal growth factor receptor$).tw. (41719)

14 (VEGF or vascular endothelial growth factor$).tw. (53588)

15 exp Luteinizing Hormone/bl (Blood) (24587)

16 leptin$.tw. (24994)

17 exp Progesterone/bl (Blood) (18412)

18 Proteolytic enzyme$.tw. (9768)

19 exp matrix metalloproteinase 1/ or exp matrix metalloproteinase 2/ or exp matrix metalloproteinase 3/ or exp matrix metalloproteinase 9/ (22968)

20 matrix metalloproteinase$.tw. (34522)

21 MMP$.tw. (44439)

22 TIMP$.tw. (10777)

23 exp "tissue inhibitor of metalloproteinase‐1"/ or exp "tissue inhibitor of metalloproteinase‐2"/ (6146)

24 exp Glycoproteins/ (637149)

25 (Ca‐125 or Ca125 or cancer antigen 125).tw. (6761)

26 (Ca‐19‐9 or Ca19‐9 or cancer antigen 19‐9).tw. (4194)

27 (PP 14 or PP14).tw. (229)

28 serum placental protein$.tw. (33)

29 exp Follistatin/ (1134)

30 Osteopontin$.tw. (6769)

31 exp intercellular adhesion molecule‐1/ or exp selectins/ (25302)

32 soluble intercellular adhesion.tw. (1588)

33 Soluble adhesion molecule$.tw. (779)

34 sICAM.tw. (2258)

35 sVCAM$.tw. (1277)

36 (sEcadherin or soluble E‐cadherin).tw. (95)

37 (sEselectin or soluble E‐selectin).tw. (689)

38 exp t‐lymphocytes/ or exp natural killer t‐cells/ (272580)

39 Immune cells alteration$.tw. (1)

40 (T helper$ or T supressor$ or T helper$ T supressor$ ratio).tw. (21275)

41 Total complement level$.tw. (23)

42 Autoantibodies.tw. (33457)

43 exp Antibodies, Antiphospholipid/ (7522)

44 Anti‐endometrial.tw. (23)

45 Antiphospholipid$.tw. (9974)

46 exp hla antigens/ or exp hla‐a1 antigen/ or exp hla‐a2 antigen/ (64462)

47 (HLA or human leucocyte antigen$).tw. (80501)

48 Anti‐laminin‐1.tw. (33)

49 Anti‐thyroid.tw. (1414)

50 Anti‐Thomsen Friedenreich antigen$.tw. (6)

51 Anti‐transferrin.tw. (275)

52 Anti‐LDL.tw. (181)

53 (Anti‐2HSG or Heremans‐Schmidt glycoprotein).tw. (3)

54 interleukin$.tw. (175195)

55 (MCP‐I or monocyte chemoattractant protein‐I).tw. (44)

56 (MIF or migration inhibitory factor$).tw. (4479)

57 (TNF‐a or tumour necrosis factor$ alfa).tw. (1344)

58 Fas ligand$.tw. (6032)

59 Endometrial marker$.tw. (11)

60 CAMs.tw. (1756)

61 cell adhesion molecule$.tw. (20903)

62 exp Integrins/ (44414)

63 Integrin$.tw. (39960)

64 Selectin$.tw. (55426)

65 Cadherin$.tw. (20780)

66 Aromatase P450.tw. (180)

67 estrogen receptor$.tw. (38819)

68 progesterone receptor$.tw. (16623)

69 MTMMP$.tw. (7)

70 cyr61.tw. (559)

71 exp Cysteine‐Rich Protein 61/ (386)

72 cysteine‐rich heparin‐binding protein$.tw. (9)

73 (ANXA 1 or ANXA1).tw. (313)

74 (Annexin 1 or Annexin1).tw. (339)

75 (PGP 9?5 or PGP9?5 or protein gene product$).tw. (2096)

76 serum marker$.tw. (5429)

77 neural marker$.tw. (925)

78 cell surface marker$.tw. (4456)

79 inflammatory marker$.tw. (10916)

80 microarray$.tw. (75404)

81 microRNA$.tw. (29731)

82 proteomic$.tw. (45292)

83 genomic$.tw. (190985)

84 (endometri$ adj2 biops$).tw. (3411)

85 Follistatin$.tw. (1663)

86 Vascular Endothelial Growth Factor A/ (35738)

87 Vitamin D‐Binding Protein/ (1282)

88 exp Cytokines/ (547522)

89 exp interleukins/ or exp interleukin‐1/ or exp interleukin‐6/ or exp interleukin‐8/ or exp interleukin‐12/ or exp interleukin‐13/ (188479)

90 exp Epidermal Growth Factor/ (21298)

91 exp Fibroblast Growth Factors/ (25075)

92 Platelet‐Derived Growth Factor/ (11030)

93 Keratin‐19/ (1090)

94 exp Clinical Laboratory Techniques/ (2132820)

95 (Luteinizing Hormone$ or LH).tw. (56679)

96 cytokeratin‐19.tw. (1469)

97 (VDBP or vitamin D‐binding protein$).tw. (1158)

98 urinary peptide$.tw. (137)

99 VDBP‐Cr.tw. (1)

100 urinary VDBP corrected for creatinine expression.tw. (1)

101 urinary marker$.tw. (638)

102 or/1‐101 (4086291)

103 Endometriosis/di (Diagnosis) (3354)

104 102 or 103 (4088946)

105 exp Endometriosis/ (17244)

106 Endometrio$.tw. (21492)

107 105 or 106 (24940)

108 104 and 107 (10490)

109 (animals not (humans and animals)).sh. (3892900)

110 108 not 109 (10113)

Additional search February 2015 ‐ May 2015

Ovid MEDLINE(R) In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE(R) Daily and Ovid MEDLINE(R) <1946 to Present (3.9.2015)>

Search Strategy:

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

1 (biomarker$ or marker$).tw. (652345)

2 Laboratory Test$.tw. (31389)

3 growth factor$.tw. (287701)

4 scatter factor$.tw. (1326)

5 cytokine$.tw. (267766)

6 hepatocyte growth factor.tw. (8585)

7 (FGF or fibroblast growth factor$).tw. (33674)

8 (PDGF or platelet derived growth factor$).tw. (20842)

9 (EGF or epidermal growth factor$).tw. (61625)

10 (IGF‐I or insulin‐like growth factor$ or IGF1).tw. (45386)

11 (TGF‐a or transforming growth factor alfa or TGFa).tw. (306)

12 (TGF‐b or transforming growth factor beta or TGFb).tw. (30559)

13 (EGFR or epidermal growth factor receptor$).tw. (46446)

14 (VEGF or vascular endothelial growth factor$).tw. (58203)

15 exp Luteinizing Hormone/bl [Blood] (24870)

16 leptin$.tw. (26783)

17 exp Progesterone/bl [Blood] (18699)

18 Proteolytic enzyme$.tw. (9992)

19 exp matrix metalloproteinase 1/ or exp matrix metalloproteinase 2/ or exp matrix metalloproteinase 3/ or exp matrix metalloproteinase 9/ (24504)

20 matrix metalloproteinase$.tw. (37055)

21 MMP$.tw. (47849)

22 TIMP$.tw. (11419)

23 exp "tissue inhibitor of metalloproteinase‐1"/ or exp "tissue inhibitor of metalloproteinase‐2"/ (6447)

24 exp Glycoproteins/ (662211)

25 (Ca‐125 or Ca125 or cancer antigen 125).tw. (7058)

26 (Ca‐19‐9 or Ca19‐9 or cancer antigen 19‐9).tw. (4399)

27 (PP 14 or PP14).tw. (232)

28 serum placental protein$.tw. (34)

29 exp Follistatin/ (1180)

30 Osteopontin$.tw. (7267)

31 exp intercellular adhesion molecule‐1/ or exp selectins/ (26225)

32 soluble intercellular adhesion.tw. (1663)

33 Soluble adhesion molecule$.tw. (795)

34 sICAM.tw. (2374)

35 sVCAM$.tw. (1360)

36 (sEcadherin or soluble E‐cadherin).tw. (97)

37 (sEselectin or soluble E‐selectin).tw. (713)

38 exp t‐lymphocytes/ or exp natural killer t‐cells/ (284378)

39 Immune cells alteration$.tw. (1)

40 (T helper$ or T supressor$ or T helper$ T supressor$ ratio).tw. (22494)

41 Total complement level$.tw. (24)

42 Autoantibodies.tw. (35161)

43 exp Antibodies, Antiphospholipid/ (7759)

44 Anti‐endometrial.tw. (22)

45 Antiphospholipid$.tw. (10351)

46 exp hla antigens/ or exp hla‐a1 antigen/ or exp hla‐a2 antigen/ (66724)

47 (HLA or human leucocyte antigen$).tw. (83856)

48 Anti‐laminin‐1.tw. (33)

49 Anti‐thyroid.tw. (1478)

50 Anti‐Thomsen Friedenreich antigen$.tw. (8)

51 Anti‐transferrin.tw. (284)

52 Anti‐LDL.tw. (183)

53 (Anti‐2HSG or Heremans‐Schmidt glycoprotein).tw. (3)

54 interleukin$.tw. (184697)

55 (MCP‐I or monocyte chemoattractant protein‐I).tw. (46)

56 (MIF or migration inhibitory factor$).tw. (4718)

57 (TNF‐a or tumour necrosis factor$ alfa).tw. (1428)

58 Fas ligand$.tw. (6204)

59 Endometrial marker$.tw. (11)

60 CAMs.tw. (1823)

61 cell adhesion molecule$.tw. (22033)

62 exp Integrins/ (46487)

63 Integrin$.tw. (42447)

64 Selectin$.tw. (58540)

65 Cadherin$.tw. (22688)

66 Aromatase P450.tw. (182)

67 estrogen receptor$.tw. (41210)

68 progesterone receptor$.tw. (17437)

69 MTMMP$.tw. (7)

70 cyr61.tw. (620)

71 exp Cysteine‐Rich Protein 61/ (425)

72 cysteine‐rich heparin‐binding protein$.tw. (9)

73 (ANXA 1 or ANXA1).tw. (355)

74 (Annexin 1 or Annexin1).tw. (358)

75 (PGP 9?5 or PGP9?5 or protein gene product$).tw. (2190)

76 serum marker$.tw. (5721)

77 neural marker$.tw. (1026)

78 cell surface marker$.tw. (4751)

79 inflammatory marker$.tw. (12244)

80 microarray$.tw. (81764)

81 microRNA$.tw. (35967)

82 proteomic$.tw. (49911)

83 genomic$.tw. (205064)

84 (endometri$ adj2 biops$).tw. (3518)

85 Follistatin$.tw. (1762)

86 Vascular Endothelial Growth Factor A/ (38477)

87 Vitamin D‐Binding Protein/ (1356)

88 exp Cytokines/ (575020)

89 exp interleukins/ or exp interleukin‐1/ or exp interleukin‐6/ or exp interleukin‐8/ or exp interleukin‐12/ or exp interleukin‐13/ (197567)

90 exp Epidermal Growth Factor/ (21875)

91 exp Fibroblast Growth Factors/ (26259)

92 Platelet‐Derived Growth Factor/ (11355)

93 Keratin‐19/ (1179)

94 exp Clinical Laboratory Techniques/ (2203416)

95 (Luteinizing Hormone$ or LH).tw. (57796)

96 cytokeratin‐19.tw. (1538)

97 (VDBP or vitamin D‐binding protein$).tw. (1262)

98 urinary peptide$.tw. (148)

99 VDBP‐Cr.tw. (1)

100 urinary VDBP corrected for creatinine expression.tw. (1)

101 urinary marker$.tw. (679)

102 or/1‐101 (4283825)

103 Endometriosis/di [Diagnosis] (3449)

104 102 or 103 (4286552)

105 exp Endometriosis/ (17833)

106 Endometrio$.tw. (22478)

107 105 or 106 (26003)

108 104 and 107 (10936)

109 (animals not (humans and animals)).sh. (4004321)

110 108 not 109 (10539)

111 (201501$ or 201502$ or 201503$ or 201504$).ed. (322721)

112 110 and 111 (215)

Appendix 4. Search strategy for EMBASE (OVID platform)

Database: EMBASE (Ovid) <1980 to 2015 Week 07 (16.02.2015)>

Search strategy:

1 Laboratory Test$.tw. (41662)

2 growth factor$.tw. (318593)

3 scatter factor$.tw. (1388)

4 cytokine$.tw. (322134)

5 hepatocyte growth factor.tw. (9594)

6 (FGF or fibroblast growth factor$).tw. (37191)

7 (PDGF or platelet derived growth factor$).tw. (23530)

8 (EGF or epidermal growth factor$).tw. (69553)

9 (IGF‐I or insulin‐like growth factor$ or IGF1).tw. (49806)

10 (TGF‐a or transforming growth factor alfa or TGFa).tw. (542)

11 (TGF‐b or transforming growth factor beta or TGFb).tw. (30820)

12 (EGFR or epidermal growth factor receptor$).tw. (64664)

13 (VEGF or vascular endothelial growth factor$).tw. (73191)

14 exp luteinizing hormone/ec (Endogenous Compound) (21924)

15 leptin$.tw. (32576)

16 exp progesterone blood level/ or exp progesterone urine level/ (6285)

17 Proteolytic enzyme$.tw. (9643)

18 exp matrix metalloproteinase/ (19364)

19 matrix metalloproteinase$.tw. (41445)

20 MMP$.tw. (58466)

21 TIMP$.tw. (14174)

22 exp "tissue inhibitor of metalloproteinase 2"/ (4824)

23 exp "tissue inhibitor of metalloproteinase 1"/ (8779)

24 exp glycoprotein/ec (Endogenous Compound) (246077)

25 (Ca‐125 or Ca125 or cancer antigen 125).tw. (9536)

26 (Ca‐19‐9 or Ca19‐9 or cancer antigen 19‐9).tw. (6054)

27 (PP 14 or PP14).tw. (244)

28 serum placental protein$.tw. (43)

29 exp follistatin/ (2148)

30 Osteopontin$.tw. (8475)

31 exp intercellular adhesion molecule 1/ (32066)

32 exp selectin/ (3082)

33 soluble intercellular adhesion.tw. (1788)

34 Soluble adhesion molecule$.tw. (919)

35 sICAM.tw. (2888)

36 sVCAM$.tw. (1793)

37 (sEcadherin or soluble E‐cadherin).tw. (120)

38 (sEselectin or soluble E‐selectin).tw. (822)

39 exp T lymphocyte/ (374675)

40 exp natural killer T cell/ (5800)

41 Immune cells alteration$.tw. (6)

42 (T helper$ or T supressor$ or T helper$ T supressor$ ratio).tw. (24786)

43 Total complement level$.tw. (20)

44 Autoantibodies.tw. (42037)

45 exp phospholipid antibody/ (9920)

46 Anti‐endometrial.tw. (23)

47 Antiphospholipid$.tw. (13777)

48 exp HLA antigen/ (81011)

49 exp HLA A1 antigen/ (597)

50 exp HLA A2 antigen/ (3288)

51 (HLA or human leucocyte antigen$).tw. (104497)

52 Anti‐laminin‐1.tw. (43)

53 Anti‐thyroid.tw. (1873)

54 Anti‐Thomsen Friedenreich antigen$.tw. (5)

55 Anti‐transferrin.tw. (290)

56 Anti‐LDL.tw. (186)

57 (Anti‐2HSG or Heremans‐Schmidt glycoprotein).tw. (4)

58 interleukin$.tw. (199692)

59 (MCP‐I or monocyte chemoattractant protein‐I).tw. (112)

60 (MIF or migration inhibitory factor$).tw. (5063)

61 (TNF‐a or tumour necrosis factor$ alfa).tw. (5998)

62 Fas ligand$.tw. (6708)

63 Endometrial marker$.tw. (18)

64 CAMs.tw. (2100)

65 cell adhesion molecule$.tw. (24039)

66 exp integrin/ (29036)

67 Integrin$.tw. (48293)

68 Selectin$.tw. (67300)

69 Cadherin$.tw. (27150)

70 Aromatase P450.tw. (202)

71 estrogen receptor$.tw. (46656)

72 progesterone receptor$.tw. (19861)

73 MTMMP$.tw. (15)

74 cyr61.tw. (755)

75 exp cysteine rich protein 61/ (753)

76 cysteine‐rich heparin‐binding protein$.tw. (12)

77 (ANXA 1 or ANXA1).tw. (452)

78 (Annexin 1 or Annexin1).tw. (425)

79 (PGP 9?5 or PGP9?5 or protein gene product$).tw. (2620)

80 serum marker$.tw. (7720)

81 neural marker$.tw. (1119)

82 cell surface marker$.tw. (5851)

83 inflammatory marker$.tw. (17339)

84 microarray$.tw. (101846)

85 microRNA$.tw. (40082)

86 proteomic$.tw. (55191)

87 genomic$.tw. (217184)

88 (endometri$ adj2 biops$).tw. (4369)

89 Follistatin$.tw. (1945)

90 exp vasculotropin/ (69810)

91 Vascular Endothelial Growth Factor A.tw. (2275)

92 exp vitamin D binding protein/ (2064)

93 exp cytokine/ (1034772)

94 exp interleukin derivative/ (2790)

95 exp interleukin 1/ (48499)

96 exp interleukin 6/ (136328)

97 exp interleukin 8/ (48884)

98 exp interleukin 12/ (31842)

99 exp interleukin 13/ (13584)

100 exp epidermal growth factor/ (32130)

101 exp fibroblast growth factor/ (13858)

102 cytokeratin 19/ (3601)

103 platelet derived growth factor/ (18930)

104 cytokeratin‐19.tw. (1918)

105 (VDBP or vitamin D‐binding protein$).tw. (1413)

106 urinary peptide$.tw. (174)

107 VDBP‐Cr.tw. (1)

108 urinary VDBP corrected for creatinine expression.tw. (1)

109 urinary marker$.tw. (830)

110 exp blood analysis/ (118854)

111 exp endometrium biopsy/ (4988)

112 exp urinalysis/ or exp biological marker/ (210153)

113 (biomarker or biomarkers).tw. (159748)

114 or/1‐113 (2734501)

115 endometriosis/di (Diagnosis) (4979)

116 114 or 115 (2738583)

117 exp endometriosis/ (25923)

118 Endometriosis.tw. (22110)

119 117 or 118 (27911)

120 116 and 119 (10326)

121 Animal/ not Human/ (1204497)

122 120 not 121 (10279)

Additional search February 2015 ‐ May 2015

Embase <1980 to 2015 Week 35 (3.09.2015)>

Search Strategy:

‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐

1 Laboratory Test$.tw. (44290)

2 growth factor$.tw. (335543)

3 scatter factor$.tw. (1407)

4 cytokine$.tw. (343623)

5 hepatocyte growth factor.tw. (10104)

6 (FGF or fibroblast growth factor$).tw. (39159)

7 (PDGF or platelet derived growth factor$).tw. (24591)

8 (EGF or epidermal growth factor$).tw. (73599)

9 (IGF‐I or insulin‐like growth factor$ or IGF1).tw. (51838)

10 (TGF‐a or transforming growth factor alfa or TGFa).tw. (583)

11 (TGF‐b or transforming growth factor beta or TGFb).tw. (32580)

12 (EGFR or epidermal growth factor receptor$).tw. (71526)

13 (VEGF or vascular endothelial growth factor$).tw. (79087)

14 exp luteinizing hormone/ec [Endogenous Compound] (22767)

15 leptin$.tw. (34921)

16 exp progesterone blood level/ or exp progesterone urine level/ (6534)

17 Proteolytic enzyme$.tw. (9903)

18 exp matrix metalloproteinase/ (20462)

19 matrix metalloproteinase$.tw. (44380)

20 MMP$.tw. (63208)

21 TIMP$.tw. (15146)

22 exp "tissue inhibitor of metalloproteinase 2"/ (5136)

23 exp "tissue inhibitor of metalloproteinase 1"/ (9381)

24 exp glycoprotein/ec [Endogenous Compound] (260024)

25 (Ca‐125 or Ca125 or cancer antigen 125).tw. (10051)

26 (Ca‐19‐9 or Ca19‐9 or cancer antigen 19‐9).tw. (6446)

27 (PP 14 or PP14).tw. (243)

28 serum placental protein$.tw. (44)

29 exp follistatin/ (2283)

30 Osteopontin$.tw. (9173)

31 exp intercellular adhesion molecule 1/ (33492)

32 exp selectin/ (3217)

33 soluble intercellular adhesion.tw. (1865)

34 Soluble adhesion molecule$.tw. (944)

35 sICAM.tw. (3049)

36 sVCAM$.tw. (1924)

37 (sEcadherin or soluble E‐cadherin).tw. (125)

38 (sEselectin or soluble E‐selectin).tw. (861)

39 exp T lymphocyte/ (394405)

40 exp natural killer T cell/ (6310)

41 Immune cells alteration$.tw. (6)

42 (T helper$ or T supressor$ or T helper$ T supressor$ ratio).tw. (26082)

43 Total complement level$.tw. (20)

44 Autoantibodies.tw. (44153)

45 exp phospholipid antibody/ (10362)

46 Anti‐endometrial.tw. (25)

47 Antiphospholipid$.tw. (14399)

48 exp HLA antigen/ (83748)

49 exp HLA A1 antigen/ (622)

50 exp HLA A2 antigen/ (3409)

51 (HLA or human leucocyte antigen$).tw. (109332)

52 Anti‐laminin‐1.tw. (43)

53 Anti‐thyroid.tw. (2059)

54 Anti‐Thomsen Friedenreich antigen$.tw. (7)

55 Anti‐transferrin.tw. (297)

56 Anti‐LDL.tw. (191)

57 (Anti‐2HSG or Heremans‐Schmidt glycoprotein).tw. (4)

58 interleukin$.tw. (210083)

59 (MCP‐I or monocyte chemoattractant protein‐I).tw. (114)

60 (MIF or migration inhibitory factor$).tw. (5342)

61 (TNF‐a or tumour necrosis factor$ alfa).tw. (6488)

62 Fas ligand$.tw. (6895)

63 Endometrial marker$.tw. (18)

64 CAMs.tw. (2198)

65 cell adhesion molecule$.tw. (25207)

66 exp integrin/ (30330)

67 Integrin$.tw. (50938)

68 Selectin$.tw. (71624)

69 Cadherin$.tw. (29496)

70 Aromatase P450.tw. (207)

71 estrogen receptor$.tw. (49530)

72 progesterone receptor$.tw. (21068)

73 MTMMP$.tw. (16)

74 cyr61.tw. (822)

75 exp cysteine rich protein 61/ (829)

76 cysteine‐rich heparin‐binding protein$.tw. (12)

77 (ANXA 1 or ANXA1).tw. (500)

78 (Annexin 1 or Annexin1).tw. (440)

79 (PGP 9?5 or PGP9?5 or protein gene product$).tw. (2760)

80 serum marker$.tw. (8158)

81 neural marker$.tw. (1234)

82 cell surface marker$.tw. (6222)

83 inflammatory marker$.tw. (19492)

84 microarray$.tw. (110181)

85 microRNA$.tw. (47554)

86 proteomic$.tw. (60599)

87 genomic$.tw. (233444)

88 (endometri$ adj2 biops$).tw. (4589)

89 Follistatin$.tw. (2081)

90 exp vasculotropin/ (74115)

91 Vascular Endothelial Growth Factor A.tw. (2526)

92 exp vitamin D binding protein/ (2196)

93 exp cytokine/ (1094317)

94 exp interleukin derivative/ (3281)

95 exp interleukin 1/ (50850)

96 exp interleukin 6/ (147379)

97 exp interleukin 8/ (52281)

98 exp interleukin 12/ (33479)

99 exp interleukin 13/ (14685)

100 exp epidermal growth factor/ (33057)

101 exp fibroblast growth factor/ (14499)

102 cytokeratin 19/ (3886)

103 platelet derived growth factor/ (19655)

104 cytokeratin‐19.tw. (2030)

105 (VDBP or vitamin D‐binding protein$).tw. (1520)

106 urinary peptide$.tw. (189)

107 VDBP‐Cr.tw. (1)

108 urinary VDBP corrected for creatinine expression.tw. (1)

109 urinary marker$.tw. (883)

110 exp blood analysis/ (124468)

111 exp endometrium biopsy/ (5197)

112 exp urinalysis/ or exp biological marker/ (232619)

113 (biomarker or biomarkers).tw. (182609)

114 or/1‐113 (2911073)

115 endometriosis/di [Diagnosis] (5173)

116 114 or 115 (2915302)

117 exp endometriosis/ (27433)

118 Endometriosis.tw. (23449)

119 117 or 118 (29532)

120 116 and 119 (10922)

121 Animal/ not Human/ (1261620)

122 120 not 121 (10862)

123 (201501$ or 201502$ or 201503$ or 201504$).em. (49200)

124 122 and 123 (34)

Appendix 5. Search strategy for CINAHL (EBSCO platform)

 Database: CINAHL Plus with Full Text (EBSCOhost) <1980 to 20.04.2015>

Search strategy:

#	Query	Results
S97	S3 AND S96	1131
S96	S4 OR S5 OR S6 OR S7 OR S8 OR S9 OR S10 OR S11 OR S12 OR S13 OR S14 OR S15 OR S16 OR S17 OR S18 OR S19 OR S20 OR S21 OR S22 OR S23 OR S24 OR S25 OR S26 OR S27 OR S28 OR S29 OR S30 OR S31 OR S32 OR S33 OR S34 OR S35 OR S36 OR S37 OR S38 OR S39 OR S40 OR S41 OR S42 OR S43 OR S44 OR S45 OR S46 OR S47 OR S48 OR S49 OR S50 OR S51 OR S52 OR S53 OR S54 OR S55 OR S56 OR S57 OR S58 OR S59 OR S60 OR S61 OR S62 OR S63 OR S64 OR S65 OR S66 OR S67 OR S68 OR S69 OR S70 OR S71 OR S72 OR S73 OR S74 OR S75 OR S76 OR S77 OR S78 OR S79 OR S80 OR S81 OR S82 OR S83 OR S84 OR S85 OR S86 OR S87 OR S88 OR S89 OR S90 OR S91 OR S92 OR S93 OR S94 OR S95	341775
S95	TX urinary peptide*	1598
S94	TX (VDBP or vitamin D‐binding protein*)	134
S93	TX cytokeratin‐19	109
S92	TX (Luteinizing Hormone* or LH)	18041
S91	(MH "Diagnosis, Laboratory+")	101773
S90	"Keratin‐19"	2
S89	(MH "Platelet‐Derived Growth Factor")	394
S88	(MH "Epidermal Growth Factors")	1264
S87	(MH "Interleukins")	6584
S86	(MH "Cytokines")	6860
S85	TX Vitamin D‐Binding Protein	131
S84	(MH "Vascular Endothelial Growth Factor A")	194
S83	TX (endometri* N2 biops*)	432
S82	TX (endometri* adj2 biops*)	0
S81	TX genomic$	7487
S80	TX proteomic*	2434
S79	TX microRNA	824
S78	TX microarray	3123
S77	TX (PGP 95 or PGP95 or protein gene product*)	9925
S76	TX (Annexin 1 or Annexin1)	472
S75	TX (ANXA 1 or ANXA1)	41
S74	TX cysteine‐rich heparin‐binding protein*	12
S73	(MH "Protein Array Analysis")	73
S72	TX cyr61	34
S71	TX MTMMP*	0
S70	TX progesterone receptor*	1927
S69	TX estrogen receptor*	5193
S68	TX Aromatase P450	38
S67	TX Cadherin*	900
S66	TX Selectin*	28411
S65	TX Integrin*	1587
S64	TX cell adhesion molecule*	1578
S63	TX CAMs	550
S62	TX Endometrial marker*	54
S61	TX Fas ligand	338
S60	TX (TNF‐a or tumour necrosis factor* alfa)	1489
S59	TX (MIF or migration inhibitory factor*)	399
S58	TX (MCP‐I or monocyte chemoattractant protein‐I)	13
S57	TX interleukin	13809
S56	TX (Anti‐2HSG or Heremans‐Schmidt glycoprotein)	7
S55	TX Anti‐LDL	9
S54	TX Anti‐transferrin	3
S53	TX Anti‐Thomsen Friedenreich antigen*	1
S52	TX Anti‐thyroid	109
S51	TX Anti‐laminin‐1	15
S50	TX (HLA or human leucocyte antigen*)	4202
S49	(MM "HLA Antigens")	638
S48	TX Antiphospholipid*	1249
S47	TX Anti‐endometrial	34
S46	(MH "Antibodies/BL/DU")	1294
S45	TX Autoantibodies	4385
S43	TX Total complement level	3
S42	TX (T helper* or T supressor*)	2341
S41	TX Immune cells alteration*	24
S40	TX natural killer t‐cells	669
S39	(MM "T Lymphocytes")	2404
S38	TX (sEselectin or soluble E‐selectin)	91
S37	TX (sEcadherin or soluble E‐cadherin)	8
S36	TX sVCAM	100
S35	TX sICAM	173
S34	TX Soluble adhesion molecule	368
S33	TX soluble intercellular adhesion	237
S32	(MM "Cell Adhesion Molecules")	52
S31	TX Osteopontin*	416
S30	TX Follistatin	74
S29	TX serum placental protein*	11
S28	TX (Ca‐19‐9 or Ca19‐9 or cancer antigen 19‐9)	262
S27	TX (Ca‐125 or Ca125 or cancer antigen 125)	831
S26	(MM "Glycoproteins/BL/DU")	224
S25	TX tissue inhibitor of metalloproteinase	423
S24	TX TIMP*	1845
S23	TX MMP*	4244
S22	TX matrix metalloproteinase*	3325
S21	TX Proteolytic enzyme*	1461
S20	(MM "Progesterone/BL/DU")	51
S19	TX leptin*	3258
S18	(MM "Luteinizing Hormone/BL/DU")	38
S17	TX (VEGF or vascular endothelial growth factor*)	7166
S16	TX (EGFR or epidermal growth factor receptor*)	6188
S15	TX (TGF‐b or transforming growth factor beta or TGFb)	2972
S14	TX (TGF‐a or transforming growth factor alfa or TGFa)	464
S13	TX (IGF‐I or insulin‐like growth factor* or IGF1)	3588
S12	TX (EGF or epidermal growth factor*)	6250
S11	TX (PDGF or platelet derived growth factor*)	3195
S10	TX (FGF or fibroblast growth factor*)	3395
S9	TX hepatocyte growth factor*	880
S8	TX cytokine*	20821
S7	TX scatter factor*	1864
S6	TX growth factor*	76163
S5	TX Laboratory Test*	82732
S4	TX (biomarker* or marker*)	84857
S3	S1 OR S2	2841
S2	TX Endometrio*	2841
S1	(MM "Endometriosis")	889
S4	TX (biomarker* or marker*)	61,794
S3	S1 OR S2	2,174
S2	TX Endometrio*	2,174
S1	(MM "Endometriosis")	1,306

Appendix 6. The search strategy for other databases

Searches for clinical studies

Database: PsycINFO (Ovid) <1806 to April Week 2 2015 (20.04.2015)>

Search strategy:

1. endometriosis.tw. (174)

Database: Web of Science Core Collection (Thomson Reuters) <1900 to Present (20.04.2015)>

Search strategy:

1. Topic=(endometrio*) AND Topic=(diagnos* OR test*); Timespan=All Years (2518)

Database: LILACS <20.04.2015>

Search strategy:

1. (tw:(endometriosis)) AND (tw:(diagnos*)) (420)

Database: OAIster (WorldCat.org) <20.04.2015>

Search strategy:

1. endometriosis and (marker* or biomarker*) (11)

2. endometriosis and diagnos* (446)

Database: TRIP <20.04.2015>

Search strategy:

1. (endometriosis and diagnos*) (1648)

Searches of trial registers for ongoing and registered trials

Database: ClinicalTrials.gov (US NIH) <20.04.2015>

Search strategy:

1. endometriosis (220)

2. endometriosis AND diagnosis (22)

Database: WHO International Clinical Trials Registry Platform (ICTRP) <20.04.2015>

Search strategy:

1. endometriosis (523)

Searches for the reviews as source of references to potentially relevant studies

Database: MEDION <10.01.2014> (the last available date; database is not updated further)

Search strategy:

ICP Code female genital system (including breast), Signssymp medical imaging, laboratory tests, histology and cytology, endoscopy and laparoscopy. Filter: systematic reviews of diagnostic studies. (2)

Database: DARE (CRD) <20.04.2015>

Search strategy:

1. endometriosis (99)

PubMed, a 'Systematic Review' search under the 'Clinical Queries' link <20.04.2015>

Search strategy:

1. (endometriosis) AND systematic(sb) (418)

Category: Diagnosis; Scope: Broad

Searches for papers recently published and not yet indexed in the major databases

Search engine: PubMed <20.10.2014 to 20.04.2015>

Search strategy:

1. marker (14979) 2. test (61151) 3. diagnos* (69743) 4. biomarker (10806) 5. or/1‐4 (7943) Filters: Publication date from 2014/10/20 to 2015/04/20	Index test(s) set
6. Endometriosis (584) Filters: Publication date from 2014/10/20 to 2015/04/20	Target condition set
7. 5 and 6 (267) Filters: Publication date from 2014/10/20 to 2015/04/20	Combined sets

Appendix 7. Summary of findings table 2 ‐ Endometrial biomarkers that do not distinguish between women with and without endometriosis

Review question	Which endometrial biomarkers are unlikely to serve as a basis for a diagnostic test for endometriosis?
Importance	Biomarkers that did not show differential expression in women with and without endometriosis are unlikely to be helpful in a diagnostic sense and hence are not worth pursuing. These findings are based on adequately designed studies that met inclusion criteria for this review.
Patients	Reproductive‐aged women with suspected endometriosis or persistent ovarian mass, or women undergoing infertility work‐up
Settings	Hospitals (public or private of any level), outpatient clinics (general gynaecology, reproductive medicine, pelvic pain) or radiology departments
Reference standard	Visualisation of endometriosis at surgery (laparoscopy or laparotomy) with or without histological confirmation
Study design	Cross‐sectional, single gate design (N = 11) or two‐gate design (N = 18); unable to determine if single‐ or two‐gate design for 2 studies. Prospective enrolment; a study could assess more than 1 test or more than 1 type of endometriosis
Risk of bias	Overall judgement	Poor quality (no studies had 'low risk' assessment in all 4 domains)
	Patient selection bias	High risk: 29 studies; unclear risk: 2 studies; low risk: 0 studies
	Index test interpretation bias	High risk: 28 studies; unclear risk: 3 studies; low risk: 0 studies
	Reference standard interpretation bias	High risk: 0 studies; unclear risk: 13 studies; low risk: 18 studies
	Flow and timing selection bias	High risk: 12 studies; unclear risk: 2 studies; low risk: 17 studies
Applicability concerns	Concerns regarding patient selection	High concern ‐ 25 studies; unclear concern ‐ 3 studies; low concern ‐ 3 studies
	Concerns regarding index test	High concern ‐ 0 studies; unclear concern ‐ 0 studies; low concern ‐ 31 studies
	Concerns regarding reference standard	High concern ‐ 0 studies; unclear concern ‐ 0 studies; low concern ‐ 31 studies
Biomarker	Expression levels	rASRM stage	Menstrual cycle phase	Sample collection method	Reference
1. Angiogenesis and growth factors and their receptors
EGF mRNA (epidermal growth factor)	Endometriosis (N = 35)a: 0.13 ± 0.01 Controls (N = 31): 0.24 ± 0.02; P = 0.120 # sub‐analysis per cycle phase ‐ similar findings	III‐IV	Proliferative/secretory	'Endometrial curettage' ‐ likely sharp	Lee 2007
FGF‐2 mRNA (fibroblast growth factor‐2)	Endometriosis (N = 35)a: 17.38 ± 1.73 Controls (N = 31): 20.0 ± 2.53; P = 0.572 # sub‐analysis per cycle phase ‐ similar findings	III‐IV	Proliferative/secretory	'Endometrial curettage' ‐ likely sharp	Lee 2007
Glycodelin mRNA (PP14 or PAEP) (placental protein 14 or progestogen‐associated endometrial protein)	Endometriosis (N = 17)a: 2.5 (0.6‐4.5) Controls (N = 10): 0.936 (0.1‐1.9); P = NS	II‐IV	Proliferative	Sharp curettage	Meola 2009
PDGF‐A mRNA (platelet‐derived growth factor A)	Endometriosis (N = 35)a: 6.52 ± 0.58 Controls (N = 31): 13.02 ± 1.60; P = 0.572 # sub‐analysis per cycle phase ‐ similar findings	III‐IV	Proliferative/secretory	'Endometrial curettage' ‐ likely sharp	Lee 2007
PIGF (placental growth factor)	Endometriosis (N = 18)a: 120 ± 31 Controls (N = 23): 87 ± 16; P = NS Endometriosis (N = 22)a: 87 ± 30 Controls (N = 27): 109 ± 15; P = NS	III‐IV	Proliferative Secretory	Aspiration curettage	Gilabert‐Estelles 2007
PIGF mRNA (placental growth factor)	Endometriosis (N = 18)a: 0.050 ± 0.017 Controls (N = 23): 0.032 ± 0.009 Endometriosis (N = 22)a: 0.037 ± 0.016 Controls (N = 27): 0.024 ± 0.004; P = NS	III‐IV	Proliferative Secretory	Aspiration curettage	Gilabert‐Estelles 2007
PKR1 mRNA (prokineticin receptor 1), EG‐VEGF receptor	Endometriosis (N = 9)b: 1.89 (0.1–5.5) Controls (N = 14): 3.92 (0.0–23.4); P = NS Endometriosis (N = 6)b: 0.43 (0–2.1) Controls (N = 19): 0.64 (0.0–6.7); P = NS	—	Proliferative Secretory	Aspiration curettage	Lee 2010
PKR2 mRNA (prokineticin receptor 2), EG‐VEGF receptor	Endometriosis (N = 9)b: 0.93 (0.0–10.7) Controls (N = 14): 0.75 (0.0–6.6); P = NS Endometriosis (N = 6)b: 0.24 (0.0–1.2) Controls (N = 19): 0 (0.0–1.0); P = NS	—	Proliferative Secretory	Aspiration curettage	Lee 2010
TSP‐1 (thrombospondin‐1)	Endometriosis (N = 18)a: 313 ± 46 ng/mg Controls (N = 23): 210 ± 27 ng/mg; P = NS Endometriosis (N = 22)a: 413 ± 63 ng/mg Controls (N = 27): 317 ± 42 ng/mg; P = NS	III‐IV	Proliferative Secretory	Aspiration curettage	Gilabert‐Estelles 2007
TSP‐1 (thrombospondin‐1)	Endometriosis (N = 58)a: 140 ± 28 ng/mg Controls (N = 38): 81 ± 16 ng/mg; P = NS Endometriosis (N = 58)a: 69 ± 16 ng/mg Controls (N = 38): 69 ± 13 ng/mg; P = NS	—	Proliferative Secretory	Aspiration curettage	Ramon 2011
TSP‐1 mRNA (thrombospondin‐1)	Endometriosis (N = 18)a: 2.621 ± 0.735 Controls (N = 23): 1.510 ± 0.270; P = NS Endometriosis (N = 22)a: 3.093 ± 0.7002 Controls (N = 27): 2.370 ± 0.910; P = NS	III‐IV	Proliferative Secretory	Aspiration curettage	Gilabert‐Estelles 2007
TSP‐1 mRNA (thrombospondin‐1)	Endometriosis (N = 58)a: 1.79 ± 0.36 Controls (N = 38): 2.36 ± 0.49; P = NS Endometriosis (N = 58)a: 2.65 ± 0.87 Controls (N = 38): 1.52 ± 0.39; P = NS	—	Proliferative Secretory	Aspiration curettage	Ramon 2011
TYMP (thymidine phosphorylase)	Endometriosis (N = 15)b: 11.8 (2.7–38.6) ng/ml Controls (N = 14): 16.8 (1.1–37.6) ng/ml; P = 0.68	III‐IV	Proliferative	Aspiration curettage	Laudanski 2014
VEGF (vascular endothelial growth factor)	Endometriosis (N = 10)b: 0.142 (0.087‐0.943) Controls (N = 7): 0.129 (0.097‐0.837); P = 0.943	II‐IV	Menstrual	Menstrual fluid	Da Silva 2014
VEGF mRNA (vascular endothelial growth factor)	Endometriosis (N = 24)b: 5.83 (2.5‐8.75) Controls (N = 23): 5.01 (1.7‐10.83); P = NS # sub‐analysis per cycle phase ‐ similar findings	III‐IV	Proliferative/secretory	'Endometrial biopsy curette' ‐ not specified	Cho 2012
EG‐VEGF mRNA (endocrine gland‐derived VEGF)	Endometriosis (N = 9)b: 0.04 (2.9 × 10‐4– 0.6) Controls (N = 14): 0.03 (5.2 × 10‐6–1.0); P = NS Endometriosis (N = 6)b: 0.28 (2.0 × 10‐4–5.6) Controls (N = 19): 1.50 (0.0–27.9); P = NS	—	Proliferative Secretory	Aspiration curettage	Lee 2010
VEGF/sFlt (VEGF and soluble fms‐like tyrosine kinase ratio)	Endometriosis (N = 24)b: 1.06 (0.71‐2.59) Controls (N = 23): 0.71 (0.35‐1.06); P = 0.064 # sub‐analysis per cycle phase ‐ similar findings	III‐IV	Proliferative/secretory	'Endometrial biopsy curette' ‐ not specified	Cho 2012
2. Apoptosis markers and regulators
Bax mRNA (BCL2‐associated X protein)	Endometriosis (15)b: 1.22 Controls (N = 30): 1.15; P = NS	I‐IV	Proliferative	Sharp curettage	Zubor 2009
Bcl‐xL (B‐cell lymphoma‐extra large, or BCL2‐like‐1 isoform 1)	Endometriosis (15)b: 1.08 Controls (N = 30): 1.07; P = NS	I‐IV	Proliferative	Sharp curettage	Zubor 2009
Bcl‐xL:Bcl‐xS ratio (ratio B‐cell lymphoma‐extra large/B‐cell lymphoma‐extra small)E	Endometriosis (15)b: 2.63 Controls (N = 30): 5.63; P = 0.2965	I‐IV	Proliferative	Sharp curettage	Zubor 2009
3. Cell adhesion molecules and other matrix‐related proteins
α2β1 integrin	Positively stained menstrual effluent: endometriosis (N = 8)c: 7/8 controls (N = 8): 6/8; P = NS Positively stained endometrium: endometriosis (N = 8)c: 8/8 controls (N = 8): 8/8; P = NS	I	Early proliferative (day 2‐5)	Aspiration curettage	Van der Linden 1994
α3β1 integrin	Positively stained menstrual effluent: endometriosis (N = 8)c: 7/8 controls (N = 8): 6/8; P = NS Positively stained endometrium: endometriosis (N = 8)c: 7/8 controls (N = 8): 7/8; P = NS	I	Early proliferative (day 2‐5)	Aspiration curettage	Van der Linden 1994
α4β1 integrin	Positively stained menstrual effluent: endometriosis (N = 8)c: 2/8 controls (N = 8): 3/8; P = NS Positively stained endometrium: endometriosis (N = 8)c: 7/8 controls (N = 8): 5/8; P = NS	I	Early proliferative (day 2‐5)	Aspiration curettage	Van der Linden 1994
α5β1 integrin	Positively stained menstrual effluent: endometriosis (N = 8)c: 7/8 controls (N = 8): 4/8; P = NS Positively stained endometrium: endometriosis (N = 8)c: 8/8 controls (N = 8): 8/8; P = NS	I	Early proliferative (day 2‐5)	Aspiration curettage	Van der Linden 1994
α6β1 integrin	Positively stained menstrual effluent: endometriosis (N = 8)c: 2/8 controls (N = 8): 3/8; P = NS Positively stained endometrium: endometriosis (N = 8)c: 8/8 controls (N = 8): 8/8; P = NS	I	Early proliferative (day 2‐5)	Aspiration curettage	Van der Linden 1994
αVβ3 integrin	Positive samples: endometriosis (N = 20)c: 10 (50) infertile controls (N = 20): 8 (40) fertile controls (N = 20): 7 (35); P = NS endometriosis (N = 16)c: 16 (100) infertile controls (N = 16): 16 (100) fertile controls (N = 19): 18 (94.7); P = NS	l‐II	Mid‐secretory Late‐secretory	—	Casals 2012
αVβ5 integrins	Glandular cells H Score: endometriosis (N = 40)b: 0.47 (0.0‐1.0) controls (N = 12): 0.6 (0.3‐0.8); P = 0.460 endometriosis (N = 40)b: 1.2 (0.5‐1.9) controls (N = 12): 1.3 (0.9‐1.7); P = 0.507	I‐IV	Proliferative Secretory	—	Puy 2002
Stromal cells H Score: endometriosis (N = 40)b: 2.0 (1.0‐2.7) controls (N = 12): 1.9 (1.4‐2.5); P = 0.920 endometriosis (N = 40)b: 1.7 (0.8‐2.7) controls (N = 12): 1.5 (1.1‐2.4); P = 0.752	I‐IV	Proliferative Secretory	—	Puy 2002
αVβ6 integrins	Glandular cells H Score : endometriosis (N = 40)b: 0.5 (0.1‐1.3) controls (N = 12): 0.75 (0.3‐1.2); P = 0.856 endometriosis (N = 40)b: 0.7 (0.3‐0.9) controls (N = 12): 1.0 (0.8‐1.2); P = 0.975	I‐IV	Proliferative Secretory	—	Puy 2002
Stromal cells H Score: endometriosis (N = 40)b: 1.4 (0.6‐2.2) controls (N = 12): 1.2 (0.3‐2.5); P = 0.651 endometriosis (N = 40)b: 1.8 (0.8‐2.6) controls (N = 12): 1.7 (1.0‐2.2); P = 0.635	I‐IV	Proliferative Secretory	—	Puy 2002
ICAM1 (intercellular adhesion molecule1)	Endometriosis (N = 8)a: 0.56 ± 0.25 Control (N = 8): 0.69 ± 0.19; P = NS # sub‐analysis per cycle phase ‐ similar findings	I‐IV	Proliferative/secretory	Aspiration curettage	Pino 2009
ICAM1 (CD54) (intercellular adhesion molecule1)	(Small/medium/large cells): endometriosis (N = 10)b: 96 (25–193)/92 (13–106)/106 (18–143) controls (12): 51 (4–453)/48 (5–129)/49 (4–103); P = NS # significant difference between the groups in secretory phase ‐ no data for 2 × 2 table	—	Proliferative	Aspiration curettage	Prefumo 2002
sICAM1 (soluble ICAM1)	Endometriosis (N = 8)a: 0.31 ± 0.16 Control (N = 8): 0.13 ± 0.13; P = NS # sub‐analysis per cycle phase ‐ similar findings	I‐IV	Proliferative/secretory	Aspiration curettage	Pino 2009
ICAM1 mRNA (intercellular adhesion molecule1)	Endometriosis (N = 6)a: 0.44 ± 0.11 Control (N = 6): 0.76 ± 0.11; P = NS	I‐IV	Proliferative/secretory	Aspiration curettage	Pino 2009
E‐cadherin	Positively stained menstrual effluent: endometriosis (N = 8)c: 3/8 controls (N = 8): 3/8; P = NS Positively stained endometrium: endometriosis (N = 8)c: 8/8 controls (N = 8): 8/8; P = NS	I	Early proliferative (day 2‐5)	Aspiration curettage	Van der Linden 1994
LAMA5 (laminin subunit alpha‐5)	Endometriosis (N = 15)b: 1500 (1060–1840) pg/ml controls (N = 14):1400 (1340–1780) pg/ml; P = 0.95	III‐IV	Proliferative	Aspiration curettage	Laudanski 2014
LFA‐3 (CD58) (leukocyte function associated molecule‐3)	(Small/medium/large cells): endometriosis (N = 10)b: 101 (64 – 302)/108 (28–155)/65 (35–197) controls (12): 315 (65 – 824)/150 (33–264)/124 (67–164); P = NS endometriosis (N = 10)b: 70 (34–252)/49 (26–184)/48 (37–118) controls (12): 171 (145–643)/134 (94–223)/162 (56–228); P = NS	—	Proliferative Secretory	Aspiration curettage	Prefumo 2002
MMP‐1 (matrix metalloproteinase‐1)	Endometriosis (N = 8)a: just above detection limit Control (N = 8): just above detection limit; P = NS	I‐IV	Proliferative/secretory	Aspiration curettage	Pino 2009
MMP‐9 (matrix metalloproteinase‐9)	Endometriosis (N = 8)a: 2.75 ± 1.5 Control (N = 8): 0.75 ± 0.13; P = NS # sub‐analysis per cycle phase ‐ similar findings	I‐IV	Proliferative/secretory	Aspiration curettage	Pino 2009
MMP‐9 mRNA (matrix metalloproteinase‐9)	Endometriosis (N = 6)a: 0.21 ± 0.06 Control (N = 6): 0.33 ± 0.07; P = NS	I‐IV	Proliferative/secretory	Aspiration curettage	Pino 2009
OPN (osteopontin)	Positive samples: endometriosis (N = 20)c: 14 (70) infertile controls (N = 20): 13 (65) fertile controls (N = 20): 14 (70); P = NS endometriosis (N = 16)c: 16 (100) infertile controls (N = 16): 16 (100) fertile controls (N = 19): 19 (100); P = NS	I‐II	Mid‐secretory Late‐secretory	—	Casals 2012
OPN (+)/αVβ3 (+)	Positive samples: endometriosis (N = 20)c: 8 (40) infertile controls (N = 20): 7 (35) fertile controls (N = 20): 6 (30); P = NS endometriosis (N = 16)c: 16 (100) infertile controls (N = 16): 16 (100) fertile controls (N = 19): 18 (94.7); P = NS	I‐II	Mid‐secretory Late‐secretory	—	Casals 2012
PAI‐1 (plasminogen activator inhibitor ‐ 1)	Antigenic levels: endometriosis (N = 21)a: 0.46 ± 0.18 ng/mg controls (N = 35): 0.73 ± 0.19 ng/mg; P = NS Functional levels: endometriosis (N = 21)a: 1.64 ± 0.59 U/mg controls (N = 35): 1.12 ± 0.35 U/mg; P = NS	III‐IV	Menstrual/proliferative/secretory	—	Gilabert‐Estelles 2003
PAI‐1 (plasminogen activator inhibitor ‐ 1)	Endometriosis (N = 18)a: 1.73 ± 0.56 ng/mg Controls (N = 23): 1.48 ± 0.40 ng/mg Endometriosis (N = 22)a: 2.51 ± 0.65 ng/mg Controls (N = 27): 1.44 ± 0.44 ng/mg; P = NS	III‐IV	Proliferative Secretory	Aspiration curettage	Gilabert‐Estelles 2007
PAI‐1 mRNA (gene encoding for (plasminogen activator inhibitor ‐ 1)	Endometriosis (N = 18)a: 0.197 ± 0.068 Controls (N = 23): 0.117 ± 0.030 Endometriosis (N = 22)a: 0.496 ± 0.158 Controls (N = 27): 0.131 ± 0.041; P = NS	III‐IV	Proliferative Secretory	Aspiration curettage	Gilabert‐Estelles 2007
PAI‐2 (plasminogen activator inhibitor ‐ 2)	Antigenic levels: endometriosis (N = 21)a: 10.69 ± 3.33 ng/mg controls (N = 35): 4.53 ± 1.75 ng/mg; P = NS	III‐IV	Menstrual/proliferative/secretory	—	Gilabert‐Estelles 2003
PAI‐3 (plasminogen activator inhibitor ‐ 3)	Antigenic levels: endometriosis (N = 21)a: 233 ± 50 ng/mg controls (N = 35): 340 ± 45 ng/mg; P = NS	III‐IV	Menstrual/proliferative/secretory	—	Gilabert‐Estelles 2003
tPA (tissue‐type plasminogen activator)	Antigenic levels: endometriosis (N = 21)a: 12.69 ± 1.52 ng/mg controls (N = 35): 9.14 ± 1.33 ng/mg; P = NS Functional levels: endometriosis (N = 21)a: 0.41 ± 0.09 U/mg controls (N = 35): 0.58 ± 0.10 U/mg; P = NS	III‐IV	Menstrual/proliferative/secretory	—	Gilabert‐Estelles 2003
uPA (urokinase‐type plasminogen activator)	Functional levels: endometriosis (N = 21)a: 0.11 ± 0.02 ng/mg controls (N = 35): 0.06 ± 0.01 ng/mg; P = NS	III‐IV	Menstrual/proliferative/secretory	—	Gilabert‐Estelles 2003
uPA‐R (urokinase‐type plasminogen activator receptor)	Antigenic levels: endometriosis (N = 21)a: 2.51 ± 0.55 ng/mg controls (N = 35): 2.90 ± 0.32 ng/mg; P = NS	III‐IV	Menstrual/proliferative/secretory	—	Gilabert‐Estelles 2003
tPA‐PAI‐3 (tissue plasminogen activator and plasminogen activator inhibitor ‐ 3 complex)	Antigenic levels: endometriosis (N = 21)a: 0.20 ± 0.10 ng/mg controls (N = 35): 0.06 ± 0.02 ng/mg; P = NS	III‐IV	Menstrual/proliferative/secretory	—	Gilabert‐Estelles 2003
uPA‐PAI‐3 (urokinase plasminogen activator and plasminogen activator inhibitor ‐ 3 complex)	Antigenic levels: endometriosis (N = 21)a: 0.27 ± 0.13 ng/mg controls (N = 35): 0.36 ± 0.18 ng/mg; P = NS	III‐IV	Menstrual/proliferative/secretory	—	Gilabert‐Estelles 2003
TIMP‐1 protein (tissue inhibitor of metalloproteinases‐1)	Endometriosis (N = 45)b: 0.253 ± 0.018 controls (N = 15): 0.267 ± 0.010; P = NS	II‐IV	Proliferative/secretory	—	Chen 2004
Endometriosis (N = 21)a: 31± 6 ng/ml controls (N = 35): 30 ± 5 ng/ml; P = NS	III‐IV	Menstrual/proliferative/secretory	—	Gilabert‐Estelles 2003
Endometriosis (N = 15)b: 56.7 (13.5–213.5) ng/ml Controls (N = 14): 58 (36.2–187.8) ng/ml; P = 0.45	III‐IV	Proliferative	Aspiration curettage	Laudanski 2014
TIMP‐1 mRNA (tissue inhibitor of metalloproteinases‐1)	Endometriosis (N = 18)a: 1.190 ± 0.228 Controls (N = 23): 1.589 ± 0.335 Endometriosis (N = 22)a: 2.028 ± 0.392 Controls (N = 27): 1.506 ± 0.363; P = NS	III‐IV	Proliferative Secretory	Aspiration curettage	Gilabert‐Estelles 2007
Endometriosis (N = 35)b: 2.31 ± 1.21 Controls (N = 20): 2.40 ± 0.89; P = NS	I‐IV	Proliferative/secretory	'Endometrial curettage' ‐ not specified	Li 2006
VCAM‐1 (CD106) (vascular cell adhesion molecule‐1)	(Small/medium/large cells): endometriosis (N = 10)b: 8 (0–42)/3 (2–22)/43 (0–57) controls (12): 34 (0–60)/14 (0–24)/21 (0–64); P = NS endometriosis (N = 10)b: 28 (5–64)/22 (6–60)/18 (10–126) controls (12): 27 (5–130)/18 (3–62)/33 (13–238); P = NS	—	Proliferative Secretory	Aspiration curettage	Prefumo 2002
4. Cell cycle regulatory molecules
Cyclin B1 mRNA	Endometriosis (N = 20)b: 0.79 ± 1.08 Controls: (N = 30): 0.51 ± 1.04; P = NS Endometriosis (N = 20)b: 0.98 ± 1.24 Controls (N = 30): 0.50 ± 0.49; P = NS # similar findings were observed for protein expression	II‐III	Proliferative Secretory	—	Tang 2009
Cdc2 mRNA (cyclin dependent kinase‐2)	Endometriosis (N = 20)b: 5.0 ± 5.0 Controls: (N = 30): 5.0 ± 5.0; P = NS Endometriosis (N = 20)b: 4.5 ± 5.5 Controls (N = 30): 4.5 ± 5.5; P = NS # similar findings were observed for protein expression	II‐III	Proliferative Secretory	—	Tang 2009
Plk1 mRNA (polo‐like kinase‐1)	Endometriosis (N = 20)b: 1.64 ± 1.18 Controls (N = 30): 1.35 ± 0.91; P = NS Endometriosis (N = 20)b: 0.84 ± 1.04 Controls (N = 30): 0.52 ± 0.44; P = NS # similar findings were observed for protein expression	II‐III	Proliferative Secretory	—	Tang 2009
5. Cell proliferation markers
Ki‐67 (Antigen KI‐67 or MKI67, marker of cellular proliferation)	Stromal cells: endometriosis (N = 13)b: 8.4 ± 1.4 controls (N = 6): 10.9 ± 4.8; P = NS endometriosis (N = 12)b: 9.6 ± 1.8 controls (N = 8): 10.5 ± 4.8; P = NS # data for glandular epithelium and blood vessels are also reported ‐ no difference between the groups	II‐III	Proliferative Secretory	Sharp curettage	Bourlev 2006
BW 495/36, endometrial epithelial marker	Positively stained menstrual effluent: endometriosis (N = 8)c: 8/8 controls (N = 8): 6/8; P = NS Positively stained endometrium: endometriosis (N = 8)c: 8/8 controls (N = 8): 7/8; P = NS	I	Early proliferative (day 2‐5)	Aspiration curettage	Van der Linden 1995
6. Cytoskeleton molecules
cytokeratin 18	Positively stained menstrual effluent: endometriosis (N = 8)c: 8/8 controls (N = 8): 7/8; P = NS Positively stained endometrium: endometriosis (N = 8)c: 8/8 controls (N = 8): 8/8; P = NS	I	Early proliferative (day 2‐5)	Aspiration curettage	Van der Linden 1995
CK19 or CYFRA 21‐1 (cytokeratin 19)	Positively stained menstrual effluent: endometriosis (N = 8)c: 8/8 controls (N = 8): 7/8; P = NS Positively stained endometrium: endometriosis (N = 8)c: 8/8 controls (N = 8): 8/8; P = NS	I	Early proliferative (day 2‐5)	Aspiration curettage	Van der Linden 1995
Vimentin	Positively stained menstrual effluent: endometriosis (N = 8)c: 8/8 controls (N = 8): 7/8; P = NS Positively stained endometrium: endometriosis (N = 8)c: 8/8 controls (N = 8): 8/8; P = NS	I	Early proliferative (day 2‐5)	Aspiration curettage	Van der Linden 1995
7. DNA‐repair and telomer maintenance molecules
Telomerase activity (relative telomerase activity (RTA))	Endometriosis (N = 30)b: 17.8 ± 30.8 Controls (N = 30): 7.6 ± 13.2; P = NS	I‐IV	Proliferative/secretory	—	Kim 2007
8. High throughput markers
mRNAome (mRNA microarray)	Endometriosis (N = 31)d: 0 Controls (N = 18): 0	I‐IV	Menstrual/secretory	Aspiration curettage	Fassbender 2012
9. Hormonal markers
EST (oestrogen sulphotransferase)	Glandular cells staining: endometriosis (N = 35)e: strong 2, medium 12, weak 8, absent 25 controls (N = 33): strong 9, medium 8, weak 6, absent 10; P = NS Stromal cells staining: endometriosis (N = 35)e: absent 35 controls (N = 33): absent 33; P = NS	I‐II	Proliferative/secretory	'Endometrial curettage' ‐ likely sharp	Hudelist 2007
LGR7 relaxin receptor mRNA (leucine‐rich G protein‐coupled receptor 7) evaluated to distinguish endometrioma from other non‐malignant ovarian masses	Endometriosis (N = 4)f: 0.95 Controls (N = 7): 1; P = NS Endometriosis (N = 9)f: 1.27 Controls (N = 12): 1; P = NS	III‐IV	Proliferative Secretory	—	Morelli 2010
Relaxin mRNA evaluated to distinguish endometrioma from other non‐malignant ovarian masses	Endometriosis (N = 13)c: 9 (69.2) Control (N = 19): 14 (73.7); P = NS # sub‐analysis per cycle phase ‐ similar findings	III‐IV	Proliferative/secretory	—	Morelli 2010
10. Immune system and inflammatory markers
A Cytokines
LIF (leukaemia‐inhibitory factor)	Endometriosis (N = 14)b: 25.53 (12.63–43.32) pg/ml Controls (N = 21): 36.26 (14.45–59.32); P = NS	I‐II	Secretory	Uterine flushing	Mikolajczyk 2006
LIF mRNA (leukaemia‐inhibitory factor gene)	Endometriosis (N = 14)b: 0.90 (0.73–1.09) Controls (N = 21): 0.92 (0.74–1.56); P = NS	I‐II	Secretory	Aspiration curettage	Mikolajczyk 2006
TNF‐a (tumour necrosis factor‐alpha)	Endometriosis (N = 10)b: 386.4 (48.31‐3636.06) Controls (N = 7): 56.02 (10.42‐619.205); P = 0.219	II‐IV	Menstrual	Menstrual fluid	Da Silva 2014
TNF‐a mRNA (tumour necrosis factor‐alpha)	Endometriosis (N = 25)b: 0.0005 (0.0003–0.0020) Controls (N = 25): 0.0002 (0.0001–0.0007); P = NS	I‐IV	Proliferative/secretory	'Endometrial curettage' ‐ not specified	Chen 2013
B Immune cells: peripheral blood mononuclear cells (PBMC)
CD56+ cells (endometrial granulated lymphocytes)	Early/mid/late secretory phase: endometriosis (N = 16)a: 1120 ± 160/1320 ± 160/2200 ± 240 controls (N = 17): 1200 ± 140/1360 ± 200/2440 ± 200; P = NS	I‐IV	Secretory	—	Klentzeris 1995
CD38+ cells (endometrial granulated lymphocytes)	Early/mid/late secretory phase: endometriosis (N = 16)a: 950 ± 200/1650 ± 300/3000 ± 300 controls (N = 17): 1550 ± 300/1750 ± 350/3000 ± 250; P = NS	I‐IV	Secretory	—	Klentzeris 1995
CD22+ cells (B lymphocytes)	early/mid/late secretory phase: endometriosis (N = 16)a: 87 ± 28/92 ± 31/120 ± 24 controls (N = 17): 90±31/104 ± 39/115 ± 28; P = NS	I‐IV	Secretory	—	Klentzeris 1995
CD68+ cells (macrophages)	Endometriosis (N = 31)b: 216.10 ± 104.41 Controls (N = 29): 175.93 ± 43.05; p = 0.06 # sub‐analysis per cycle phase ‐ significant difference between the groups only in proliferative phase ‐ no data for 2 × 2 table	I‐IV	Any phase	Aspiration curettage	Cetin 2013
CD68+ cells (macrophages)	Early/mid/late secretory phase: endometriosis (N = 16)a: 1178 ± 186/1271 ± 287/1643 ± 248 controls (N = 17): 1054 ± 155/1116 ± 155/1519 ± 287; P = NS	I‐IV	Secretory	—	Klentzeris 1995
CD16+ cells (natural killer cells)	Early/mid/late secretory phase: endometriosis (N = 16)a: 1276 ± 377/1305 ± 290/1421 ± 348 controls (N = 17): 957 ± 348/1044 ± 377/1189 ± 290; P = NS	I‐IV	Secretory	—	Klentzeris 1995
CD8+ cells (T suppressor/cytotoxic lymphocytes)	Early/mid/late secretory phase: endometriosis (N = 16)a: 691 ± 109/855 ± 127/891 ± 91 controls (N = 17): 782 ± 73/891 ± 91/964 ± 91; P = NS	I‐IV	Secretory	—	Klentzeris 1995
CD4+ cells (T helper/inducer lymphocytes)	Early/mid/late secretory phase: endometriosis (N = 16)a: 321 ± 57/357 ± 50/400 ± 64 controls (N = 17): 293 ± 57/307 ± 57/343 ± 57; P = NS	I‐IV	Secretory	—	Klentzeris 1995
C Interleukins
IL‐1β mRNA (interleukin ‐1β)	Endometriosis (N = 25)b: 0.0295 (0.0056–0.1039) Controls (N = 25): 0.008 (0.0040–0.0251); P = NS	I‐IV	Proliferative/secretory	'Endometrial curettage' ‐ not specified	Chen 2013
IL‐11 (interleukin ‐ 11)	Endometriosis (N = 14)b: not detected Controls (N = 21): not detected; P = NS	I‐II	Secretory	Uterine flushing	Mikolajczyk 2006
IL‐11 mRNA (interleukin ‐ 11)	Endometriosis (N = 14)b: 0.48 (0.39–0.59) Controls (N = 21): 0.52 (0.42–0.61); P = NS	I‐II	Secretory	Aspiration curettage	Mikolajczyk 2006
IL‐1R1 mRNA (interleukin ‐1 receptor type II)	Stroma positive samples: endometriosis (N = 17)c: 9 (53) controls (N = 17): 14 (82); P = 0.299 Glandular cells positive samples: endometriosis (N = 16)c: 13 (81) controls (N = 17): 14 (82); P = 0.411 # sub‐analysis per cycle phase ‐ similar findings	I‐II	Proliferative/secretory	—	Lawson 2008
D other immune/inflammatory markers
MPO (myeloperoxidase)	Endometriosis (N = 10)b: 1.229 (0.778‐2.094) Controls (N = 7): 1.69 (0.95‐1.84); P = 0.669	II‐IV	Menstrual	Menstrual fluid	Da Silva 2014
NAG (N‐acetyl‐β‐D‐Glucosaminidase)	Endometriosis (N = 10)b: 723.26 (536.7‐768.73) Controls (N = 7): 663.62 (357.5‐1214.1); P = 1.0	II‐IV	Menstrual	Menstrual fluid	Da Silva 2014
11. Mediators of prostaglandin biosynthesis
Akr1B1 mRNA (aldoketoreductase ‐1B1, PGF2a synthase)	Endometriosis (N = 22)b: 216.9 ± 25.9 Controls (N = 12): 217.6 ± 15.6; P = NS Endometriosis (N = 23)b: 121.2 ± 18.7 Controls (N = 17): 188.3 ± 31.9; P = NS	I‐IV	Proliferative Secretory	—	Rakhila 2013
Akr1C3 mRNA (aldoketoreductase ‐1C3, PGF2a synthase)	Endometriosis (N = 22)b: 132.8 ± 21.1 Controls (N = 12): 127.9 ± 26.5; P = NS Endometriosis (N = 23)b: 185.9 ± 31.1 Controls (N = 17): 196.0 ± 45.9; P = NS	I‐IV	Proliferative Secretory	—	Rakhila 2013
Cox‐1 mRNA (cyclo‐oxygenase‐1)	Endometriosis (N = 22)b: 263.6 ± 103.1 Controls (N = 12): 237.9 ± 78.3; P = NS Endometriosis (N = 23)b: 1045.0 ± 313.8 Controls (N = 17): 564.6 ± 151.2; P = NS	I‐IV	Proliferative Secretory	—	Rakhila 2013
15‐PGDH mRNA (15‐hydroxyprostaglandin dehydrogenase)	Endometriosis (N = 22)b: 41.3 ± 11.4 Controls (N = 12): 18.4 ± 3.2; P = NS Endometriosis (N = 23)b: 77.2 ± 51.4 Controls (N = 17): 48.1 ± 26.8; P = NS	I‐IV	Proliferative Secretory	—	Rakhila 2013
cPGES mRNA (cytosolic PGE2 synthase)	Endometriosis (N = 22)b: 63.6 ± 4.6 Controls (N = 12): 68.3 ± 4.8; P = NS Endometriosis (N = 23)b: 55.2 ± 4.3 Controls (N = 17): 61.4 ± 3.9; P = NS	I‐IV	Proliferative Secretory	—	Rakhila 2013
12. Nerve sheath and nerve growth markers
NF (neurofilament)	Endometriosis (N = 20)b: 0.02 ± 0.10 Controls (N = 20): 0.025 ± 1.04; P = NS	I‐II	Secretory	Aspiration curettage	Bokor 2009
NF (neurofilament)	Endometriosis (N = 31): no staining Controls (N = 29): no staining	I‐IV	Any phase	Aspiration curettage	Cetin 2013
PGP 9.5 (protein gene product 9.5)	Endometriosis (N = 31): no staining Controls (N = 29): no staining	I‐IV	Any phase	Aspiration curettage	Cetin 2013
13. Other peptides and proteins
hBD‐2 mRNA (human b‐defensin‐2)	Endometriosis (N = 25)b: 0.0343 (0.0025–2.0326) Controls (N = 25): 0.0034 (0.0025–0.0424); P = NS	I‐IV	Proliferative/secretory	'Endometrial curettage' ‐ not specified	Chen 2013
hBD‐2 (human b‐defensin‐2)	Negative samples: endometriosis (N = 25)c: 8 (32) controls (N = 25): 10 (40); P = NS	I‐IV	Proliferative/secretory	'Endometrial curettage' ‐ not specified	Chen 2013
14. Transcription factors and signalling molecules
AKT1 (RAC‐alpha serine/threonine‐protein kinase)	Endometriosis (N = 15)b: 7.4 (1.4–15.1) ng/ml Controls (N = 14): 5.6 (1.4–15.4) ng/ml; P = 0.9	III‐IV	Proliferative	Aspiration curettage	Laudanski 2014
JAG1 (jagged 1 protein)	Endometriosis (N = 15)b: 1.5 (0.3–6.5) ng/ml Controls (N = 14): 1.5 (0.4–2.8) ng/ml; P = 0.82	III‐IV	Proliferative	Aspiration curettage	Laudanski 2014
NS = Not stated

Appendix 8. Endometrial biomarkers that have limited diagnostic vaue in endometriosis

Testa

Angiogenesis and growth factors and their receptors

VEGF (vascular endothelial growth factor) protein or mRNA

Cell adhesion molecules and other matrix‐related proteins

TIMP‐1 (tissue inhibitor of metalloproteinases‐1) protein or mRNA

Hormonal markers

CYP19 (aromatase cytochrome P450)

Notes:a Limited diagnostic value was assigned when there were at least three studies demonstrating low diagnostic estimates that do not meet or approach the criteria for either replacement or triage test, or demonstrating no differential expression in endometriosis. We advise against further evaluation of these biomarkers for the diagnosis of endometriosis.

Appendix 9. Endometrial biomarkers that possibly have limited diagnostic value in endometriosis

Testa

1. Angiogenesis and growth markers and their receptors

EGF (epidermal growth factor)

FGF‐2 (fibroblast growth factor‐2)

glycodelin A (PP14 or PAEP) (placental protein 14 or progestogen‐associated endometrial protein)

PDGF (platelet derived growth factor)

PIGF (placental growth factor)

PKR1 (prokineticin receptor 1), EG‐VEGF receptor

PKR2 (prokineticin receptor 2), EG‐VEGF receptor

PROK‐1 (prokineticin 1)

TSP‐1 (thrombospondin‐1)

TYMP (thymidine phosphorylase)

2. Apoptosis markers and regulators

Bax (BCL2‐associated X protein)

Bcl‐xL (B‐cell lymphoma‐extra large, or BCL2‐like 1 isoform 1)

Bcl‐xL:Bcl‐xS ratio (ratio B‐cell lymphoma‐extra large/B‐cell lymphoma‐extra small)

3. Cell adhesion molecules and other matrix‐related proteins

α2β1 integrin

α3β1 integrin

α4β1 integrin

α5β1 integrin

α6 β1 integrin

αVβ3 integrin

αVβ5 integrins

αVβ6 integrins

β1 integrin

depolarised α6 integrin

ICAM‐1 (intercellular adhesion molecule‐1) or sICAM‐1 (soluble form of intercellular adhesion molecule‐1)

E‐cadherin

LAMA5 (laminin subunit alpha‐5)

LFA‐3 (CD58) (leukocyte function associated molecule‐3)

MMP‐1 (matrix metalloproteinase‐1)

MMP‐9 (matrix metalloproteinase‐9)

OPN (osteopontin)

PAI‐1/‐2/‐3 (plasminogen activator inhibitors 1/2/3)

PAs: tPA, uPA (plasminogen activators: tissue‐type PA, urokinase‐type PA)

VCAM‐1 (CD106) (vascular cell adhesion molecule‐1)

4. Cell cycle regulatory molecules

cyclin B1

cdc2 (cyclin dependent kinase‐2)

Plk1 (polo‐like kinase‐1)

5. Cell proliferation markers

Ki‐67 (antigen KI‐67 or MKI67, marker of cellular proliferation)

BW 495/36, endometrial epithelial marker

6. Cytoskeleton molecules

cytokeratin 18

CK19 or CYFRA 21‐1 (cytokeratin 19)

vimentin

7. DNA‐repair and telomer maintenance molecules

hTERT (human telomerase reverse transcriptase)

telomerase activity

8. High throughput markers

mitochondrial proteome

mRNAome (mRNA micro‐array)

9. Hormonal markers

EST (oestrogen sulphotransferase)

ER ‐ α (oestrogen receptor ‐ alpha)

ER ‐ β (oestrogen receptor ‐ beta)

LGR7 (leucine‐rich G protein‐coupled receptor 7), relaxin receptor

relaxin

10. Immune system and inflammatory markers

Cytokines

LIF (leukaemia‐inhibitory factor)

TNF‐α (tumour necrosis factor alpha)

Immune cells: peripheral blood mononuclear cells (PBMC)

lymphocytes (all, B‐, T‐cells)

monocytes/ macrophages

NK (natural killer cells)

Interleukins

IL‐1β

IL‐11

IL‐1R1 (interleukin‐1 receptor type II)

other Immune/ inflammatory markers

MPO (myeloperoxidase)

NAG (N‐acetyl‐b‐D‐Glucosaminidase)

11. Mediators of prostaglandin biosynthesis

Akr1B1 mRNA (aldoketoreductase ‐1B1, PGF2a synthase)

Akr1C3 mRNA (aldoketoreductase ‐1C3, PGF2a synthase)

Cox‐1 mRNA (cyclo‐oxygenase‐1)

15‐PGDH mRNA (15‐hydroxyprostaglandin dehydrogenase)

cPGES mRNA (cytosolic PGE2 synthase)

12. Myogenic markers (markers of smooth muscle differentiation )

CALD1 (gene encoding for caldesmon)

13. Nerve sheath and nerve growth markers

NF (neurofilament)

14. Other peptides and proteins

hBD‐2 (human b‐defensin‐2)

15. Transcription factors and signalling molecules

AKT1 (RAC‐alpha serine/threonine‐protein kinase)

JAG1 (jagged‐1 protein)

16. Tumour markers

Ca‐125 (cancer antigen‐125) in menstrual fluid

Notes:a This group comprises the tests that likely to have limited diagnostic value, but where there are insufficient data to confidently comment on their diagnostic role (fewer than three studies with low diagnostic estimates or no differential expression in endometriosis). Further investigation of these biomarkers is possible with a focus on specific phases of the menstrual cycle, specific phenotypes of endometriosis, by implementation of different cut‐off values or by utilising different laboratory methods.

Data

Presented below are all the data for all of the tests entered into the review.

Tests. Data tables by test.

Test	No. of studies	No. of participants
1 PROK‐1 (glandular)	1	24
2 depolarised α‐6 integrin (glandular)	1	49
3 α3β1 integrin (glandular)	1	32
4 α3β1 integrin (stroma)	1	32
5 α4β1 integrin (glandular)	1	32
6 α4β1 integrin (stroma)	1	32
7 β1 integrin (glandular)	1	32
8 β1 integrin (stroma)	1	32
9 hTERT mRNA	1	69
10 Endometrial proteome	2	53
11 Mitochondrial proteome	1	53
12 CYP19	8	444
13 17βHSD2 mRNA	1	53
14 ER‐α (glandular)	1	90
15 ER‐α (stroma)	1	90
16 ER‐β (glandular)	1	90
17 ER‐β (stroma)	1	90
18 IL‐1R2 mRNA (glandular)	1	31
19 IL‐1R2 mRNA (stroma)	1	32
20 IL‐1R2 mRNA (glandular secretory)	1	19
21 IL‐1R2 mRNA (stroma secretory)	1	20
22 Caldesmon (proliferative)	1	35
23 Caldesmon (secretory)	1	35
24 CALD1 mRNA (proliferative)	1	35
25 CALD1 mRNA (secretory)	1	35
26 PGP 9.5	8	429
27 VIP	1	40
28 CGRP	1	40
29 SP	1	40
30 NPY	1	40
31 NF	1	40
32 Combined test (VIP, PGP 9.5, SP)	1	40
33 CA 125 (menstrual fluid)	1	104

1. Test.

PROK‐1 (glandular).

2. Test.

depolarised α‐6 integrin (glandular).

3. Test.

α3β1 integrin (glandular).

4. Test.

α3β1 integrin (stroma).

5. Test.

α4β1 integrin (glandular).

6. Test.

α4β1 integrin (stroma).

7. Test.

β1 integrin (glandular).

8. Test.

β1 integrin (stroma).

9. Test.

hTERT mRNA.

10. Test.

Endometrial proteome.

11. Test.

Mitochondrial proteome.

12. Test.

CYP19.

13. Test.

17βHSD2 mRNA.

14. Test.

ER‐α (glandular).

15. Test.

ER‐α (stroma).

16. Test.

ER‐β (glandular).

17. Test.

ER‐β (stroma).

18. Test.

IL‐1R2 mRNA (glandular).

19. Test.

IL‐1R2 mRNA (stroma).

20. Test.

IL‐1R2 mRNA (glandular secretory).

21. Test.

IL‐1R2 mRNA (stroma secretory).

22. Test.

Caldesmon (proliferative).

23. Test.

Caldesmon (secretory).

24. Test.

CALD1 mRNA (proliferative).

25. Test.

CALD1 mRNA (secretory).

26. Test.

PGP 9.5.

27. Test.

VIP.

28. Test.

CGRP.

29. Test.

SP.

30. Test.

NPY.

31. Test.

NF.

32. Test.

Combined test (VIP, PGP 9.5, SP).

33. Test.

CA 125 (menstrual fluid).

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Al‐Jefout 2007.

Study characteristics
Patient sampling	Primary objective: to evaluate endometrial biopsy and curettage in detecting small nerve fibres in eutopic endometrium for diagnosis of endometriosis Participants: reproductive‐aged women undergoing laparoscopy for suspected endometriosis or infertility Selection criteria: exclusion criteria: current hormonal treatment for endometriosis, pregnancy and unwillingness to participate Study design: observational single‐gate, prospective recruitment and sample collection
Patient characteristics and setting	Clinical presentation: chronic pelvic pain, infertility or both Age: reproductive age, not specified Number enrolled: 37 women Number available for analysis: 37 women (menstrual cycle phase not specified) Setting: Royal Prince Alfred Hospital, a tertiary referral centre Place of study: Sydney, Australia Period of study: 1 January 2006 to 1 December 2006 Language: English
Index tests	Index test: endometrial nerve fibres: PGP 9.5 Description of positive case definition by index test as reported: presence of nerve fibres in the functional layer of endometrium, measured by IHC staining for PGP 9.5 (immunostaining was carried out on a Dako Autostainer Model S3400 (Dako Cytomation, Inc, CA); images analysed by using an Olympus BX51 digital camera (Olympus, Japan)); laboratory technique described Examiners: 3 pathologists, 2 of whom had good experience in nerve fibre counting; 'blinded counting' Interobserver variability: close correlation between the 3 reviewers
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample n/N = 20/37 (54%); stage I‐II 15, stage III‐IV 5 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: visualisation of endometriotic lesions with surgical staging according to rASRM system and peritoneal biopsy confirmation in most cases Examiners: 3 gynaecologists with extensive experience in endometriosis
Flow and timing	TIme interval between index test and reference standard: "prior to laparoscopy" Withdrawals: none reported
Comparative
Notes	Conclusion: Carefully taken endometrial biopsy maybe a reliable means of making a diagnosis of endometriosis. Nerve fibres may be present in other pathologies (adenomyosis, fibroids, endometritis) Comments: The reported estimates for index test performed on specimens obtained by curettage were similar to those obtained by Pipelle endometrial biopsy; not presented in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Al‐Jefout 2009.

Study characteristics
Patient sampling	Primary objective: to detect small unmyelinated nerve fibres immunohistochemically (using the pan‐neuronal marker protein gene product 9.5 (PGP 9.5)) in the functional layer of endometrium in women undergoing diagnostic laparoscopy for pelvic pain or infertility Participants: reproductive‐aged women undergoing laparoscopy for infertility, pelvic pain or both Selection criteria: exclusion criteria: hormonal treatment for 3/12 months prior to surgery, pregnancy, unwillingness to participate Study design: observational single‐gate, prospective recruitment and sample collection
Patient characteristics and setting	Clinical presentation: pelvic pain symptoms alone (n = 52), infertility alone (n = 24), pelvic pain + infertility (n = 20), no pain and no infertility (n = 3) Age: mean age 33.9 years (range 20‐50 years) Number enrolled: 103 women Number available for analysis: 99 women (menstrual cycle phase n = 15; proliferative n = 39; mid‐cycle n = 14; secretory n = 31) Setting: Royal Prince Alfred Hospital, a tertiary referral centre Place of study: Sydney, Australia Period of study: 12 December 2007 to 10 December 2008 Language: English
Index tests	Index test: endometrial nerve fibres: PGP 9.5 Description of positive case definition by index test as reported: presence of endometrial nerve fibres in functional layer by IHC staining for PGP 9.5 (Immunostaining on a Dako Autostainer Model S3400 (Dako, Australia); image analysis by using an Olympus microscope BX51 and digital camera DP70 (Olympus, Japan)); laboratory technique described Examiners: 2 people with experience in nerve fibre counting, blinded to the patients' data and each others' results Interobserver variability: close (98%) correlations between the 2 operators
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample (n/N) = 64/99 (64%): stage I‐II 33, stage III‐IV 31; controls n/N = 35/99 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: visualisation of endometriotic lesions with surgical staging of the disease according to rAFS system; biopsy confirmation of lesions was available in almost all cases. Examiners: 5 gynaecologists with extensive experience in endometriosis
Flow and timing	TIme interval between index test and reference standard: prior to laparoscopy Withdrawals: 4 participants were excluded due to poor sample quality (assessed prior to nerve fibre counting)
Comparative
Notes	Conclusion: Endometrial biopsy, with detection of nerve fibres, provided a reliable diagnosis of endometriosis that is close to the accuracy of laparoscopic assessment by experienced gynaecological laparoscopists
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Yes
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		Low	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Low	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Bokor 2009.

Study characteristics
Patient sampling	Primary objective: to assess density of sensory small‐diameter nerve fibres in the functional layer of endometrium in women with minimal to mild endometriosis and in women with a normal pelvis in order to develop a possible semi‐invasive diagnostic tool for minimal to mild endometriosis Participants: endometrial samples selected from tissue bank, which were collected from women undergoing laparoscopies for infertility, pain or both Selection criteria: inclusion criteria: no medical treatment for 3/12 months preceding surgery, secretory phase of menstrual cycle Study design: observational single‐gate design, prospective sample collection, retrospective selection of cases
Patient characteristics and setting	Clinical presentation: infertility, 100%; dysmenorrhoea, 25% Age: mean age 33 ± 10 years, endometriosis; 32 ± 5 years, controls Number enrolled: 40 women (retrospective selection) Number available for analysis: 40 women (all in secretory phase of menstrual cycle) Setting: University Hospital Gasthuisberg Place of study: Leuven, Belgium Period of study: not provided Language: English
Index tests	Index test: endometrial neural markers: PGP 9.5, VIP, CGRP, SP, NPY, NF Description of positive case definition by index test as reported: nerve fibre density was defined as total number of nerve fibres divided by the total surface area of the examined endometrium; nerve fibres were evaluated by IHC for each marker and counted in HPF areas for the slide section (antibody detection with REAL Detection System, Alkaline Phosphatase/RED, Rabbit/Mouse (Dako); analysis by image analysis software KS400 3.0 (Zeiss, Germany) linked to a Zeiss microscope); the whole surface of each section was evaluated on high‐power images; procedure described; thresholds not pre‐specified; reported cut‐off values: PGP 9.5 − 0.49, VIP − 0.08, CGRP − 0.23, SP − 0.2, NPY − 0.13, NF − 0.19 Examiners: 1 examiner who was blinded to the diagnosis Interobserver variability: NA
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 20/40 (50%), all stage I‐II Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: laparoscopically and histologically confirmed endometriosis, staged according to rASRM Examiners: none mentioned
Flow and timing	TIme interval between index test and reference standard: prior to laparoscopy Withdrawals: none reported
Comparative
Notes	Conclusion: The combined analysis of neural markers PGP 9.5, VIP and SP could predict the presence of minimal to mild endometriosis with 95% sensitivity, 100% specificity and 97.5% accuracy. Prospective studies are required to confirm our findings.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Bourlev 2006.

Study characteristics
Patient sampling	Primary objective: to examine the relationship between microvessel density, mitotic activity in blood vessels and expression of VEGF‐A and its receptors VEGFR‐1 and VEGFR‐2 in eutopic endometrium from women with and without endometriosis and in peritoneal endometriotic lesions Participants: women with laparoscopically confirmed endometriosis and women who underwent laparoscopic sterilisation with confirmed normal pelvis Selection criteria: exclusion criteria: patients with irregular menstrual cycles or gynaecological disorders other than endometriosis and those who had received hormonal treatment within the last 3 months Study design: observational two‐gate design, prospective sample collection, retrospective selection of cases
Patient characteristics and setting	Clinical presentation: not specified Age: mean age 32.7 years (range 22‐44), endometriosis; 29.0 years (range 20–37); controls Number enrolled: 39 women (retrospective selection) Number available for analysis: 39 women (19 in proliferative and 20 in secretory phase of menstrual cycle) Setting: Research Centre of Obstetrics, Gynaecology and Perinatology, Russian Academy of the Medical Sciences Place of study: Moscow, Russia Period of study: not provided Language: English
Index tests	Index test: MVD, Ki‐67, VEGF‐A, VEGFR‐1, VEGFR‐2 Description of positive case definition by index test as reported: MVD was determined by IHC staining as number of microvessels per mm2, calculated as mean value of 5 randomly chosen fields, 0.109 mm2 each; proliferative index (PI) was calculated as the percentage of cells that were Ki‐67 positive (IHC method); VEGF‐A, VEGFR‐1 and VEGFR‐2 levels defined by IC staining intensity and classified 0 when comparable to the negative control, 1 when weak (clearly visible but no more), 2 when between weak and strong, 3 when strong; laboratory technique described in details Examiners: examination by 2 independent observers, unclear if were blinded to diagnosis Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 25/39 (64%), all stage II‐III Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: laparoscopically and histologically confirmed endometriosis, staged according to rASRM Examiners: none mentioned
Flow and timing	TIme interval between index test and reference standard: prior to laparoscopy Withdrawals: none reported
Comparative
Notes	Conclusion: There seems to be a dysregulation of angiogenic activity in the eutopic endometrium of women with endometriosis, and endometriotic lesions with high proliferative activity were accompanied by higher local angiogenic activity and higher levels of VEGF in serum and peritoneal fluid. Comments: For Ki‐67 there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7) For MVD, VEGF‐A, VEGFR‐1 and VEGFR‐2 there was a statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables ‐ not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Casals 2012.

Study characteristics
Patient sampling	Primary objective: to investigate the endometrial expression and co‐expression of αvβ3 integrin and OPN in women with early stages of endometriosis Participants: women undergoing investigations for infertility and women undergoing tubal ligation Selection criteria: inclusion criteria: regular menstruation and not taking any medications Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis and infertile controls ‐ infertility, normal fertility work‐up (other than endometriosis); fertile controls ‐ healthy asymptomatic women with no history of infertility or miscarriage, requesting tubal ligation Age: mean age 31.5 ± 0.7 years, endometriosis; 32.1 ± 0.5 years, unexplained infertility controls; 34.9 ± 0.9 years, fertile controls Number enrolled: 60 women Number available for analysis: 60 women (all in early and 51 in late luteal phase of the cycle) Setting: University hospital ‐ Hospital Clınic, IDIBAPS, University of Barcelona Place of study: Barcelona, Spain Period of study: not provided Language: English
Index tests	Index test: osteopontin (OPN) and αvβ3 integrin expression Description of positive case definition by index test as reported: For all biomarkers positive staining was defined when IHC staining of any intensity was detected in both endometrial glands and luminal surface epithelium (detected by the automated immunohistochemical system TechMate 500 TM (Dako Co., Carpinteria, USA)); in addition, intensity of staining was evaluated by mean staining score and by H‐score method (defined); laboratory technique described Examiners: no information provided; unclear if were blinded to the result of reference standard Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 20/60 (33%): all stage I‐II; controls = 40 Reference standard: laparoscopy Description of positive case definition by reference test as reported: none given Examiners: no information provided
Flow and timing	TIme interval between index test and reference standard: "Endometrial evaluation was performed in all women as a part of a routine infertility work‐up and always before laparoscopyE Withdrawals: In 9 patients late luteal biopsy could not be performed
Comparative
Notes	Conclusions: Endometrial OPN and αvβ3 integrin expression or co‐expression is not impaired during the window of implantation in patients with Stage I–II endometriosis. Comment: For OPN and αvβ3 integrin there was no statistically significant difference between the groups; no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Unclear	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Unclear
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Cetin 2013.

Study characteristics
Patient sampling	Primary objective: to determine whether the number of macrophage cells in the endometrium, the detection of nerve fibres or both can be used in the diagnosis of endometriosis Participants: patients undergoing and having undergone laparoscopy/laparotomy Selection criteria: exclusion criteria: adenomyosis, malignancy, use of hormonal treatment for 3/12 months before surgery Study design: cross‐sectional two‐gate, prospective collection of samples (prospective recruitment, n = 31 and retrospective random selection, n = 34)
Patient characteristics and setting	Clinical presentation: indications for surgery: ovarian cysts, pelvic pain, infertility, tubal ligation, endometrial hyperplasia, AUB, myoma uteri Age: mean age 38.29 ± 6.24 years, endometriosis group; 37.24 ± 5.40 years, controls Number enrolled: 65 women Number available for analysis: 60 women (any stage of menstrual cycle) Setting: university hospital: Istanbul University Istanbul School of Medicine Hospital Place of study: Istanbul, Turkey Period of study: 2006‐2011 Language: English
Index tests	Index test: endometrial nerve fibres ‐ PGP 9.5 and NF; macrophages ‐ CD68 + Description of positive case definition by index test as reported: neural marker results were noted as present or absent via IHC staining (evaluation by using Olympus BX‐51 microscope); macrophages markers reported as sum of IHC stained macrophage cells (evaluation under 400 × magnification in 10 fields); threshold was not reported Examiners: All samples were evaluated by a single pathologist who was blinded to patient data and who is highly experienced in gynaeco‐pathology Interobserver variability: NA
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 31/60 (52%): stage I‐II 6, stage III‐IV 25; controls 29 Reference standard: laparoscopy/laparotomy + histology Description of positive case definition by reference test as reported: visual inspection confirmed by biopsy; staging according to the rAFS Examiners: no information provided
Flow and timing	TIme interval between index test and reference standard: samples were taken preoperatively Withdrawals: 5 women were excluded due to insufficient tissue sample
Comparative
Notes	Conclusions: The detection of nerve fibres in the eutopic endometrium with the markers of PGP 9.5 and NF is not found to be helpful in the diagnosis of endometriosis. Macrophage cells may be helpful in the diagnosis only in the proliferative phase Comments: Nerve fibres markers were not detected in either group For macrophages there was no statistically significant difference between the groups; no data available for meta‐analysis (Appendix 7); subanalysis revealed differential expression only in proliferative phase of the cycle, but there was no data for construction of 2 × 2 tables
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Chen 2004.

Study characteristics
Patient sampling	Primary objective: to study the expression of matrix metalloproteinase‐9 (MMP‐9) and Tissue Inhibitor of Metalloproteinase‐1 (TIMP‐1) in ectopic and eutopic endometrium in patients with endometriosis Participants: patients who underwent surgery for ovarian endometriosis (study group) and for fibroid uterus (controls) Selection criteria: inclusion criteria: no history of hormone therapy and endometriosis drug treatment for 3/12 months before surgery Study design: cross‐sectional two‐gate, prospective collection of samples
Patient characteristics and setting	Clinical presentation: not specified Age: mean age 39 ± 3 years, endometriosis group; 41 ± 3 years, controls Number enrolled: 77 women (54 in proliferative and 23 in secretory cycle phase) Number available for analysis: 60 women Setting: university hospital: Xiamen First Hospital of Xiamen University Place of study: Fujian Xiamen, China Period of study: May 1999 to February 2003 Language: Chinese
Index tests	Index test: MMP‐9, TIMP‐1 Description of positive case definition by index test as reported: the expression of MMP‐9 and TIMP‐1 evaluated by IHC; cells with yellow granules present in the cytoplasm were called positive cells, and the scanning areas for each positive cell were calculated (4 visual fields were selected for each sliced sample, and each visual field was measured 3 times and then their average taken); threshold was not reported Examiners: all samples were evaluated by the same person by using the same equipment Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: ovarian endometriosis Prevalence of target condition in the sample: n/N = 45/77 (58%): stage II 5, stage III‐IV 40; controls 32 (15 included in analysis) Reference standard: laparoscopy/laparotomy, not specified Description of positive case definition by reference test as reported: staging according to rAFS Examiners: no information provided
Flow and timing	TIme interval between index test and reference standard: samples taken at surgery Withdrawals: 17 women from control group were not included in the analysis; reason not specified
Comparative
Notes	Conclusions: The change of expression of MMP‐9 and TIMP‐1 in ectopic endometrium may be related to the pathogenesis of endometriosis. Comments: For TIMP‐1 there was no statistically significant difference between the groups; no data available for meta‐analysis (Appendix 7) For MMP‐9 there was a statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Chen 2013.

Study characteristics
Patient sampling	Primary objective: to investigate the expression of human b‐defensin‐2 (hBD‐2) in the endometrium of patients with endometriosis and explore the potential role of hBD‐2 in pathogenesis of endometriosis Participants: women who were hospitalised at the authors' institution and underwent laparoscopy for ovarian mass (endometriosis group) and hysterectomy for fibroids (controls) Selection criteria: inclusion criteria: regular menstruation, no other endocrine, immune, metabolic or serious medical disorders; exclusion criteria: history of long‐term hormone use or hormone treatment 3/12 months preceding surgery, endometrial hyperplasia or endometrial inflammatory diseases Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: not specified Age: mean age 35.8 ±5.6 years (endometriosis), 37.5 ± 6.8 years (controls) Number enrolled: 50 women Number available for analysis: 50 women (29 in proliferative and 21 in secretory cycle phase) Setting: university teaching hospital ‐ The First Affiliated Hospital, Sun Yat‐sen University Place of study: Guangzhou, China Period of study: February 2008 to April 2010 Language: English
Index tests	Index test: hBD‐2 mRNA and protein, IL‐1β mRNA, TNF‐α mRNA Description of positive case definition by index test as reported: hBD‐2, IL‐1β, and TNF‐α mRNA expression levels defined as the ratio of the target gene to the reference gene (gene expression assessed by using qRT‐PCR; automatically analysed; quantified by a standard curve method). hBD‐2 protein expression according to the percentage of positively IHC stained cells: negative (−),< 5%; positive (+), 5%‐24%; positive (++), 25%‐49%; positive (+++), ≥50% (qualitative positioning method); laboratory techniques described; threshold not provided Examiners: IHC ‐ double‐blinded by two experienced pathologists; PCR ‐ no information; unclear if were blinded to the results of reference standard Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 25/50 (50%): stage I‐II 9, stage III‐IV 16; controls 25 Reference standard: laparoscopy/laparotomy + histology Description of positive case definition by reference test as reported: visual inspection and pathological examination; staging according rAFS system Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: endometrial sample was collected at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: High levels of hBD‐2 gene and protein expressions in the ectopic endometrium of endometriosis patients may be an important contributor in the pathogenesis of endometriosis. TNF‐α and IL‐1β may promote the up regulation of hBD‐2 expression. Comment: #For hBD‐2, IL‐1β, and TNF‐α there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Cho 2012.

Study characteristics
Patient sampling	Primary objective: to evaluate the expression of VEGF and its soluble receptor (sFlt‐1) in peritoneal fluid, peritoneal endometriotic lesions and eutopic endometrial tissues of patients with endometriosis. Participants: women undergoing laparoscopy for different indications (pelvic masses, pelvic pain, suspicious endometriosis, infertility, and diagnostic evaluation) Selection criteria: exclusion criteria: postmenopausal women, previous hormone or GnRH agonist users, patients who had adenomyosis, endometrial pathology, infectious diseases, malignancy, autoimmune diseases, or cardiovascular diseases Study design: observational single‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: pelvic masses, pelvic pain, infertility Age: mean age 31.57 ± 1.04 years, endometriosis group; 34.60 ± 1.48 years, controls Number enrolled: 82 women Number available for analysis: 47 women (22 in proliferative and 25 in secretory phase of menstrual cycle) Setting: Gangnam Severance Hospital, Yonsei University, College of Medicine Place of study: Seoul, Republic of Korea Period of study: not provided Language: English
Index tests	Index test: VEGF mRNA, sFlt‐1 mRNA Description of positive case definition by index test as reported: The amount of target normalised to the endogenous reference (GAPDH), compared to the calibrator and quantified by the 2‐ΔΔCt method (expression levels were measured by SYBR green real‐time PCR using the ABI 7300 instrument (Applied Biosystems, CA)); laboratory technique described; threshold not provided Examiners: no information provided; unclear if blinded to the result of reference standard Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 24/47 (51%): all stage III‐IV Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: visual inspection and pathological confirmation; staging according to rASRM Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: endometrial samples obtained prior to surgery Withdrawals: for 35 women, endometrial samples were not collected ‐ excluded from analysis
Comparative
Notes	Conclusion: No cyclic differences in VEGF mRNA expressions between and within the 2 groups were noted, whereas sFlt‐1 mRNA expressions were significantly lower in patients with endometriosis than in the controls during the proliferative phase. Comments: For VEGF there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7) For sFlt‐1 there was statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables ‐ not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	No
Was a 'two‐gate' design avoided?	Yes
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Da Silva 2014.

Study characteristics
Patient sampling	Primary objective: to evaluate evaluate the presence of MPO, NAG, TNF‐α and VEGF in peripheral and menstrual blood in women with and without endometriosis Participants: women undergoing laparoscopy for chronic pelvic pain, infertility or tubal sterilisation Selection criteria: inclusion criteria: regular menstrual cycles in the 6 months preceding sample collection, no use of hormonal nor anti‐inflammatory agents in the previous 3 months and surgical confirmation or exclusion of endometriosis in agreement with the ESHRE guidelines Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis group ‐ pelvic pain, infertility; controls ‐ infertility or request for sterilisation; none of the women had a significant past medical history Age: median age 36 years, range 31‐48 years Number enrolled: 17 women Number available for analysis: 17 women (all in early proliferative cycle phase, day 1‐4) Setting: university hospital ‐ Hospital das Clýnicas at Universidade Federal de Minas Gerais (HC‐UFMG) Place of study: Belo Horizonte, Brazil Period of study: February 2011 to December 2012 Language: English
Index tests	Index test: NAG, MPO, TNF‐α, VEGF in menstrual fluid Description of positive case definition by index test as reported: NAG and MPO activity (expressed as change in absorbance (OD) at 400 and 450 nm, respectively); TNF‐α and VEGF levels measured by using commercial specific ELISA kits; laboratory technique described; threshold not provided Examiners: no information provided; unclear if blinded to the result of reference standard Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: unable to estimate as two‐gate design (endometriosis n = 10: stage II 5, stage IV 2, undetermined stage 3; controls 7) Reference standard: laparoscopy Description of positive case definition by reference test as reported: diagnosis according to ESHRE guidelines; staging according to rASRM Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: not specified, from the context ‐ perioperative sample collection Withdrawals: none
Comparative
Notes	Conclusion: These findings point to the existence of an increased local inflammatory activity in women with endometriosis. Comments: For NAG, MPO, TNF‐α and VEGF there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Unclear
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Unclear

Dheenadayalu 2002.

Study characteristics
Patient sampling	Primary objective: to determine whether expression of aromatase P450 mRNA in eutopic endometrium is predictive of the presence of pelvic endometriosis Participants: women undergoing laparoscopy for infertility, pelvic pain or tubal sterilisation Selection criteria: not provided Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: infertility (n = 38), pelvic pain (n = 13) and tubal sterilisation (n = 9); 56 (93%) women had regular spontaneous menstrual cycles; none of the patients had been treated surgically for endometriosis; 4 (7%) women were taking the combined oral contraceptive pill Age: mean age 33.2 ± 5.2 years (range 21–46) Number enrolled: 60 women Number available for analysis: 56 women (25 in proliferative and 27 in secretory phase of menstrual cycle) Setting: 4 tertiary centres for reproductive medicine Place: London, UK Period of study: not provided Language: English
Index tests	Index test: aromatase P450 mRNA Description of positive case definition by index test as reported: presence of aromatase P450 mRNA (detected by RT‐PCR with amplification by using Access RT‐PCR System (Promega) positive staining for P450; southern blot for sequence detection); laboratory technique described Examiners: no information provided, unclear if blinded to the results of reference standard Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 34/56 (61%): stage I‐II 28, stage III‐IV 6; controls 22 Reference standard: laparoscopy Description of positive case definition by reference test as reported: visual inspection, rAFS classification Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: Endometrial biopsies were taken at the time of laparoscopy Withdrawals: 4 women were excluded (2 of the biopsies were insufficient for RNA extraction and in 2 specimens RT‐PCR failed to amplify GAPDH mRNA)
Comparative
Notes	Conclusion: Most women with pelvic endometriosis or uterine proliferative disorders express aromatase P450 in the endometrium, regardless of the phase of the menstrual cycle. However, the potential of endometrial aromatase P450 mRNA detection as a clinically useful diagnostic marker of pelvic disease is limited by the observation that approximately 30% of aromatase P450‐negative women were found to have visual evidence of endometriosis.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Unclear
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Unclear	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Ding 2010.

Study characteristics
Patient sampling	Primary objective: to detect specific mitochondrial proteins in eutopic endometrial samples from women with and without endometriosis and to build diagnostic models Participants: women who underwent laparoscopy for benign indications, without clinical symptoms of endometriosis (controls) and women who had endometriosis with laparoscopy Selection criteria: exclusion criteria: hormonal therapy or antiinflammatory medications for 6/12 months prior to surgery; adenomyosis detected at surgery Study design: observational two‐gate, prospective collection of samples
Patient characteristics and setting	Clinical presentation: not specified; control group with "no symptoms of endometriosis" Age: mean age 35.2 ± 6.4 years (controls), 38.4 ± 5.8 years (endometriosis) Number enrolled: 53 women Number available for analysis: 53 women (menstrual cycle phase not provided) Setting: university hospital ‐ Second Affiliated Hospital, Medical College, Zhejiang University Place: Hangzhou, Zhejiang, China Period of study: not stated Language: English
Index tests	Index test: mitochondrial proteome by SELDI‐TOF‐MS (3 potential biomarkers with mass‐to‐charge ratio (m/z) of 15,334, 15,128, and 16,069) Description of positive case definition by index test as reported: SVM model with the highest Youden index (proteins with different physicochemical properties were detected by using Ciphergen SELDI‐TOF‐MS (PBS‐IIplus; Ciphergen Biosystems Inc., CA) and CM10 protein chip); analysis by using SVM classifier and validated by leave‐one cross‐validation); technique described; diagnostic classifiers not pre‐specified Examiners: not stated; unclear if blinded to the results of reference standard Interobserver variability: CV of the selected normalised peaks = 15.2%, CV of the selected peaks' mass = 0.03%; little variation with day‐to‐day sampling, instrumentation or chip variation
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 24/53 (45%): stage I‐II 19, stage III‐IV 5; controls 29 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: visible evidence of endometriosis confirmed by histopathology ‐ 'pelvic organs were examined carefully for the presence and extent of endometriosis; all of the patients were confirmed by pathologic examination of biopsied tissue'; staging according to the rAFS system Examiners: no information provided
Flow and timing	TIme interval between index test and reference standard: endometrial tissues were obtained during the operation Withdrawals: none
Comparative
Notes	Conclusion: The mitochondrial protein fingerprint models identified by protein chip technology had a relative feasibility. 3 biomarkers could be used for early detection and screening of endometriosis
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Elgafor el Sharkwy 2013.

Study characteristics
Patient sampling	Primary objective: to evaluate the diagnostic value of serum measurement of IL‐6 combined with the presence of nerve fibres in the functional layer of endometrium for diagnosis of minimal‐mild endometriosis Participants: women undergoing laparoscopy for infertility, pelvic pain or both Selection criteria: inclusion criteria: reproductive age, follicular phase of the cycle and regular menstrual cycle; exclusion criteria: any current infection, any medication within 1 month prior to laparoscopy, previous surgery for endometriosis and smoking or drinking alcohol Study design: observational, single‐gate, prospective recruitment and sample collection
Patient characteristics and setting	Clinical presentation (n/N): infertility ‐ 91/114; dysmenorrhoea ‐ 64/114; dyspareunia ‐ 17/114; dyschezia ‐ 6/114; other pelvic pain ‐ 35/114 Age: mean age 29 ± 0.6 years, controls; 31 ± 1.1 years, endometriosis Number enrolled: 114 women Number available for analysis: 78 women (all in follicular cycle phase; only control and endometriosis stage I‐II were analysed) Setting: University hospital ‐ Zagazig University Hospital Place of study: Zagazig, Egypt Period of study: December 2010 to April 2012 Language: English
Index tests	Index test: endometrial nerve fibres ‐ PGP 9.5 Description of positive case definition by index test as reported: presence of nerve fibres in the functional layer of endometrium, assessed by IHC staining for PGP 9.5 (an average of 4–5 sections per specimen were examined by using an Olympus microscope) Examiners: 2 pathologists, both of whom have good experience in nerve fibre identification Interobserver variability: close (96%) correlation between the 2 pathologists
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 74/114 (65%): stage I‐II 38, stage III‐IV 36; controls 40 Reference standard: laparoscopy Description of positive case definition by reference test as reported: visualisation of the endometriotic lesions with surgical staging according to rASRM classification Examiners: 3 experienced gynaecologists in endometriosis
Flow and timing	Time interval between index test and reference standard: endometrial biopsy was obtained prior to laparoscopy Withdrawals: data were not available for all women with advanced endometriosis (stage III‐IV), n = 36
Comparative
Notes	Conclusion: Serum IL‐6 and nerve fibres in the functional layer of endometrium will both allow more accurate detection of women who are at risk of having early stages of endometriosis. Comment: The reported data on the serum marker or combination of serum‐endometrial marker are not presented in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Unclear	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Fassbender 2012.

Study characteristics
Patient sampling	Primary objective: to perform a combined mRNA microarray and proteomic analysis on the same eutopic endometrium sample obtained from patients with and without endometriosis Participants: biobank samples collected from women undergoing laparoscopy for infertility, pain or both Selection criteria: exclusion criteria: patients using the oral contraceptive pill or being on chronic medication; surgery within 6 months prior to the time of sample collection Study design: observational single‐gate, retrospective recruitment, prospective sample collection
Patient characteristics and setting	Clinical presentation (n/N): infertility + pain ‐ 27/49; infertility only ‐ 19/49, pain only ‐ 3/49; regular cycle ‐ 35/49 Age: mean age 31.96 ± 4.54 years, median 32 years, range 24‐43 years Number enrolled: 49 women Number available for analysis: 49 women (22 in menstrual and in 27 in early luteal phase of menstrual cycle) Setting: University Hospital Gasthuisberg Place of study: Leuven, Belgium Period of study: Samples were collected between 2005 and 2009 Language: English
Index tests	Index test: endometrial mRNAome by microarray and endometrial proteome by SELDI‐TOF MS (5 peptide peaks of 2,072 m/z; 2,973 m/z; 3,623 m/z; 3,680 m/z and 21,133 m/z) Description of positive case definition by index test as reported: mRNAome ‐ probe sets with an absolute fold change > 2 and a corrected P‐value < 0.05 (detection by using hybridisation microarray (Affymetrix GeneChip Human Gene 1.0 ST Arrays); analysis based on the RMA expression values through xps package 1.7.2 of Bioconductor) Proteome ‐ mass spectrometry peaks, which best discern between endometriosis and controls detected by LS‐SVM model in training and validated in tests set; the average performance of the model was calculated over the 100 splits (detection by using ProteinChip System, Series 4000 SELDI‐TOF‐MS instrument (Bio‐Rad); analysis by using repeated random sub‐sampling cross‐validation); laboratory techniques and analyses described in details; diagnostic classifiers were not pre‐defined Examiners: no information provided; the study was not a blinded study Interobserver variability: proteome ‐ intra‐assay CV of 9% and 10%; interassay CV of 8% and 11% using the reference sample spotted on the CM10 SPA or IMAC CHCA SELDI surface, respectively
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 31/49 (63%): stage I‐II 16, stage III‐IV 15; controls 18 Reference standard: laparoscopy Description of positive case definition by reference test as reported: rAFS classification Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: endometrial samples were collected "from women undergoing laparoscopies"; implies prior to surgery Withdrawals: the proteome data was not available for all women in menstrual cycle phase, n = 22
Comparative
Notes	Conclusions: mRNA expression of eutopic endometrium was comparable in women with and without endometriosis. Proteomic analysis of luteal phase endometrium allowed the diagnosis of endometriosis with high sensitivity and specificity in training and test sets. Comments: For endometrial mRNAome there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7) The reported data on proteome per different stages of endometriosis is not presented The data for proteome in diagnosing endometriosis was available only for secretory phase
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	No
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Unclear
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Gilabert‐Estelles 2003.

Study characteristics
Patient sampling	Primary objective: to analyse several components of the plasminogen activator (PA) pathway and the matrix metalloproteinase (MMP) system in endometriotic tissue, endometrium and peritoneal fluid from women with and without endometriosis Participants: women undergoing laparoscopy/laparotomy at the authors' institution for various indications Selection criteria: exclusion criteria: no hormonal treatment for 3/12 months preceding or pregnant/breastfeeding for 6/12 months before surgery Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis group ‐ pelvic pain (n = 16) and infertility (n = 23); controls ‐ pelvic pain (n = 4), infertility (n = 3), pelvic floor defect (n = 3), asymptomatic (n = 25) Age: mean age 35 years (range 23‐47), endometriosis; 40 years (range 29‐47), controls Number enrolled: 74 women (19 in proliferative, 51 in secretory and 4 in menstrual cycle phase) Number available for analysis: 56 women Setting: university hospital ‐ Hospital Universitario La Fe Place of study: Valencia, Spain Period of study: not provided Language: English
Index tests	Index test: plasminogen activators: tPA, uPA, uPAR, PAI‐1, PAI‐2, PAI‐3, tPA‐PAI3, uPA‐PAI3 and MMPs: TIMP‐1, MMP‐3 Description of positive case definition by index test as reported: antigen levels of each of the tested markers assessed with a commercially available ELISAs (in cytosolic and membrane extracts); functional levels of tPA, PAI‐1 by chromogenic assay and functional level of uPA by immunosorbent activity assay (in cytosolic extract); laboratory techniques described Examiners: no information provided; unclear if blinded to the result of reference standard Interobserver variability: the intra‐ and interassay variabilities for tPA were 8% and 12%; uPA: 4% and 10%; uPAR: 5% and 7%; PAI‐1: 3% and 7%, PAI‐2: 5% and 8%, PAI‐3: 4‐8% and 6‐9%; tPA‐PAI3, uPA‐PAI3: 5‐9% and 7‐12%, MMP‐3: 5% and 9%, TIMP‐1: 4% and 7%
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 39/74 (53%): all stage III‐IV; controls 35 Reference standard: laparoscopy, n = 35/ laparotomy, n = 39 Description of positive case definition by reference test as reported: visual inspection with systematic examination of the abdominal cavity and attention for early and atypical lesions; staging according to the rASRM classification Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: not specified, but from the context appears that the samples were obtained at surgery Withdrawals: 18 endometriosis patients were not included because endometrial tissue was not collected
Comparative
Notes	Conclusion: The increase in antigenic levels of uPA and MMP‐3 in endometrium of women with endometriosis might contribute to the invasive potential of endometrial cells. Once the ovarian endometriotic cyst is developed, an increase in PAI‐1 and TIMP‐1 is detected and significant proteolytic activity is no longer observed, which could explain the frequent clinical finding of isolated endometriotic cyst without invasion of the surrounding ovarian tissue Comments: For tPA, uPAfc, PAI‐1, PAI‐2, PAI‐3, tPA‐PAI3, uPA‐PAI3, TIMP‐1 there was no statistically significant difference between the groups; no data available for meta‐analysis (Appendix 7) For uPAag and MMP‐3 there was a statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Unclear
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Gilabert‐Estelles 2007.

Study characteristics
Patient sampling	Primary objective: to analyse mRNA expression and protein levels of VEGF, PlGF and TSP‐1 in endometrium and peritoneal fluid of women with and without endometriosis and in ovarian endometrioma patients Participants: women of reproductive age undergoing laparoscopy for various indications Selection criteria: exclusion criteria: menstrual phase, irregular menstruation, pregnancy or breastfeeding in the previous 6/12 months, hormonal treatment for 3/12 months before surgery Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: indications for surgery ‐ clinical or sonographic suspicion of endometriosis (endometriosis group) and pelvic pain (12%), sterility (18%) or tubal sterilisation (70%) in controls Age: mean age: 32.7 years (range 19–46), endometriosis; 36.5 years (range 20–52 years), controls Number enrolled: 121 women (56 in proliferative and 65 in secretory cycle phase) Number available for analysis: 90 women Setting: University hospital ‐ Hospital Universitario La Fe Place of study: Valencia, Spain Period of study: not provided Language: English
Index tests	Index test: VEGF, PlGF, TSP‐1, uPA, PAI‐1, MMP‐3, TIMP‐1 Description of positive case definition by index test as reported: mRNA expression by RT‐PCR (analysis in a LightCycler apparatus v 3.5 (Roche Molecular Biochemicals, Germany), normalised to b‐actin); antigen protein levels of VEGF, PIGF, uPA, PAI‐1, MMP‐3 and TIMP‐1 by commercially available ELISAs and TSP‐1 ‐ by indirect ELISA; laboratory technique and sample handling described Examiners: no information provided; unclear if blinded to the result of reference standard Interobserver variability: CV for TSP‐1 protein assay was 4.8%, not reported for other tests
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 71/121 (59%): all stage III‐IV; controls 50 Reference standard: laparoscopy + histopathology Description of positive case definition by reference test as reported: visual inspection of abdominal cavity confirmed by histology; staging according to rASRM classification Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: not specified, but from the context appears that the samples were obtained at surgery Withdrawals: 31 endometriosis patients were excluded due to insufficient sample quantity
Comparative
Notes	Conclusion: Endometrium and peritoneal fluid from women with endometriosis have increased levels of VEGF, uPA and MMP‐3. Therefore, the development of endometriotic implants at ectopic sites may be facilitated, promoting the progress of endometriosis Comments: For PIGF, TSP‐1 and PAI‐1, TIMP‐1 mRNA there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7) For VEGF, uPA, MMP‐3 and TIMP‐1 protein there was statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Unclear
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Hatok 2011.

Study characteristics
Patient sampling	Primary objective: to prospectively evaluate whether detection of CYP19 mRNA in eutopic endometrium could be used as specific screening tool for endometriosis as well as reflect its severity Participants: women with regular menstrual period undergoing laparoscopic surgery for pelvic pain Selection criteria: exclusion criteria: irregular menstrual cycle, secretory phase, menopause, history of hormonal treatment before surgery, previous abdominal surgery, pregnancy, history of IVF prior to laparoscopy Study design: observational single‐gate, prospective recruitment and sample collection
Patient characteristics and setting	Clinical presentation: pelvic pain Age: mean age 43.02 ± 6.8 years (range 25‐55) Number enrolled: 105 women Number available for analysis: 101 women (all in early proliferative phase) Setting: University hospital ‐ Jessenius Faculty of Medicine, Comenius University Place of study: Bratislava, Slovak Republic Period of study: not provided Language: English
Index tests	Index test: Aromatase P450 (CYP19) mRNA expression Description of positive case definition by index test as reported: "[i]ncreased aromatase mRNA expression" ‐ not specified (detection by RT‐PCR; analysis by gel electrophoresis with GeneTools software; quantified by relative quantification: expression of target relative to housekeeping gene (β‐actin); laboratory technique described; diagnostic threshold not provided Examiners: no information provided Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 30/101 (30%): stage I‐II 15, stage III‐IV 15; adenomyosis ‐ 25, healthy controls ‐ 46, total non‐endometriosis controls 71 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: surgical inspection confirmed by histological examination; rAFS classification Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: tissue samples were obtained at surgery Withdrawals: 4 participants were excluded for low sample quality
Comparative
Notes	Conclusion: Significantly increased level of CYP19 mRNA was detected in patients, and this expression was significantly dependent on disease severity. These findings provide direct evidence that screening for endometrial aromatase could be beneficial in the prediction of oestrogen dependent disease. Comment: For the purpose of this review diagnostic estimates were calculated for endometriosis versus non‐endometriosis (controls included women with adenomyosis)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Unclear
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		Unclear	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Hudelist 2007.

Study characteristics
Patient sampling	Primary objective: to analyse the expression of both aromatase and EST in the uterine and ectopic endometrium of patients with endometriosis Participants: women undergoing surgery for suspected endometriosis or non‐malignant conditions (e.g. fibroids) Selection criteria: exclusion criteria: history of PID, malignancy, adenomyosis uteri, intake of GnRH agonists, or exposure to steroids within the 6/12 months prior to surgery Study design: observational two‐gate, prospective collection of samples
Patient characteristics and setting	Clinical presentation: pain ‐ 26/35 and infertility ‐ 9/35, endometriosis; controls ‐ not specified Age: median age 39.6 years, endometriosis; 37.8 years, controls Number enrolled: 68 women Number available for analysis: 68 women (38 in proliferative and 27 in secretory cycle phase) Setting: Division of Special Gynecology, University of Vienna and LKH Villach hospital Place of study: Villach and Vienna, Austria Period of study: 2002‐2005 Language: English
Index tests	Index test: aromatase and EST Description of positive case definition by index test as reported: the expression was quantified by immunoreactive score (IRS), defined by IHC stain intensity: IRS > 8 ‐ strong immunoreactivity; > 4 and ≤ 8 ‐ moderate; > 0 and ≤ 4 ‐ weak, 0 ‐ negative; laboratory techniques described; no threshold provided Examiners: independent blinded assessment by 2 pathologists Interobserver variability: not reported; decision by consensus
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 35/68 (40%) (all stage I‐II); controls 33 Reference standard: laparoscopy/laparotomy Description of positive case definition by reference test as reported: visual inspection, AFS classification Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: eutopic tissue specimens were obtained at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: The elevated expression of aromatase in eutopic and ectopic endometrium from patients with endometriosis in the presence of comparable EST provides further evidence for unopposed local biosynthesis of estrogens in endometriosis. Comments: For aromatase the data is presented only for glandular cells, diagnostic threshold defined by the review authors as presence or absence of expression; there was no statistically significant difference between the groups in the overall uterine endometrium (not presented) For EST there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Johnson 2004.

Study characteristics
Patient sampling	Primary objective: to study the estrogenic microenvironment in eutopic endometria of women with and without endometriosis. Participants: women undergoing laparoscopy for suspected endometriosis or sterilisation Selection criteria: exclusion criteria: endocrine therapy, such as GnRH agonists, danazol or oral contraceptives 6/12 months prior to surgery Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis group ‐ CPP, severe dysmenorrhoea, infertility or both; controls ‐ eumenorrhoeic fertile women requesting sterilisation Age: mean age 32.6 ± 5.8 years, endometriosis; 36.2 ± 4.8 years, controls Number enrolled: 43 women Number available for analysis: 36 women (8 in proliferative and 28 in secretory cycle phase) Setting: University hospital ‐ Hospital Clínico San Borja Arriarán, Universidad de Chile Place of study: Santiago, Chile Period of study: not specified Language: Spanish
Index tests	Index test: aromatase cytochrome P450 mRNA, E2, Aromatase activity Description of positive case definition by index test as reported: presence of aromatase mRNA expression as detected by RT‐PCR with amplification (semi‐quantification by using the System Analysis and Kodak Electrophoresis Documentation (EDAS 290)/Kodak 1D Image Analysis (Rochester, NY)); aromatase activity was determined by conversion of testosterone to E2 (assessed as E2 level after incubation with testosterone); baseline and stimulated E2 was measured by RIA (Diagnostic Sys Lab, Webster, Texas) with sensitivity 5.0 pg/ml; laboratory techniques described; no thresholds provided Examiners: no information provided; unclear if were blinded to the result of reference standard Interobserver variability: For E2 intra‐/interassay CV = 4.1%/6.7%; for aromatase ‐ none reported
Target condition and reference standard(s)	Target condition: Endometriosis Prevalence of target condition in the sample: n/N = 23/43 (53%): stage I‐II 11, stage III‐IV 12; controls 20 Reference standard: laparoscopy Description of positive case definition by reference test as reported: classification by rASRM Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: endometrial samples were obtained at laparoscopy Withdrawals: 7 patients were not included in the experiments, reasons not provided
Comparative
Notes	Conclusion: In women with endometriosis, the microenvironment in the endometria is estrogenic as a consequence of an increased expression and activity of the P450 aromatase. Comment: For aromatase activity and baseline Estradiol there was a statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Yes
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Kim 2007.

Study characteristics
Patient sampling	Primary objective: to evaluate the replication potential of the endometrium in patients with endometriosis by examining hTERT mRNA expression and telomerase activity. Participants: women undergoing laparotomy or operative laparoscopy at the authors institution Selection criteria: exclusion criteria: postmenopausal women, previous hormone or GnRH agonist users, and patients with adenomyosis or endometrial cancer/hyperplasia/polyps Study design: observational, unclear if single‐ or two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: not specified Age: mean age 32.7 ± 6.8 years, endometriosis; 34.6 ± 6.7 years, controls Number enrolled: 60 women Number available for analysis: 60 women (26 in proliferative and 34 in secretory cycle phase) Setting: University hospital ‐ Yongdong Severance Hospital, Yonsei University College of Medicine Place of study: Seoul, South Korea Period of study: September 2005 to March 2006 Language: English
Index tests	Index test: hTERT mRNA, telomerase activity Description of positive case definition by index test as reported: hTERT mRNA expression level as measured by RT–PCR (TaqMan methodology, normalised to the endogenous control GAPDH and compared to the calibrator as per 2‐ΔΔCt method); telomerase ‐ relative telomerase activity (RTA), defined as the activity equivalent to that in 100 molecules of TSR (determined by the telomerase repeat amplification protocol assay, referenced to primary source (Kim et al.,1994); laboratory techniques described; no thresholds provided Examiners: no information provided Interobserver variability: no information provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 30/60 (50%): stage I‐II 8, stage III‐IV 22; controls 30 Reference standard: laparoscopy/laparotomy Description of positive case definition by reference test as reported: disease stages established by rASRM Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: biopsies were performed at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: The demonstrated overexpression of hTERT mRNA and telomerase activity in the endometrium of endometriosis patients suggests that replication potential of endometrial cells may have an important role in the pathogenesis of endometriosis. Comments: For hTERT there was a statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review For telomerase activity there was no statistically significant difference between the groups; no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Unclear
		High	Unclear
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	No
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Klentzeris 1995.

Study characteristics
Patient sampling	Primary objective: to investigate whether the endometrium of women with endometriosis differs immunologically from the endometrium of normal fertile women Participants: women undergoing laparoscopy for infertility or requesting tubal sterilisation/reversal Selection criteria: inclusion criteria: reproductive age (20‐40), regular cycles, no hormonal treatment at biopsy, no IUD for 3/12 months preceding surgery Study design: observational, two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis group ‐ infertility; controls ‐ fertile women requesting tubal sterilisation or reversal Age: mean age 33.2 ± 1.6 years (range 22‐41), endometriosis; 35.4 ± 2.8 years (range 24‐40), controls Number enrolled: 36 women Number available for analysis: 33 women (all in secretory cycle phase) Setting: unclear; the authors affiliations are academic departments: Queens Medical Centre and Nottingham City Hospital Place of study: Nottingham, UK Period of study: not reported Language: English
Index tests	Index test: endometrial leukocytes: T lymphocytes (CD8+, CD4+ cells), endometrial granulated lymphocytes (CD56+ and CD38+ cells), macrophages (CD68+ cells), natural killers (CD16+ cells) and B lymphocytes (CD22+ cells) Description of positive case definition by index test as reported: positive IHC staining ‐ the area of precipitate within the endometrial stroma for each monoclonal antibody, which was proportional to the number of positive cells (detected by using Quantimet 970 image analyser (Cambridge instruments Ltd, UK); 5 non‐overlapping fields of view were examined per biopsy in a systematic random sampling pattern; magnification × 250); laboratory techniques described; no thresholds provided Examiners: no information provided Interobserver variability: no information provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 16/33 (59%): 7 stage 1, 4 stage 2, 2 stage 3, 3 stage 4; controls 17 Reference standard: laparoscopy Description of positive case definition by reference test as reported: disease stages established by rASRM Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: endometrial biopsies were performed at surgery Withdrawals: 3 women were excluded because of "retarded endometrium"
Comparative
Notes	Conclusion: This study has shown that the endometrial lymphoid tissue of women with endometriosis does not differ qualitatively or qualitatively from that of normal fertile controls. However, functional differences of endometrial leukocytes between the two groups cannot be excluded. Comments: For all the endometrial leukocytes there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Laudanski 2014.

Study characteristics
Patient sampling	Primary objective: to compare the expression level of the most relevant angiogenesis‐related genes in the eutopic endometrium of women with and without endometriosis Participants: regularly menstruating premenopausal women, undergoing diagnostic or surgical laparoscopic surgery for non‐malignant ovarian lesions Selection criteria: Exclusion criteria: patients with autoimmune disease, PID, adenomyosis, myomas, uterine adhesion and anomalies, DUB and those who took NSAIDs, GnRH agonists and steroids for the past 3/12 months Study design: observational single‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: infertility ‐ 19/32, ovarian cysts ‐ 22/32 Age: mean age 31.85 ± 0.93 years, endometriosis; 29.73 ± 0.94 years, controls Number enrolled: 32 women Number available for analysis: 29 women (all in proliferative cycle phase) Setting: University hospital ‐ Medical University of Bialystok Place of study: Bialystok, Poland Period of study: not provided Language: English
Index tests	Index test: AKT1, TYMP, JAG1, LAMA5 and TIMP‐1 mRNAs and proteins Description of positive case definition by index test as reported: gene expression values measured by PCR array and calculated by 2‐ΔΔCt method (experiments on ABI 7900HT instrument (Life Technologies, Applied Biosystems, Poland); protein levels assessed by Western Blot and ELISA; laboratory techniques described; no thresholds provided Examiners: no information provided, unclear if blinded to the results of reference standard Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 18/32 (56%): all stage III‐IV; controls 14 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: laparoscopic findings according to the rAFS classification; each case was confirmed by histopathology Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: endometrial samples were collected in the operating room before the laparoscopic procedure Withdrawals: In 3 women from endometriosis group protein analysis was not performed (insufficient sample)
Comparative
Notes	Conclusion: Changes in the expression of selected genes might lead to or be a consequence of an early defect in the physiological activity of proliferative endometrium ultimately resulting in its overgrowth outside the uterine cavity. Comments: The reported data for mRNA levels are not included in the review as data of high throughput method was not corrected for multiplicity and has not verified in singleplex experiments For AKT1, TYMP, JAG1, LAMA5, TIMP‐1 proteins there was no statistically significant difference between the groups; no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	No
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Lawson 2008.

Study characteristics
Patient sampling	Primary objective: to elucidate the endometriosis associated changes in expression of IL‐1 receptors and to investigate further at the mRNA level the expression of IL‐1R1 and IL‐1R2 in ectopic endometrial tissue in comparison with endometrial tissue from matched women with endometriosis and normal controls Participants: women undergoing laparoscopy for infertility, pelvic pain or tubal ligation Selection criteria: inclusion criteria: no signs of hyperplasia or neoplasia, no use of hormonal or anti‐inflammatory medications for 3/12 months before surgery Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: infertility, pelvic pain or tubal ligation Age: mean age 35 ± 4 years Number enrolled: unclear Number available for analysis: 34 women (13 in proliferative and 21 in secretory phase of menstrual cycle) Setting: University hospital ‐ Hopital Saint‐Francois d’Assise, Centre Hospitalier Universitaire de Québec Place of study: Quebec, Canada Period of study: not stated Language: English
Index tests	Index test: IL‐1R1 and IL‐1R2 mRNA Description of positive case definition by index test as reported: mRNA hybridisation score defined as intensity of hybridisation signal green to yellow from 3 evaluations in 3 randomly selected areas, scored as following: 0 = absent, 1 = light, 2 = moderate and 3 = intense signal (detection by FISH, viewing using a Leica microscope (Leica Corp. Germany) with image analysis system (ISIS, Metasystems, Germany)); no threshold cut‐off given; personal communication with the authors: for epithelial cells positive test defined as HS < 3, for stromal cells ‐ HS < 2 Examiners: 2 observers who had no knowledge of clinical status and diagnosis of the patients Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N =17/34 (50%), all stage I‐II; controls 17 Reference standard: laparoscopy Description of positive case definition by reference test as reported: staged according to rAFS classification Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: tissues obtained during laparoscopy Withdrawals explained: unclear number of recruited, different number for stroma and glandular cells experiments; withdrawals not explained
Comparative
Notes	Conclusion: The present study has shown a combined defect in IL‐1R1 and IL‐1R2 gene expression in endometriosis. Comments: For mRNA levels by RT‐PCR there was no data available to construct 2 × 2 tables; not presented in the review; demonstrated similar trends to the presented FISH data For IL‐1R2 there was statistically significant difference in secretory and no difference in proliferative cycle phase, hence the diagnostic estimates reported for the pooled phases and for sub‐analysis in secretory phase For IL‐1R1 there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Lee 2007.

Study characteristics
Patient sampling	Primary objective: to compare the levels of expression of EGF, FGF‐2 and PDGF‐A mRNA in the eutopic endometrium of women with and without endometriosis Participants: patients of reproductive age who underwent laparoscopic surgery, exploratory laparotomy, or transabdominal hysterectomy Selection criteria: inclusion criteria: regular menstrual cycles, no hormonal therapy in the past 6/12 months; exclusion criteria: PID, DUB or any endometrial pathology; patients with visually suspected endometriotic lesions without histological confirmation Study design: observational, unclear if single‐ or two‐gate design, prospective collection of samples
Patient characteristics and setting	Clinical presentation: endometriosis ‐ not specified; controls ‐ fertile women undergoing surgery for ovarian cysts Age: mean 35.2 ± 6.3 years (endometriosis), 37.5 ± 5.1 years (controls) Number enrolled: 66 women Number available for analysis: 66 women (29 in proliferative and 37 in secretory cycle phase) Setting: College of Medicine, University of Ulsan, Asan Medical Centre Place of study: Seoul, South Korea Period of study: not provided Language: English
Index tests	Index test: PDGF‐A, EGF, FGF‐2 mRNA Description of positive case definition by index test as reported: levels of mRNAs assessed by RT‐PCR (cycle threshold was defined as the cycle number at which the fluorescent signal passed a fixed value; the expression of each target gene was normalised to the expression of the control gene (h‐GAPDH)); laboratory technique described; no threshold provided Examiners: none stated, unclear if were blinded to the results of reference standard Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 35/66 (53%), all stages III‐IV; controls 31 Reference standard: laparoscopy/ laparotomy + histology Description of positive case definition by reference test as reported: endometriosis was staged according to rASRM Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: tissue samples were obtained at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: The expression level of PDGF‐A, but not EGF and FGF‐2, might be decreased during the secretory phase in the eutopic endometrium of women with advanced stage endometriosis. Comment: For PDGF‐A, EGF, FGF‐2 mRNA there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Unclear
		Unclear	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Lee 2010.

Study characteristics
Patient sampling	Primary objective: to study the expression of vascular endothelial growth factor (VEGF), endocrine gland‐derived VEGF (EG‐VEGF/PK1), and its receptors (PKR1 and PKR2) in eutopic and ectopic endometrial tissues Participants: infertile women undergoing diagnostic laparoscopy for tubal patency Selection criteria: inclusion criteria: regular cycles, no inflammatory or hormonal medication for at least 3/12 months before surgery Study design: observational single‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: infertility Age: mean age 34 years (range 29–38), endometriosis; 33 years (range 24–39), controls Number enrolled: 48 women Number available for analysis:48 women (23 in proliferative and 25 in secretory phase of the cycle) Setting: University reproduction unit, LKS Faculty of Medicine, the University of Hong Kong Place of study: Hong Kong, China Period of study: not stated Language: English
Index tests	Index test: EG‐VEGF, PKR 1 and PKR2 mRNA Description of positive case definition by index test as reported: mRNA levels detected by RTPCR following laser‐captured microdissection (detection by ABI 7500 Sequence Detector (Applied Biosystems, CA); relative gene expression values were calculated by the 2‐ΔΔCt method, normalised to endogenous control 18S); laboratory technique described; no threshold provided Examiners: none stated; unclear if were blinded to the results of reference standard Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 15/48 (31%); stages not specified; controls 33 Reference standard: laparoscopy Description of positive case definition by reference test as reported: rAFS scoring was used to classify the stages of endometriosis Examiners: none stated
Flow and timing	Time interval between index test and reference standard: not specified, but from context appears that tissue were collected at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: Expression of VEGF mRNA and protein was higher in red peritoneal endometriotic lesion than in other types of peritoneal lesions. There was no difference in the mRNA expression of VEGF between eutopic endometrium and ovarian endometriotic tissues. Comment: For EG‐VEGF, PKR 1 and PKR2 mRNA there was no statistically significant difference between the groups; no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Unclear
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Unclear

Leslie 2013.

Study characteristics
Patient sampling	Primary objective: to assess presence of nerve fibres in endometrial functional layer as a specific and sensitive marker of concurrent peritoneal endometriosis Participants: patients undergoing laparoscopy for suspected endometriosis Selection criteria: exclusion criteria: histological diagnosis not available (ablated lesions). Hormonal pretreatment was not an exclusion Study design: observational single‐gate, prospective sample collection
Patient characteristics and setting	Clincial presentation: pain ‐ 45/68, infertility ‐ 14/68; adnexal mass/ menorrhagia ‐ 7/68; hormonal therapy ‐ 11/68; information was not available in 1 control and 11 cases Age: mean age 35 years (range 21–53) Number enrolled: 68 women Number available for analysis: 68 women (25 in proliferative, 19 in secretory cycle phase; 24 ‐ unclear/hormonal treatment) Setting: university hospital ‐ King Edward Memorial Hospital and private hospital ‐ Hollywood Hospital Place of study: Perth, Australia Period of study: 2006‐2011 Language: English
Index tests	Index test: endometrial functional layer nerve fibres ‐ PGP 9.5 Description of positive case definition by index test as reported: presence of functional layer nerve fibres as detected by PGP 9.5 IHC staining (lower uterine, cervical and basal layer staining was not considered; magnification using a Leica DM2500 light microscope); laboratory technique described Examiners: Single pathologist unaware of the results for the reference standard; positive and equivocal biopsies were blindly reviewed by the 2nd pathologist, disagreement resolved by consensus Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition: n/N = 47/68 (69%); stage I‐IV, number of women in each subgroup of severity not reported; controls 21 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: not stated (all laparoscopic procedures and endometrial biopsies were performed according to the individual clinician’s standard practice) Examiners: not stated
Flow and timing	Time interval between reference standard and index test: "Laparoscopy and concurrent endometrial biopsy" ‐ implies at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: Endometrial nerve fibre assessment in a general patient population undergoing laparoscopy and using routine endometrial sampling and immunohistochemical analysis proved neither sensitive nor specific in the diagnosis of endometriosis.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Yes
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		Low	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Low	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Li 2006.

Study characteristics
Patient sampling	Primary objective: to investigate mRNA expression of matrix metalloproteinase (MMP‐9) and tissue inhibitor of metalloproteinase (TIMP‐1) in ectopic endometriosis tissue and uterine endometrium from women with and without endometriosis Participants: endometriosis patients who underwent surgical treatment (study group) and women who underwent surgery for ovarian cyst or for fibroid uterus (controls) Selection criteria: inclusion criteria: normal menstrual regularity, no other internal or surgical diseases, and no history of hormonal therapy 3 months prior to surgery Study design: cross‐sectional two‐gate, prospective collection of samples
Patient characteristics and setting	Clinical presentation: not specified Age: mean age 33 ± 5 years Number enrolled: 58 women Number available for analysis: 55 women (23 in proliferative and 32 in secretory cycle phase) Setting: university hospital: Peking Union Medical College Hospital, Peking Union Medical College, Chinese Academy of Medical Science Place of study: Peking, China Period of study: January 2000 to December 2000 Language: Chinese
Index tests	Index test: MMP‐9, TIMP‐1 Description of positive case definition by index test as reported: The expression of MMP‐9 and TIMP‐1 evaluated by RT‐PCR (normalised to β‐actin; detection and semi‐quantification of PCR products by using gel electrophoresis; a gel scanner was used to determine the presence of the target band and the stripe absorbance (A) value); laboratory technique described; threshold was not reported Examiners: no information provided Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 38/58 (66%): stage I‐II 12, stage III‐IV 26; controls 20 Reference standard: laparoscopy/laparotomy Description of positive case definition by reference test as reported: staging according to rAFS Examiners: no information provided
Flow and timing	TIme interval between index test and reference standard: samples were taken at surgery Withdrawals: 3 women from endometriosis group were not included (failure to obtain endometrial samples)
Comparative
Notes	Conclusions: An increase of MMP‐9 mRNA expression of eutopic endometrium with endometriosis might enhance the endometrial implantation ability, thus facilitating the ectopic implantation of endometrium. Ectopic lesions express significantly less TIMP‐1 mRNA, indicating they have increased invasive ability, which might facilitate the development of endometriosis. Comments: For TIMP‐1 there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7) For MMP‐9 there was statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Unclear
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Liu 2008.

Study characteristics
Patient sampling	Primary objective: to investigate the distribution of ER isoforms in endometriosis and eutopic endometrium Participants: patients who underwent surgical treatment of ovarian endometrioma (study group) and total hysterectomy for cervical lesions in the authors' institution (controls) Selection criteria: inclusion criteria: regular menstrual cycle and no history of hormonal therapy 6 months prior to surgery Study design: cross‐sectional two‐gate, prospective collection of samples
Patient characteristics and setting	Clinical presentation: not specified Age: mean age 36 years (range 24‐49), endometriosis; 40 years (range 27‐46), controls Number enrolled: 90 women Number available for analysis: 90 women (45 in proliferative and 45 in secretory cycle phase) Setting: university hospital: Chinese People's Liberation Army General Hospital, affiliated to Peking Union Medical College Place of study: Beijing, China Period of study: January 2004 to December 2006 Language: Chinese
Index tests	Index test: ER‐α and ER‐β Description of positive case definition by index test as reported: positive expression of ER‐α and ER‐β was evaluated by IHC and classified by using Sinicrope Improved Semi‐quantitative Method (the percentage of stained cells and the intensity of staining was scored each by a 4 tier point system in each sliced section; both scores were added up and interpreted as: < 0.5 points means negative; 0.5‐1.5 points means weakly positive (+); 1.6‐2.5 points means averagely positive (++); > 2.5 points means strongly positive (+++).; laboratory technique described; threshold was not pre‐specified Examiners: no information provided Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 60/90 (67%): stage not reported; controls 30 Reference standard: laparotomy/laparoscopy Description of positive case definition by reference test as reported: no information provided Examiners: no information provided
Flow and timing	TIme interval between index test and reference standard: samples were taken at surgery Withdrawals: none reported
Comparative
Notes	Conclusions: Both ER‐α and ER‐β have higher expression levels in eutopic endometrium of patients with ovarian endometriotic cysts. ER‐β is predominantly expressed in endometriotic cysts, whereas the expression of ER‐α is limited. The different distribution of ER‐α and ER‐β may play an important role in the development of ovarian endometriosis. Comments: The authors also report findings for endometriotic cysts; not included in this review The authors also report findings for each cycle phase; not presented in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Mafra 2014.

Study characteristics
Patient sampling	Primary objective: to evaluate the expression of telomerase in the endometrium and peritoneal endometriotic lesions from women with endometriosis and controls Participants: women who underwent laparoscopy for infertility ± pain (recruited from endometriosis clinic) and women undergoing sterilisation (recruited from family planning clinic) Selection criteria: inclusion criteria: regular cycle, no use of hormonal therapy for 3/12 months prior to surgery; exclusion criteria: personal/family history of autoimmune or chronic disorders Study design: observational, two‐gate design, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis: infertility ‐ 25/25, pelvic pain ‐ 15/25; controls: fertile women requesting tubal ligation Age: mean age 33.9 ± 3.3, endometriosis; 35.3 ± 4.6 years, controls Number enrolled: 69 women Number available for analysis: 69 women (all in luteal phase of menstrual cycle) Setting: Human Reproduction and Genetics Center of the Faculdade de Medicina do ABC Place of study: Santo Andre, Brazil Period of study: not stated Language: English
Index tests	Index test: hTERT mRNA Description of positive case definition by index test as reported: Positive hTERT mRNA expression by qRT‐PCR (detection by the TaqMan methodology using StepOne Real‐Time PCR System (Applied Biosystems, CA); target was normalised to the endogenous reference (GAPDH) and compared to calibrator by the 2‐ΔΔCt method); laboratory technique described Examiners: not stated; unclear if were blinded to reference standard Interobserver variability: not stated
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 25/69 (36%): stage I‐II 15, stage III‐IV 10; controls 44 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: visual inspection with obligatory histological confirmation of the disease; staging according ASRM classification Examiners: not stated
Flow and timing	Time interval between index test and reference standard: < 12 months Withdrawals: none reported
Comparative
Notes	Conclusion: There was an association between the expression of telomerase (hTERT mRNA) and the genesis and progression of endometriosis.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Yes
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Unclear	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Makari 2012.

Study characteristics
Patient sampling	Primary objective: to test the hypothesis that nerve fibres are present in a higher density in endometrium of patients with endometriosis when compared to women with a normal pelvis Participants: patients that presented for laparoscopy for infertility, pelvic pain or both Selection criteria: inclusion criteria: reproductive age (18‐45 years); exclusion criteria: hormonal treatment 3/12 months before surgery, pregnancy or oncology cases Study design: observational single‐gate, prospective collection of samples
Patient characteristics and setting	Clinical presentation: dysmenorrhoea ‐ 10/20, chronic pelvic pain ‐ 11/20, infertility, dyspareunia, dysuria, dyschezia Age: mean age 36.1 ± 6.10, endometriosis; 30 13 ± 6.38 years, controls Number enrolled: 20 women Number available for analysis: 20 women (15 in proliferative and 5 in secretory cycle phase) Setting: university hospital ‐ Hospital of Lithuanian University of Health Sciences Kaunas Clinics Place of study: Kaunas, Lithuiania Period of study: 2009‐2011 Language: Lithuianian
Index tests	Index test: endometrial nerve fibres ‐ PGP 9.5 Description of positive case definition by index test as reported: presence of nerve fibres as detected by IHC staining for PGP 9.5 (evaluatIoin under × 400 magnification, microscope Olympus BX51; the number of immunoreactive nerve fibres was also calculated for each cross‐sectional area to assess nerve fibre density) Examiners: not stated, unclear if were blinded to the reference standard Interobserver variability: not stated
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 10/20 (50%): stage I‐II 4, stage III‐IV 6; controls 10 Reference standard: laparoscopy + histology Description of positive case definition by reference standard as reported: staging according to rASRM classification Examiners: not stated
Flow and timing	Time interval between index test and reference standard: tissue samples were collected at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: Endometrial biopsy for detecting density of nerve fibres by usage of PGP 9.5 provided a reliable marker for diagnosing endometriosis. The density of small nerve fibres was significantly higher in endometrium of patients with endometriosis when compared to women with a normal pelvis.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Unclear	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Matsuzaki 2006a.

Study characteristics
Patient sampling	Primary objective: to investigate mRNA expression of aromatase and βHSD2 in eutopic and ectopic endometrium from patients with endometriosis Participants: patients undergoing laparoscopy for deep endometriosis (cases) and women undergoing tubal reversal or sterilisation (controls) Selection criteria: inclusion criteria ‐ regular cycle, exclusion criteria: no hormonal remedies or IUCD for 6/12 months prior to surgery Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis ‐ not specified; controls ‐ fertile women requesting tubal sterilisation or reversal Age: reproductive age Number enrolled: 54 women Number available for analysis: 54 women (15 in late proliferative, 12 in early secretory, 14 in mid secretory, 13 in late secretory phase of the cycle) Setting: university hospital ‐ Polyclinique de l'Hotel Dieu, CHU Clermont‐Ferrand Place of study: Clermont‐Ferrand, France Period of study: beginning May 2001 ‐ concluding date not reported Language: English
Index tests	Index test: aromatase mRNA and 17HSD2 mRNA Description of positive case definition by index test as reported: SYBR Green I kit (Roche, Germany); target concentration was normalised to a housekeeping gene GAPDH; melting curve was used to verify specificity of reaction); laboratory technique described; threshold not pre‐specified Examiners: none stated; unclear if blinded to the result of reference standard Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: DIE Prevalence of target condition in the sample: n/N = 30/54 (55%): stages not specified; controls 24 Reference standard: laparoscopy Description of positive case definition by reference test as reported: DIE lesion deeper than 5 mm from peritoneal surface Examiners: none stated
Flow and timing	Time interval between index test and reference standard: samples were collected at surgery Withdrawals: 1 patient was not included in analysis of 17HSD2 mRNA
Comparative
Notes	Conclusion: Local oestrogen concentration may be much higher in epithelial cells than in stromal cells in deep endometriotic tissue. Comments: For 17HSD2 mRNA the diagnostic estimates are reported only for stromal cells; there was no statistically significant difference between the group for glandular epithelium For 17HSD2 mRNA in stromal cells there was no statistically significant difference between the groups in proliferative phase, but separate accuracy estimates for secretory phase were not available ‐ pooled data presented For aromatase the diagnostic estimates are presented for pooled cycle phases and pooled cell types
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Meibody 2011.

Study characteristics
Patient sampling	Primary objective: to test the hypothesis that small demyelinated nerve fibres in functional layer of endometrium are present in higher density in women with endometriosis, which is more significant in secretory phase. Participants: women undergoing laparoscopy/laparotomy for infertility or pelvic pain Selection criteria: inclusion criteria: reproductive age, regular menstrual cycle; exclusion criteria: unwillingness to participate and use of hormonal medications for the past 3/12 months Study design: observational single‐gate, prospective collection of samples
Patient characteristics and setting	Clinical presentation: chronic pelvic pain ‐ 23/27, dyspareunia ‐ 5/27, dysmenorrhoea ‐ 7/27, infertility ‐ 5/27 Age: mean age 39.5 ± 5.9 years, endometriosis; 41.6 ± 5.7 years, controls Number enrolled: 27 women Number available for analysis: 27 women (all in proliferative cycle phase) Setting: university hospital ‐ Minimally Invasive Surgery Research Center, Rassoul Akram Hospital, Iran University of Medical Sciences Place of study: Tehran, Iran Period of study: 2007‐2009 Language: English
Index tests	Index test: endometrial small nerve fibres in eutopic endometrium ‐ PGP 9.5 Description of positive case definition by index test as reported: Presence of nerve fibres detected by IHC staining for PGP 9.5 seen in 10 HPF (IHC by using Dako Denmark A/S Produktionsej42 DK‐2600, Denmark and Olympus microscope; assessment of 3‐4 sections per slide; density of NF was also calculated by intensity of staining); laboratory technique described Examiners: pathologist was blinded to reference standard result Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition: n/N = 12/27 (44%): stage not reported; controls 15 Reference standard: Laparoscopy/laparotomy + histology Description of positive case definition by index test as reported: visualisation of the endometriotic lesions with surgical staging of the disease and histological confirmation Examiners: 3 experienced gynaecologists in endometriosis
Flow and timing	Time interval between index test and reference standard: endometrial biopsy was obtained prior to surgery Withdrawals: none reported
Comparative
Notes	Conclusion: Assessment of neural fibre density in endometrial biopsy by PGP 9.5 staining is as accurate as surgical assessment by experienced gynaecologists.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	Unclear
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Low	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Meola 2009.

Study characteristics
Patient sampling	Primary objective: to analyse the expression of the glycodelin gene to better understand the molecular environment of endometriotic lesions and to elucidate the potential mechanisms that underlie the complex physiopathology of endometriosis Participants: women undergoing laparoscopy for pelvic pain, infertility or sterilisation Selection criteria: inclusion criteria: reproductive age (18‐40 years), early proliferative phase of the menstrual cycle (days 5‐8), regular cycles, and no history of any hormone therapy during the 6/12 months prior to surgery Study design: observational two‐gate, prospective collection of samples
Patient characteristics and setting	Clinical presentation: endometriosis: pelvic pain, infertility or both; controls referred for tubal ligation Age: range 18‐40 Number enrolled: 28 women Number available for analysis: 28 women (all in early proliferative phase) Setting: tertiary referral centre, School of Medicine of Ribeirao Preto, University of Sao Paulo Place of study: Ribeirao Preto, Brazil Period of study: not provided Language: English
Index tests	Index test: glycodelin mRNA Description of positive case definition by index test as reported: mRNA expression by RT‐PCR (TaqMan methodology, target level normalised to endogenous control (GAPDH) and the calibrator sample, calculated by 2‐ΔΔCt method); laboratory technique described; threshold not pre‐specified Examiners: none stated Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 17/28 (61%): stage II 4, stage III‐IV 13; controls 11 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: staging according to rASRM classification, diagnosis was confirmed through histopathologic analysis Examiners: none stated
Flow and timing	Time interval between index test and reference standard: before procedure Withdrawals: none reported
Comparative
Notes	Conclusion: Glycodelin may be one of the molecules that contributes to the loss of cellular homeostasis in endometriotic lesions.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Yes
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	No
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Meola 2013.

Study characteristics
Patient sampling	Primnary objective: to investigate whether average CALD1 expression and caldesmon protein levels were altered in the endometrium and endometriotic lesions and to evaluate the performance of CALD1 gene and caldesmon protein as potential biomarker for endometriosis Participants: women who underwent laparoscopy for suspected endometriosis and for tubal ligation Selection criteria: inclusion criteria: reproductive age, no hormone therapy for at least 6/12 months and no history of reproductive tract disorders or tumours Study design: observational two‐gate, prospective collection of the samples
Patient characteristics and setting	Clinical presentation: endometriosis group: infertility ‐ 21/40, pelvic pain ‐ 19/40; controls ‐ healthy fertile women requesting tubal ligation Age: mean age 33.5 ± 0.72 years (endometriosis), 33.87 ± 0.69 years (controls) Number enrolled: 55 women Number available for analysis: 55 women (35 in proliferative and 35 in secretory phase of menstrual cycle; controls samples × 2 in each cycle phase) Setting: tertiary hospital of the School of Medicine of Ribeirao Preto (University of Sao Paulo) Place of study: Ribeirao Preto, Brazil Period of study: not stated Language: English
Index tests	Index test: CALD1 mRNA and caldesmon protein Description of positive case definition by index test as reported: relative expression of CALD 1 gene detected by RT‐PCR (Taqman methodology, relative quantification by the 2‐ΔΔCt method, calibrated and normalised to GAPDH and ACTB genes); caldesmon protein level by IHC stain intensity; laboratory techniques described; thresholds not pre‐specified Examiners: examiners blinded to the clinical data Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition: n/N = 40/55 (72%): stage I‐II 17, stage III‐IV 23; controls 15 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: laparoscopic and histological diagnosis of endometriosis, staging according to rASRM classification Examiners: not stated
Flow and timing	Time interval between reference test and index test: during laparoscopy Withdrawals: none reported
Comparative
Notes	Conclusion: Caldesmon is a possible predictor of endometrial dysregulation in patients with endometriosis.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Mikolajczyk 2006.

Study characteristics
Patient sampling	Primary objective: to test whether there is an altered secretion of IL‐11 and LIF, which could explain receptivity defect in patients with minimal endometriosis Participants: women undergoing laparoscopy for infertility, pelvic pain or both Selection criteria: exclusion criteria: any abnormality in infertility workup, hormonal medications for 3/12 months preceding surgery, advanced endometriosis diagnosed at the index surgery Study design: observational singe‐gate, prospective collection of samples
Patient characteristics and setting	Clinical presentation: endometriosis group ‐ Infertility; controls ‐ pelvic pain, fertile women Age: mean age 26 years (range 22‐34), endometriosis; 24 years (range 22‐34), controls Number enrolled: 84 women Number available for analysis: 35 women (all in luteal phase) Setting: Division of Reproduction, University of Medical Sciences Place of study: Poznan, Poland Period of study: not provided Language: English
Index tests	Index test: LIF and IL‐11 protein in uterine flushing; LIF and IL‐11 mRNA in endometrium Description of positive case definition by index test as reported: concentration of IL‐11 and LIF in uterine flushing detected by ELISA; mRNA expression levels in endometrium detected by RT‐PCR (Qiagen OneStep RT–PCR kit with amplification, normalised to GAPDH); experiments described; thresholds not pre‐specified Examiners: no information provided, unclear if blinded to the result of index test Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 14/35 (40%): all stage I‐II; controls 21 Reference standard: laparoscopy ± histology Description of positive case definition by reference test as reported: diagnosis of endometriosis was based on visualisation of endometriotic lesions found during laparoscopy; in 70% of patients, the diagnosis was also confirmed by histopathologist; staging according to rAFS Examiners: All laparoscopies were performed by a skilled surgeon with over 15 years of experience in detection and treatment of endometriosis
Flow and timing	Time interval between index test and reference standard: uterine flushing and tissue biopsy were performed at surgery Withdrawals: 48 women were excluded because did not meet inclusion criteria ‐ before sample processing
Comparative
Notes	Conclusion: There is no receptivity defect with regard to LIF and IL‐11 secretions by eutopic endometrium in infertile women with endometriosis. Comment: For LIF and IL‐11 in uterine flushing and in endometrium there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	No
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Morelli 2010.

Study characteristics
Patient sampling	Primary objective: to determine whether relaxin and or its LGR7 receptor are expressed In human endometriotic tissues and whether their expression differs from that in endometrium of normal controls Participants: women undergoing surgery for findings of an ovarian cyst on USS Selection criteria: exclusion criteria: laparoscopic evidence and histologic confirmation of extra ovarian endometriosis at index surgery Study design: observational single‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: ovarian cyst, not specified Age: range 24‐37 years Number enrolled: 40 women Number available for analysis: 40 women (11 in proliferative and 29 in secretory phase of the cycle) Setting: Department of Paediatrics, Obstetrics and Reproductive Medicine, University of Siena Place of study: Siena, Italy Period of study: October 2006 to May 2007 Language: English
Index tests	Index test: Relaxin mRNA, LGR7 receptor mRNA Description of positive case definition by index test as reported: LGR7 ‐ mRNA expression detected by RT‐PCR; Relaxin ‐ proportion of tissues expressing relaxin (detection by using the Rotor‐Gene 3000 PCR system (Corbett Research, CA) with using Platinum SYBR Green qPCR SuperMix (Invitrogen); normalised to β‐actin mRNA expression via 2‐ΔΔCt method; experiments described; thresholds not pre‐specified Examiners: none stated; unclear if blinded to the results of reference standard Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: ovarian endometriosis Prevalence of target condition in the sample: n/N = 21/40 (52%): all stage III‐IV; controls 19 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: laparoscopic evidence and histologic confirmation of extra ovarian endometriosis; staging according to the rASRM classification Examiners: none stated
Flow and timing	Time interval between index test and reference standard: samples were collected at surgery (appears from the context) Withdrawals: 8 patients with endometriosis were excluded as no tissue was available
Comparative
Notes	Conclusion: Significantly decreased expression of relaxin mRNA and LGR7 relaxin receptor mRNA in ectopic endometriotic tissues relative to their expression in eutopic endometrium and in endometrium from normal controls suggests that relaxin exerts a protective effect against endometriosis. Comment: For LGR7 relaxin receptor and relaxin mRNA there was no statistically significant difference between the groups; no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	No
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Unclear
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Pino 2009.

Study characteristics
Patient sampling	Primary objective: to investigate the association between MMP‐1 and MMP‐9 activities and ICAM1 cleavage mediated by TNF in eutopic endometrial stromal cells from women with and without endometriosis during culture Participants: women without endometriosis who underwent laparoscopic tubal sterilisation or hysterectomy for a benign non‐endometrial gynaecologic condition (control group), and women undergoing laparoscopy for diagnosis of endometriosis, which was later surgically confirmed Selection criteria: exclusion criteria: hormonal treatment or contraceptives during the previous 6 or 3 months, respectively; neoplastic, endocrine or infectious diseases Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis group ‐ not specified, controls ‐ eumenorrhoeic reproductively normal women without endometriosis who underwent laparoscopic tubal sterilisation Age: mean age 35.2 ± 0.7, endometriosis group; 38.4 ± 0.5 years, controls Number enrolled: 20 women Number available for analysis: 20 women (7 in proliferative and 13 in secretory phase of the cycle) Setting: Institute of Maternal and Child Research, San Borja Arriarán Clinical Hospital, University of Chile Place of study: Santiago, Chile Period of study: not reported Language: English
Index tests	Index test: MMP‐1, MMP‐9, ICAM1, sICAM1 mRNA and protein Description of positive case definition by index test as reported: MMP‐1, MMP‐9, ICAM1 and sICAM1 mRNA expression by RT‐PCR (internal control 18S rRNA, semi quantification by using agarose gel electrophoresis); ICAM1 and MMP1 protein levels by Western blot; activities of MMP1 and MMP9 by using casein zymography in cell culture (results are expressed as arbitrary units regarding the total µg of proteins loaded (MMP1) or as values normalised to the internal standard (MMP9)); concentration of active MMP‐1 by using enzymatic activity assay (Human Active MMP‐1 Fluorescent Assay kit (R&D System Inc., Minneapolis, USA) per manufacturer’s instructions; detection limit 0.052 ng/ml); sICAM concentration by using ELISA (Inmunoassay Human sICAM1 kit (R&D System) per manufacturer’s instructions, minimal detectable concentration of 0.096 ng/mL); sample processing and laboratory technique described in details; thresholds not provided Examiners: none stated; unclear if blinded to the results of reference standard Interobserver variability: intra‐ and interassay reproducibility for sICAM: 3.3%‐4.8% and 6.0%‐10.1%, for MMP‐1: 9.6%‐10.0% and 8.7%‐17.7%
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 10/20 (50%): stage I‐II 6, stage III‐IV 4; controls 10 Reference standard: laparoscopy/laparotomy Description of positive case definition by reference test as reported: staging according to the rASRM classification Examiners: none stated
Flow and timing	Time interval between index test and reference standard: samples were collected at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: The deregulation of MMP‐9, and the TNF participation in the MMP‐1 and proMMP‐9 secretions, in the MMP‐9 expression and in the expression and cleavage of ICAM1 may contribute to the pathophysiology of this disease Comment: The reported estimates for the assessed biomarkers are for basal conditions only, prior to incubation with TNF For ICAM mRNA, ICAM protein, sICAM protein, MMP‐9 mRNA, ‐ protein and MMP‐1 protein there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7) For MMP‐1 mRNA there was statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Prefumo 2002.

Study characteristics
Patient sampling	Primary objective: to quantify the expression of CD54, CD58 and CD106, three adhesion molecules with a crucial role in cytotoxic mechanisms, on fresh endometrial cells by immunofluorescence and flow cytometry in women with and without endometriosis Participants: regularly cycling patients undergoing laparoscopy because of pelvic pain, infertility or both Selection criteria: inclusion criteria: no hormonal treatment in 3 months before surgery Study design: observational single‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: pelvic pain, infertility or both Age: median age 29.5 years (range 20–44), endometriosis; 34 years (range 26‐46), controls Number enrolled: 22 women Number available for analysis: 22 women (11 in proliferative and 11 in secretory phase of menstrual cycle) Setting: university hospital ‐ UO di Ostetricia e Ginecologia, Istituto Giannina Gaslini, Università di Genova Place of study: Genoa, Italy Period of study: not reported Language: English
Index tests	Index test: adhesion molecules CD54, CD58, CD106 Description of positive case definition by index test as reported: cell size and number of molecules expressed on the cell surface by flow cytometric methods (analyses by using a flow cytometer (Coulter EPICS XL, Milan, Italy) on cells gated to exclude nonviable cells; a minimum of 1 × 104 cells were examined per sample; quantitative analysis of the expression of surface molecules on the endometrial cells were calculated using different standard beads of known diameter and number of FITC equivalent (Coulter Science, Miami, USA); experiments described; thresholds not provided Examiners: no information provided; unclear if blinded to the result of reference standard Interobserver variability: not stated
Target condition and reference standard(s)	Target condition: peritoneal endometriosis Prevalence of target condition in the sample: n/N = 10/22 (45%): stages not reported; controls 12 Reference standard: laparoscopy Description of positive case definition by reference test as reported: peritoneal endometriosis was defined as the presence of red, black or white lesions on the peritoneal surface Examiners: none stated
Flow and timing	Time interval between index test and reference standard: endometrial samples were obtained at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: These findings could account for an apparently cyclic defect in the expression of CD54 that could result in poor binding of immune effectors to secretory endometrial cells in vivo. Comments: For CD58, CD106 and CD54 follicular there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7) For CD54 luteal there was a statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	Unclear
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Puy 2002.

Study characteristics
Patient sampling	Primary objective: To compare the glandular and stromal immunoreactivity patterns of αvβ5 and αvβ6 integrins between eutopic and ectopic endometrium from women with endometriosis and in eutopic endometrium of women without endometriosis; to analyse vascular αvβ5 and αvβ6 immunoreactivity in these tissue specimens Participants: patients who underwent laparoscopy at the authors' institution, including tubal ligation Selection criteria: exclusion criteria: known prior endometrial disease or patients receiving hormonal therapy Study design: observational two‐gate, prospective recruitment and sample collection
Patient characteristics and setting	Clinical presentation: endometriosis group: not reported; controls ‐ women requesting tubal ligation Age: reproductive age group; not specified Number enrolled: 57 women Number available for analysis: 52 women (22 in proliferative and 30 in secretory cycle phase) Setting: university‐affiliated tertiary care centre ‐ Women’s College Health Sciences Centre Place of study: Toronto, Canada Period of study: not reported Language: English
Index tests	Index test: epithelial, stromal and vascular αvβ5 and αvβ6 integrins Description of positive case definition by index test as reported: staining intensity of endometrial epithelial and nonepithelial cells by IHC using HSCORE (0 ‐ no staining, 1 ‐ weak, 2 ‐ moderate, 3 ‐ strong staining) and quantification of the percentage of immunolabelled blood vessels (evaluation by using a computerised image analysis system (CIAS): endometrial cells ‐ mean of 5 fields at × 100 magnification; blood vessel cells ‐ mean of 10 fields at × 1000 magnification; percentage of immunolabelled blood vessels was evaluated as mean of 10 fields at × 400 magnification); technique described; thresholds not provided Examiners: All samples were evaluated independently by the two observers, and each was blinded to all clinical data Interobserver variability: The overall correlation coefficient for intraobserver median HSCOREs was 0.97 (P < 0.0001) and for interobserver variation was 0.96 (P < 0.0001).
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 40/52 (77%): stage I‐II 14, stage III‐IV 26; controls 12 Reference standard: laparoscopy Description of positive case definition by reference test as reported: staging according to rAFS classification Examiners: none stated
Flow and timing	Time interval between index test and reference standard: endometrial samples were obtained at surgery Withdrawals explained: 5 endo cases excluded due to asynchronous histological dating (excluded before the experiments)
Comparative
Notes	Conclusion: The differences that we observed for αvβ5 and αvβ6 integrin immunostaining between eutopic and ectopic endometrium remain to be elucidated. Comments: For epithelial and stromal αvβ5 and αvβ6, there was no statistically significant difference between the groups; no data available for meta‐analysis (Appendix 7) For αvβ5 and αvβ6 integrins in blood vessel walls there was statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Unclear
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Rakhila 2013.

Study characteristics
Patient sampling	Primary objective: to investigate prostaglandin (PG) biosynthesis and catabolism pathways in eutopic and ectopic endometrium of women with endometriosis Participants: women undergoing laparoscopy for pelvic pain, infertility or both (endometriosis group) and women scheduled for tubal ligation Selection criteria: exclusion criteria: no other pelvic pathology, signs of endometrial hyperplasia, or neoplasia and any nonsteroidal anti‐inflammatory drugs (NSAIDs) or hormonal medication for 3 months before surgery Study design: observational two‐gate, prospective recruitment and sample collection
Patient characteristics and setting	Clinical presentation: endometriosis group ‐ chronic pelvic pain, infertility or both; controls – disease‐free women requesting tubal ligation Age: Mean age 34.2 ± 3.6 yeas (endometriosis group), 35.3 ± 3.8 years (controls) Number enrolled: 74 women Number available for analysis: 74 women (34 in proliferative and 40 in secretory cycle phase) Setting: Centre Hospitalier Universitaire de Québec, Hôpital Saint‐François d'Assise Place of study: Quebec, Canada Period of study: not specified Language: English
Index tests	Index test: prostaglandin synthases: Cox‐1, Cox‐2, mPGES‐1, mPGES‐2, cPGES, AKR‐1B1, AKR‐1C3; prostaglandin catabolic enzyme 15‐PGDH mRNA Description of positive case definition by index test as reported: expression levels of the biomarkers by RT‐PCR (using SYBR Green chemistry, carried out in ABI 7000 Thermal Cycler (Applied Biosystems); quantification by using a relative quantification method, normalised to GAPDH mRNA levels); sample handling and laboratory technique described Examiners: no information provided Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 45/74 (61%); stage I‐IV, number per subgroup not provided; controls 29 Reference standard: laparoscopy Description of positive case definition by reference test as reported: staging according to the rASRM system Examiners: no information provided
Flow and timing	Time interval between index test and reference standard: tissue samples were obtained at laparoscopy Withdrawals: none reported
Comparative
Notes	Conclusion: This study reveals for the first time multiple defects in PG biosynthesis pathways, which differ between eutopic intrauterine and ectopic endometrial tissues and may, owing to the wide spectrum of PG properties, contribute to the initial steps of endometrial tissue growth and development and have an important role to play in the pathogenesis and symptoms of this disease Comments: For Cox‐1, cPGES, Akr1B1, Akr1C3 and 15‐PGDH mRNA there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7) For Cox‐2, mPGES1 and mPGES2 mRNA there was a statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Ramon 2011.

Study characteristics
Patient sampling	Primary objective: to analyse several miRNAs related to angiogenesis and the angiogenic factors, VEGF‐A and TSP‐1, in endometriotic lesions and eutopic endometrium from women with endometriosis. Participants: women undergoing surgery for pelvic pain, sterilisation, infertility or endometriosis Selection criteria: exclusion criteria: menstrual phase of the cycle, irregular menstruation, pregnancy or breastfeeding in the previous 6/12 months, hormonal treatment for the last 3/12 months Study design: observational two‐gate, prospective sample selection
Patient characteristics and setting	Clinical presentation: indications for surgery ‐ endometriosis group: clinical or ultrasonographic suspicion of endometriosis; controls: pelvic pain ‐ 7.90%, sterility ‐ 21.0%, tubal sterilisation ‐ 71.1% Age: mean age 34.9 years (range 24‐47), endometriosis; 36.3 years (range 21‐47), controls Number enrolled: 96 women Number available for analysis: 96 women (46 in proliferative and 50 in secretory cycle phase) Setting: university hospital ‐ Hospital Universitario La Fe and Hospital Arnau de Vilanova Place of study: Valencia, Spain Period of study: not reported Language: English
Index tests	Index test: VEGF‐A and TSP‐1 mRNA and protein Description of positive case definition by index test as reported: relative mRNA expression by RT‐PCR (in a LightCycler apparatus, using version 3.5 software (Roche Molecular Biochemicals, Mannheim, Germany); normalised to β‐actin; protein level as quantified by ELISA; experiments described; no threshold provided Examiners: no information provided; unclear if blinded to the result of reference standard Interobserver variability: ELISA intra‐ and interassay variabilities for VEGF: 4–6% and 7–10%; for TSP‐1: 5–6% and 8–11%
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 58/96 (60%): stage not specified; controls 38 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: meticulous examination of the peritoneum, ovaries, intestine and diaphragm to detect any typical or atypical endometriotic lesion; biopsies of suspicious areas; complete excision of the endometriotic tissue Examiners: none stated
Flow and timing	Time interval between index test and reference standard: endometrial samples were obtained at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: expression levels of mRNAs related to angiogenesis were different in eutopic endometrium from that observed in ovarian endometrioma. This could influence the expression of angiogenic factors and play a role in the pathogenesis of endometriosis. Comments: For VEGF A mRNA and protein there was statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review For TSP‐1 mRNA and protein there was no statistically significant difference between the groups; no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Szymanowski 2003.

Study characteristics
Patient sampling	Primary objective: to describe the expression of α3β1, α4β1 and β1 integrin chain in endometrial biopsy in women suffering from endometriosis Participants: women with infertility, pelvic pain or both undergoing laparoscopy and hysteroscopy Selection criteria: inclusion criteria: regular cycles, no hormonal treatment at least 1/12 month prior surgery; exclusion criteria: pregnancy, non‐peritoneal endometriosis Study design: observational single‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: infertility ‐ 23/32, pelvic pain ‐ 11/32 Age: mean age 29.2 ± 4.19 years Number enrolled: 32 women Number available for analysis: 32 women (all in secretory phase) Setting: university hospital ‐ Division of Reproduction, K. Marcinkowski University of Medical Sciences Place of study: Poznan, Poland Period of study: January 1998 to June 2000 Language: English
Index tests	Index test: α3β1, α4β1 and β1 Integrin chain Description of positive case definition by index test as reported: Positive IHC staining (staining intensity was defined as weak +, moderate ++, strong +++; not defined in glandular and stromal cells); threshold not pre‐specified Examiners: blinded observer Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: peritoneal endometriosis Prevalence of target condition in the sample: n/N = 17/32 (53%): stage not specified; controls 15 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: direct visualisation of endometrial foci with histological confirmation; rAFS classification Examiners: not stated
Flow and timing	Time interval between reference test and index test: tissue sample was obtained at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: The selected integrins do not yet allow identification of a pattern characteristic for endometriosis.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	Unclear
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Takahashi 1990.

Study characteristics
Patient sampling	Primary objective: to discover clinical usefulness of levels of CA‐125 in menstrual blood in patients with endometriosis Participants: women undergoing laparoscopy or laparotomy for infertility or pelvic pain Selection criteria: not stated Study design: observational single‐gate, prospective recruitment and sample collection; consecutive patients
Patient characteristics and setting	Clinical presentation: infertility/pelvic pain Age: reproductive‐aged women Number enrolled: 104 women Number available for analysis: 104 women (all in menstrual phase of the cycle) Setting: university hospital ‐ Department of O&G, Shimane Medical University Place of study: Izumo, Japan Period of study: not stated Language: English
Index tests	Index test: CA‐125 in menstrual fluid Description of positive case definition by index test as reported: CA‐125 level by RIA (commercial kit Centocor Inc, Malvern, PA); sample handling described; threshold > 100,000 U/ml ‐ not pre‐specified Examiners: not stated; unclear if blinded to reference standard Interobserver variability: not provided
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in sample: n/N = 38/104 (36%): stage I‐II 20, stage III‐IV 14; controls 66: normal pelvis 30, other pelvic pathology 36 Reference standard: laparoscopy/laparotomy Description of positive case definition by reference test as reported: visual typical appearance of disease, staged according to rAFS classification Examiners: not stated
Flow and timing	Time interval between reference test and index test: sample was collected pre‐operatively with interval of 1‐4 weeks Withdrawals: none reported
Comparative
Notes	Conclusion: Obtaining CA‐125 concentration in menstrual discharge was found to be more sensitive than serum CA‐125 levels drawn during menses in endometriosis.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Yes
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		Low	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Unclear
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Tang 2009.

Study characteristics
Patient sampling	Primary objective: to investigate the possible roles of cyclin B1, cyclin‐dependent kinase (cdc2) and Polo‐like kinase 1 (Plk1) in the pathogenesis of endometriosis Participants: women who were undergoing laparoscopic surgery or hysterectomy Selection criteria: inclusion criteria: regular menstrual cycles (21–35 days), not on hormonal treatment, not pregnant, not breastfeeding, and not using IUD in the past 6/12 months; exclusion criteria: PID, adenomyosis, DUB Study design: observational study, two‐gate design, prospective sample collection
Patient characteristics and setting	Clinical presentation: not specified Age: range 24‐46 years Number enrolled: 59 women Number available for analysis: 50 women (27 in proliferative and 23 in secretory cycle phase) Setting: university hospital ‐ Women’s Hospital of Zhejiang University Medical School Place of study: Hangzhou, Zhejiang, China Period of study: 2005‐2006 Language: English
Index tests	Index test: cyclin B1, cdc2, Plk1 mRNA and protein Description of positive case definition by index test as reported: mRNA expression by RT‐PCR (using multiscribe reverse transcriptase (Invitrogen), single tube reaction with amplification; β‐actin gene used as internal reference); protein levels detected by Western Blot (Bradford assay (Bio‐Rad Laboratories, CA), normalised to β‐actin); laboratory technique described; thresholds not provided Examiners: none stated; unclear if blinded to the result of reference standard Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 20/50 (40%): all stages II‐III; controls 30 Reference standard: laparoscopy/laparotomy Description of positive case definition by reference test as reported: staging according to rAFS Examiners: none stated
Flow and timing	Time interval between index test and reference standard: not stated; the context suggests timing around surgery Withdrawals: 9 participants were not included in the analyses, reasons not explained
Comparative
Notes	Conclusion: the increased expression of cyclin B1, cdc2, and Plk1 in ectopic endometrium may be involved in the up‐regulated proliferation activity and neoplastic features of endometriosis. Comment: For Cyclin B1, cdc2, Plk1 mRNA and protein, there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Unclear
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Tiberi 2010.

Study characteristics
Patient sampling	Primary objective: to examine prokineticin 1 (PROK1) mRNA expression in eutopic endometrial glands obtained from patients with or without endometriosis Participants: women of reproductive age undergoing laparoscopy for subfertility or suspected endometriosis Selection criteria: inclusion criteria: regular menstrual cycles, written consent; exclusion criteria: no hormonal medications for the 3 months preceding surgery, endometrium with pathology or hormonal alterations Study design: observational single‐gate, prospective enrolment and sample collection
Patient characteristics and setting	Clinical presentation: infertility (blocked tubes n = 10, unexplained infertility n = 1, without previous pregnancy n = 1) or suspected endometriosis (CPP n = 2, suspected endometriomas n = 18) Age: range 26‐40 years Number enrolled: 32 women Number available for analysis: 24 women (all in secretory cycle phase) Setting: university hospital ‐ Universita Cattolica del Sacro Cuore Place of study: Rome, Italy Period of study: not provided Language: English
Index tests	Index test: PROK1 mRNA in glandular epithelium Description of positive case definition by index test as reported: Positive or negative mRNA expression by RT‐PCR; negative mRNA expression was defined as no amplification detected after 40 cycles (TaqMan methodology, expression was evaluated using the i‐Cycler iQTM system (Bio‐Rad Laboratories); normalised to GAPDH gene and calibrator sample by the 2−ΔΔCt method); sample handling and laboratory technique described Examiners: none stated; unclear if blinded to the result of reference standard Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 12/24 (50%): all stage III‐IV; controls 12 Reference standard: laparoscopy Description of positive case definition by reference test as reported: staging according to rASRM classification Examiners: none stated
Flow and timing	Time interval between index test and reference standard: "All samples were collected by uterine curettage specifically during the putative window of implantation . . . from women undergoing laparoscopy" suggests that the samples were collected before surgery and the time interval sample collection‐surgery was reasonably short Withdrawals: 8 women from control group were excluded prior to analysis because of other ovarian pathology
Comparative
Notes	Conclusion: PROK1 is a newly discovered angiogenic factor implicated in the vascular function of peri‐implantation endometrium and early pregnancy. An altered expression of PROK1 could be one of the several biochemical abnormalities characterising eutopic endometrium in endometriosis.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Unclear	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Unclear
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Unclear
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Van der Linden 1994.

Study characteristics
Patient sampling	Primary objective: to detect the expression of integrins and E‐cadherin in cells from peritoneal fluid (PF), endometrium, menstrual effluent, peritoneum, and endometriotic lesions during the early follicular phase of the menstrual cycle Participants: women who underwent a diagnostic laparoscopy as part of a subfertility work‐up Selection criteria: inclusion criteria: regular ovulatory cycles Study design: observational single‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: infertiliy Age: reproductive age, not specified Number enrolled: 16 women Number available for analysis: 16 women (all in early proliferative cycle phase) Setting: tertiary care university medical centre ‐ the University of Limburg Place of study: Maastricht, the Netherlands Period of study: not reported Language: English
Index tests	Index test: Integrins α2β1, α3β1, α4β1, α5β1, α6β1, E‐cadherin in menstrual fluid and endometrium Description of positive case definition by index test as reported: positive IHC staining; laboratory technique described Examiners: none stated; unclear if blinded to the result of reference standard Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 8/16 (50%): all stage I; controls 8 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: visual inspection confirmed by histopathology, staging according to rAFS Examiners: none stated
Flow and timing	Time interval between index test and reference standard: samples were collected at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: Integrins α2β1, α3β1, α4β1, α5β1, α6β1 and E‐cadherin, important cell adhesion molecules, are expressed in endometriotic lesions and in cells and tissues that are potentially involved in the development of endometriosis. These cell adhesion molecules could be involved in the shedding of endometrial tissue during menstruation and the attachment of endometrial tissue fragments to the peritoneum Comment: For α2β1, α3β1, α4β1, α5β1, α6β1 and E‐cadherin there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Unclear
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Unclear	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Van der Linden 1995.

Study characteristics
Patient sampling	Primary objective: to examine the immunohistochemical properties of epithelial cells in peritoneal fluid and to compare the staining characteristics with cells of endometrium, menstrual effluent, peritoneum, and endometriotic lesions Participants: women who underwent a diagnostic laparoscopy as part of a subfertility work‐up Selection criteria: inclusion criteria: regular ovulatory cycles Study design: observational single‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: infertility Age: reproductive age, not specified Number enrolled: 16 women Number available for analysis: 16 women (all in early proliferative cycle phase) Setting: tertiary care university medical centre ‐ the University of Limburg Place of study: Maastricht, the Netherlands Period of study: not reported Language: English
Index tests	Index test: vimentin, cytokeratin 18, cytokeratin 19 and endometrial epithelial marker BW 495/36 in menstrual fluid and endometrium Description of positive case definition by index test as reported: positive IHC staining; laboratory technique described Examiners: none stated; unclear if blinded to the result of reference standard Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 8/16 (50%): all stage I; controls 8 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: visual inspection confirmed by histopathology, staging according to rAFS Examiners: none stated
Flow and timing	Time interval between index test and reference standard: samples were collected at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: These results support the contention of transport of menstrual detritus to the peritoneal cavity in women with patent fallopian tubes. Comment: For vimentin, cytokeratin 18, cytokeratin 19 and endometrial epithelial marker BW 495/36, there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Unclear
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Unclear	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Vernet‐Tomas 2006a.

Study characteristics
Patient sampling	Primary objective: to compare the expression and localisation of integrin subunit α6 in women with and without endometriosis Participants: women who underwent surgery for various indications (pelvic pain, infertility, menstrual disturbances, adnexal masses or tubal ligation) Selection criteria: exclusion criteria: history of immune disease, malignancy, hormone therapy, or immune therapy Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis: pelvic pain ‐ 12/30, infertility ‐ 8/30, adnexal mass ‐ 6/30, menstrual disturbances ‐ 4/30; controls: pelvic pain ‐ 5/19, infertility ‐ 3/19, tubal sterilisation ‐ 11/19 Age: mean age 32.8 ± 7.11 years (range 20‐50), endometriosis; 33.95 ± 3.36 years (range 27‐41), controls Number enrolled: 52 women Number available for analysis: 49 women (19 in proliferative and 30 in secretory cycle phase) Setting: university hospital ‐ O&G Department Hospital Universitari del Mar and Hospital Gemans Trias i Pujol Place of study: Barcelona, Spain Period of study: not reported Language: English
Index tests	Index test: depolarised α6 integrin Description of positive case definition by index test as reported: the percentage of positively stained glandular cells and the localisation of expression: polarised α6 ‐ expression exhibited only on the basal side of the cell, depolarised ‐ when expression was observed in the basal side, as well as in any other side of the cell (IHC staining; examining ten non‐overlapping fields per biopsy with a magnification of × 400, counting a total of 300‐400 cells per case); technique described Examiners: 2 independent observers who were blinded to the study group Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 30/52 (58%); stage I‐II 4, stage III‐IV 26; controls 19 Reference standard: laparoscopy n = 32/laparotomy n = 17 Description of positive case definition by reference test as reported: rAFS classification Examiners: none stated
Flow and timing	Time interval between index test and reference standard: endometrial samples were collected at surgery Withdrawals: 3 women were excluded due to ineligible tissue samples
Comparative
Notes	Conclusion: The endometria of women with endometriosis more frequently show a depolarised expression of integrin subunit α6, a characteristic usually found in highly proliferating cells with migrating and invasive abilities.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	Yes
		Low	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Visnovsky 2008.

Study characteristics
Patient sampling	Primary objective: to determine clinical benefits of mRNA aromatase expression in entopic endometrium as a diagnostic marker of endometriosis Participants: women undergoing laparoscopy or laparotomy for pelvic pain, infertility of benign pelvic tumours Selection criteria: inclusion criteria: confirmed benign pelvic condition by pre‐operative ultrasound Study design: observational single‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis: pelvic pain, infertility, pelvic mass Age: mean age 39.7; SD and range not reported Number enrolled: 23 women Number available for analysis: 23 women (menstrual cycle phase not specified) Setting: university hospital ‐ O&G Department Jessenius Medical Faculty and Martin University Hospital Place of study: Martin, Czech Republic Period of study: May 2006 to August 2007 Language: English
Index tests	Index test: aromatase mRNA Description of positive case definition by index test as reported: mRNA expression by RT‐PCR, semi‐quantitative expression scored 0‐3: 0 ‐ no expression; 1 ‐ low; 2 ‐ medium; 3 ‐ high expression; technique not described; threshold ‐ low expression (score 1 and above, not pre‐specified) Examiners: not reported; unclear if were blinded to the results of reference standard Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 12/23 (52%); stage not specified; controls 11 Reference standard: laparoscopy n = 17/laparotomy n = 6 + histology Description of positive case definition by reference test as reported: visual inspection confirmed by histology, rAFS staging Examiners: none stated
Flow and timing	Time interval between index test and reference standard: endometrial samples were collected at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: Aromatase expression in eutopic endometrium is a good diagnostic marker for endometriosis.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Unclear
Was a 'two‐gate' design avoided?	Yes
		High	Unclear
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Wang 2010a.

Study characteristics
Patient sampling	Primary objective: to find the biomarkers of eutopic endometrium in endometriosis patients by using SELDI‐TOF‐MS protein chip array technology Participants: women who underwent laparoscopy for suspected endometriosis (cases) and for tubal ligation or hysterectomy for benign reasons (controls) Selection criteria: inclusion criteria: normal menstrual cycles, no hormonal treatment for at least 6/12 months before operation Study design: observational two‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: endometriosis group: pelvic pain ‐ 2/13, infertility ‐ 4/13, dysmenorrhoea ‐ 3/13, ovarian cyst ‐ 4/13; controls: women requesting tubal ligation, symptoms not specified Age: median age 34 years (range 20‐45), endometriosis; 36 years (range 21‐48), controls Number enrolled: 26 women Number available for analysis: 26 women Setting: university hospital ‐ the 2nd Affiliated Hospital, Zhejiang University School of Medicine Place of study: Hangzhou, China Period of study: not specified Language: English
Index tests	Index test: endometrial proteome by SELDI‐TOF‐MS array (5 protein peaks of 5385 m/z, 5425 m/z, 6898 m/z, 5891 m/z, 6448 m/z) Description of positive case definition by index test as reported: protein peaks significantly different between the groups with higher ROC (detection on the Protein Biological System II (PBSII) and mass spectrometer reader (Ciphergen Biosystems); analysed by using ProteinChip Software 3.1 (Ciphergen): protein mass‐dependent velocities (m/z) peaks were analysed using an artificial neural network); technique described; threshold not pre‐specified Examiners: no information provided; unclear if blinded to the reference standard Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 13/26 (50%): stage I‐II 8, stage III‐IV 5; controls 13 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: visual inspection (careful assessment of the pelvic organs) confirmed by histology; staged according to rAFS classification Examiners: none stated
Flow and timing	Time interval between index test and reference standard: tissue samples were collected at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: 5 potential biomarkers were found, and the diagnostic system separated the endometriosis from the healthy samples with a sensitivity of 91.7%, a specificity of 90.0%. This method showed great potential in screening better biomarkers for endometriosis.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	No
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	No
		High	High
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Wolfler 2005.

Study characteristics
Patient sampling	Primary objective: to evaluate whether endometrial biopsy prior to laparoscopy in symptomatic women to screen for the presence of aromatase by RT‐PCR and IHC combined with select patient characteristics is of value to predict endometriosis Participants: women attending for diagnosis of unexplained infertility, dysmenorrhoea, dyspareunia or chronic pelvic pain scheduled for laparoscopy Selection criteria: exclusion criteria: oestrogen dependent disease other than endometriosis, no endocrine therapy prior to inclusion (GnRH analogues, danazol or oral contraceptives) Study design: observational single‐gate, prospective enrolment and sample collection
Patient characteristics and setting	Clinical presentation: unexplained infertility ‐ 32/48, dysmenorrhoea ‐ 13/48, dyspareunia ‐ 1/48, chronic pelvic pain ‐ 12/48 Age: mean age 32.9 years, range 21‐48 years Number enrolled: 64 women Number available for analysis: 48 women (cycle phase not reported) Setting: tertiary centre ‐ Division of Gynaecological Endocrinology and Reproductive Medicine, Medical University of Vienna Place of study: Vienna, Austria Period of study: not stated Language: English
Index tests	Index test: aromatase mRNA and protein Description of positive case definition by index test as reported: mRNA expression levels by RT‐PCR (Taqman methodology, normalised to internal reference GAPDH gene according to the 2‐ΔΔCt method); protein level ‐ positive staining on IHC (detection by ChemMate Antibody Detection Kit (Dako); microscopical examination of 6 section per slide); laboratory techniques described; threshold pre‐defined for IHC Examiners: IHC ‐ 3 independent trained physicians who were unaware of the patients' history Interobserver variability: not stated
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of reference condition in the sample: n/N = 25/48 (52%): stages I‐IV, numbers not specified; controls 23 Reference standard: laparoscopy + histology Description of positive cased definition by index test as reported: laparoscopic visualisation followed by histopathological assessment; rAFS classification; visual diagnosis that could not be confirmed by histopathology was considered as negative Examiners: not stated
Flow and timing	TIme interval between reference standard and index test: tissue samples were collected prior to laparoscopy Withdrawals: 16 patients were excluded from the analysis: 7 patients ‐ poor quality samples, 2 patients ‐ insufficient RNA extraction, 2 patients ‐ insufficient sample for duplicates, 5 patients ‐ were using COCs
Comparative
Notes	Conclusion: Screening for eutopic endometrial aromatase in combination with clinical data could be of discriminative value in the prediction of disease. Comments: The reported diagnostic estimates of predictive model based on endometrial sample and clinical data are beyond the scope of this review. For aromatase mRNA there was statistically significant difference between the groups, but no data available to construct 2 × 2 tables; not presented in this review.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Unclear
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		Unclear	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	No
		High

Yadav 2013.

Study characteristics
Patient sampling	Primary objective: to evaluate the diagnostic accuracy of detection of nerve fibres in eutopic endometrium and biopsy from peritoneal implants and endometriomas in cases of endometriosis Participants: patients who underwent laparoscopy for infertility/pelvic pain/suspected endometriosis Selection criteria: exclusion criteria: hormonal therapy in the preceding 3/12 months, acute PID, suspected pregnancy, suspected or diagnosed genital malignancy, undiagnosed vaginal bleeding, documented genital tuberculosis, contraindication for laparoscopy or unwillingness to undergo surgery Study design: observational study, single gate, prospective recruitment and sample collection
Patient characteristics and setting	Clinical presentation: infertility ‐ 32/60, CPP ‐ 19/60, infertility + pain symptoms (dysmenorrhoea, dyspareunia, dyschezia) ‐ 9/60; regular menstrual cycle ‐ 57/60 Age: range 15‐45 years Number enrolled: 60 women Number available for analysis: 60 women (cycle phase not specified) Setting: university hospital ‐ O&G Department, University College of Medical Sciences and Guru Teg Bahadur Hospital Place of study: Delhi, India Period of study: November 2009 to April 2012 Language: English
Index tests	Index test: endometrial nerve fibres Description of positive case definition by index test as reported: positive IHC staining for PGP 9.5 identified as single cell positive or linear nerve fibres; technique described Examiners: senior pathologist blinded to patients' data Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of reference condition in the sample: n = 30/60 (50%): stages not reported; controls 30 Reference standard: laparoscopy + histology Description of positive cased definition by index test as reported: laparoscopic visualisation with histological confirmation of suspected lesions; staging according to rAFS classification Examiners: not stated
Flow and timing	TIme interval between reference standard and index test: endometrial biopsy was collected at surgery Withdrawals: none reported
Comparative
Notes	Conclusion: Eutopic endometrium exhibited positivity for nerve fibres in 24 out of 30 cases of proven endometriosis, though the density of nerve fibres was low as compared to that reported in a few studies.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Unclear
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		Unclear	Unclear
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Yes
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Zeng 2005.

Study characteristics
Patient sampling	Primary objective: to evaluate the diagnostic value of evaluating endometrial biopsy specimens for aromatase CYP450 and CA‐125 for endometriosis Participants: women who underwent laparoscopy or laparotomy for pelvic pain, infertility or both Selection criteria: inclusion criteria: reproductive age regular menstrual cycle; exclusion criteria: hormonal treatment for 3/12 months prior reproductive age, preoperative diagnosis of uterine fibroids, adenomyosis Study design: observational study, single‐gate, prospective sample collection
Patient characteristics and setting	Clinical presentation: infertility or pelvic pain Age: Mean age 33 ± 4 years (range 26‐40), endometriosis; 32 ± 4 years (range 25‐39), controls Number enrolled: 58 women Number available for analysis: 58 women (31 women in proliferative and 27 women in secretory phase of the cycle) Setting: Department of Obstetrics and Gynecology, Third Xiangya Hospital, Central South University Place of study: Changsha, China Period of study: March 2003 to February 2004 Language: Chinese
Index tests	Index test: aromatase protein Description of positive case definition by index test as reported: positive IHC staining indicated by presence of brown particles within the cytoplasm; laboratory technique not described Examiners: not stated; unclear if blinded to the result of reference standard Interobserver variability: not stated
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in sample: n/N = 36/58 (62%): stage I‐II 20, stage III‐IV 16; controls 22 Reference standard: laparoscopy/laparotomy Description of positive case definition by reference test as reported: visual inspection; rAFS classification Examiners: not stated
Flow and timing	TIme interval between reference standard and index test: endometrial biopsy was collected intraoperatively Withdrawals: none reported
Comparative
Notes	Conclusion: The combination assay of aromatase cytochrome P450 in eutopic endometrium and CA‐125 can be used as a diagnostic test for endometriosis, especially for the early stage of endometriosis, which is superior to the assay of CA‐125. Comment: The reported diagnostic estimates for combined test, including blood and endometrial sample is beyond the scope of this review.
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Unclear
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		Unclear	Low
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Was a menstrual cycle phase considered in interpreting the index test	No
		High	Unclear
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Unclear
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Unclear	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

Zubor 2009.

Study characteristics
Patient sampling	Primary objective: to investigate the expression of pro‐apoptotic and anti‐apoptotic genes in eutopic endometrium from women with endometriosis and healthy controls in relation to disease occurrence and severity Participants: women undergoing laparoscopy for pelvic pain Selection criteria: inclusion criteria: reproductive age, regular menstrual cycle (26‐34 days), proliferative phase of the menstrual cycle, no history of hormonal treatment, chronic pelvic pain lasting at least 6/12 months, written informed consent; exclusion criteria: previous abdominal surgery, pregnancy; history of in vitro fertilisation treatment Study design: observational single‐gate, prospective recruitment and sample collection
Patient characteristics and setting	Clinical presentation: pelvic pain Age: mean age 41.5 ± 5.6 years (range 28–50) Number enrolled: 51 women Number available for analysis: 45 women (all at proliferative cycle phase) Setting: university hospital ‐ Clinic of O&G, Jessenius Faculty of Medicine, Comenius University Place of study: Martin, Slovakia Period of study: not reported Language: English
Index tests	Index test: p53 mRNA, Bcl‐Xs mRNA, Bax mRNA, Bcl‐XL mRNA, Bcl‐XL:BclXS ratio Description of positive case definition by index test as reported: mRNA expression by RT‐PCR (Detection by using (RevertAidTM H Minus First Strand cDNA Synthesis Kit, Fermentas–EU); normalised to b‐actin mRNA; the products tested with gel electrophoresis, expression intensity analysed by Gene Tools software (Syngene, UK); technique described; thresholds not provided Examiners: none stated; unclear if blinded to the reference standard Interobserver variability: not reported
Target condition and reference standard(s)	Target condition: endometriosis Prevalence of target condition in the sample: n/N = 15/45 (33%): stages I‐II 9, stages III‐IV 6 ; controls 30 Reference standard: laparoscopy + histology Description of positive case definition by reference test as reported: endometriosis was defined as the presence of glandular epithelium and endometroid stroma in lesions outside of the uterus and disease was staged according to the rAFS classification Examiners: none stated
Flow and timing	Time interval between index test and reference standard: tissue biopsies obtained at surgery Withdrawals: 6 women did not fulfil inclusion criteria (excluded before the experiments)
Comparative
Notes	Conclusion: Results suggest that an increased transcription of pro‐apoptotic genes (p53 and Bcl‐xS) in eutopic endometrium is significantly associated with endometriosis, which indicates dysregulation of apoptotic gene transcription associated with disease. Comments: For p53, Bcl‐XS there was statistically significant difference between the groups, but there were insufficient data to construct 2 × 2 tables; not included in this review For Bax, Bcl‐XL, Bcl‐XL:BclXS ratio there was no statistically significant difference between the groups ‐ no data available for meta‐analysis (Appendix 7)
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Unclear
Did the study avoid inappropriate exclusions?	Yes
Was a 'two‐gate' design avoided?	Yes
		Unclear	Unclear
DOMAIN 2: Index Test All tests
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	No
Was a menstrual cycle phase considered in interpreting the index test	Yes
		High	Low
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
		Low	Low
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Yes
Did all patients receive the same reference standard?	Yes
Were all patients included in the analysis?	Yes
		Low

(r)AFS: (revised) American Fertility Society; (r)ASRM: (revised) American Society for Reproductive Medicine; AUB: abnormal uterine bleeding; CA‐125: cancer antigen; COC: combined oral contraceptive; CPP: chronic pelvic pain; CV: coefficient of variation; DIE: deep infiltrating endometriosis; DUB: dysfunctional uterine bleeding; EGF: epidermal growth factor; ER: oestrogen receptor; ESHRE: European Society of Human Reproduction and Embryology; FISH: fluorescence in situ hybridisation; GnRH: gonadotropin‐releasing hormone; HPF: high‐power fields; HS: hybridisation score; IC: intracellular;IHC: immunohistochemistry/immunohistochemical; IL: interleukin; IRS: immunoreactive score; IUD: intrauterine device; LS‐SVM: least squares support vector machine; MMP: matrix metalloproteinase; mRNA: messenger RNA; N: total sample size; n: number of events; NA: not applicable; NSAIDS: nonsteroidal anti‐inflammatory drugs; PCR: polymerase chain reaction; PID: pelvic inflammatory disease; qRT‐PCR: quantitative reverse transcription PCR; RIA: radioimmunoassay; SD: standard deviation; SVM: support vector machine; TNF: tumour necrosis factor; TSR: telomerase substrate oligonucleotide.

For a full list of endometrial biomarkers and their biological annotation, please see Appendix 1.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Abae 1992	Target condition outside inclusion criteria (only healthy women included)
Absenger 2004	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Abu Musa 1992	Population likely overlapped with Takahashi 1990, unable to clarify with the authors
Acien 2007	Index test outside inclusion criteria (comparisons between endometrium in control group and endometriotic lesions)
Adamyan 1993	Insufficient description of the study methods and population (unclear patient selection); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Aghajanova 2009a	Insufficient description of study population (unclear if all controls underwent abdominal surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm the reported negative result)
Aghajanova 2009b	Insufficient description of study population (unclear if all controls underwent abdominal surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm the reported negative result)
Aghajanova 2011	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Agic 2007	Insufficient description of the study methods and population
Akoum 1999	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Akoum 2001	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Akoum 2006	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Akoum 2007	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table); population likely overlapped with Lawson 2008
Alizadeh 2011	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ametzazurra 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Andersson 2014	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Anger 2007	Study design outside exclusion criteria (archived samples, poorly defined population)
Antsiferova 2005	Reference standard outside inclusion criteria (no laparoscopy in control group); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Badawy 1984	Insufficient description of the study methods and population (unclear timing of sample collection); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Baka 2011	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Balasch 1985	Reference standard outside inclusion criteria (controls did not have pelvic surgery)
Ballester 2012	Insufficient description of the study methods and population
Barrier 2006	Study design outside exclusion criteria (archived samples, poorly defined population)
Bartosik 1987	Insufficient description of study population (unclear if all the participants were of reproductive age, unable to clarify with the authors)
Bellehumeur 2005	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Bellelis 2013	Index test outside inclusion criteria (comparisons between endometrium in control group and endometriotic lesions)
Berbic 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Berbic 2010	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Bergqvist 2001	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Bohler 2007	Reference standard outside inclusion criteria (no laparoscopy in control group); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Braun 2002	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Braun 2007	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to verify negative result)
Browne 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Budrys 2012	Index test outside inclusion criteria (functional study of endometrium; no direct measurement or comparison of biomarker expression between the groups)
Bukulmez 2008	Insufficient description of population recruitment and characteristics; insufficient diagnostic accuracy information (unable to construct 2 × 2 table or confirm negative findings)
Bulmer 1998	Study design outside exclusion criteria (archived samples, poorly defined population); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Burlev 2005	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Burlev 2006	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Burney 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Calcagno 2011	Index test outside inclusion criteria (evaluation of ectopic endometrium in endometriosis vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis)
Carneiro 2007	Insufficient diagnostic accuracy information [unable to construct 2x2 table or to confirm negative result]
Carneiro 2008	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm negative result)
Carvalho 2013	Index test outside inclusion criteria (evaluation of peritoneal fluid and ectopic endometrium; no comparative data on eutopic endometrium between the groups)
Chand 2007	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Chang 2013	Insufficient description of study population (unclear if all controls underwent abdominal surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Chegini 2003	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Chehna‐Patel 2010	Insufficient description of study methods and population; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Chen 2006	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Chen 2012a	Insufficient description of the study methods and study population
Chen 2012b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Chen 2012c	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Cho 2009	Reference standard outside inclusion criteria (no surgery in 15% of participants); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Cho 2010	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Chung 2001	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table; unable to confirm the reported negative result)
Chung 2002a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table; unable to confirm the reported negative result)
Chung 2002b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Colette 2004	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Colette 2006	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table; unable to confirm the reported negative result)
Cosin 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Cunha‐Filho 2001	Index test outside inclusion criteria (luteal insufficiency measured as combined blood‐endometrium test)
D'Amico 2013	Index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Daftary 2004	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Dai 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
De Graaff 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Debrock 2004	Review article
Delbandi 2013	Index test outside inclusion criteria (functional study of endometrium; no direct measurement or comparison of biomarker expression between the groups)
Delvoux 2009	Index test outside inclusion criteria (focus on biological events in endometrium; no direct measurement or comparison of biomarker expression between the groups)
Dharmaraj 2014	Insufficient description of study population (unclear if all controls had pelvic surgery)
Di Carlo 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Dimitriadis 2006	Insufficient description of study methods and population (unclear timing of sample collection)
Dmowski 2001	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Donnez 1998	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Donnez 2013	Study design outside exclusion criteria (archived samples, poorly defined population)
Ejskjaer 2009	Reference standard outside inclusion criteria (no laparoscopy in control group); insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm the reported negative result)
Fang 2003	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Fasciani 2010	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Fazleabas 1987	Insufficient description of study methods and population
Fedele 1988	Insufficient description of study methods and population
Fernandez‐Shaw 1995a	Insufficient description of study methods and population (unclear age of the participants)
Fernandez‐Shaw 1995b	Index test outside inclusion criteria (focus on biological events in endometrium; no direct measurement or comparison of biomarker expression between the groups)
Ferriani 1993	Index test outside inclusion criteria (focus on biological events in endometrium; no direct measurement or comparison of biomarker expression between the groups)
Finas 2008	Study design outside exclusion criteria (archived samples, poorly defined population)
Fowler 2007	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Fujino 2006	Insufficient description of the study methods and population; the study presents negative findings
Fukaya 1999	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Gaetje 2006	Insufficient description of study methods and population (unclear if all controls had abdominal surgery and timing of sample collection)
Gaetje 2007a	Reference standard outside inclusion criteria (not all controls had abdominal surgery); index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis)
Gaetje 2007b	Insufficient description of study methods and population (unclear if all controls had abdominal surgery)
Gagné 2003	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Gebel 1998	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Giannelli 2007	Index test outside inclusion criteria (focus on biological events in endometrium; no direct measurement or comparison of biomarker expression between the groups)
Gonzalez Ramos 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Gori 2013	Index test outside inclusion criteria (focus on biological events in endometrium; no direct measurement or comparison of biomarker expression between the groups)
Guay 2011	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Guo 2013	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Hapangama 2008	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Hapangama 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Hapangama 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Hassa 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm the reported negative result)
Hawkins 2011	Index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; functional study of endometrium)
Hey‐Cunningham 2013	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Hii 1998	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm negative findings)
Hsu 2014	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Huang 2012	Index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis)
Huang 2013	Index test outside inclusion criteria (functional study of endometrium)
Hudelist 2005a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm negative findings)
Hudelist 2005b	Index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis)
Hudelist 2008	Index test outside inclusion criteria (functional study of endometrium)
Hur 2006	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Hurst 2014	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Hwang 2013	Insufficient information on study population and sample collection; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Igarashi 2005	Reference standard outside inclusion criteria (controls did not have pelvic surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Isaacson 1990	Insufficient description of study methods and population
Jana 2013	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Johnson 2005	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Jolicoeur 1998	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Jones 1995	Study design outside exclusion criteria (archived samples, poorly defined population)
Jones 1996	Study design outside exclusion criteria (archived samples, poorly defined population)
Jones 1998	Study design outside exclusion criteria (archived samples, poorly defined population)
Jones 2009	Reference standard outside inclusion criteria (controls did not have pelvic surgery)
Juhasz‐Boss 2011	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm the reported negative result)
Jurgensen 1996	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm negative result)
Kao 2003	Insufficient description of study methods (unclear timing of sample collection); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Karalok 2014	Reference standard outside inclusion criteria (controls did not have pelvic surgery)
Khan 2003	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Khan 2005a	Index test outside inclusion criteria (functional study of endometrium; no direct measurement or comparison of biomarker expression between the groups)
Khan 2005b	Index test outside inclusion criteria (evaluation of peritoneal macrophages, not endometrial biomarker)
Khan 2010	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Khan 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Khan 2013	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Kharfi 2001	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Kharfi 2002	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Kharfi 2003	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Khorram 2002	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Kim 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Kim 2013a	Index test outside inclusion criteria (functional study of endometrium, focus on endometrial activity in response to TNF‐α treatment)
Kim 2013b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Kim 2013c	Index test outside inclusion criteria (functional study of endometrium, focus on endometrial activity following manipulation of Pak4)
Kitawaki 1999a	Target condition outside inclusion criteria (diseased vs disease‐free tissue, no separate diagnostic data for endometriosis)
Kitawaki 2000a	Insufficient information on study population; insufficient diagnostic accuracy information (unable to construct 2 × 2 table or confirm negative findings)
Kitawaki 2000b	Target condition outside inclusion criteria (diseased vs disease‐free tissue, no separate diagnostic data for endometriosis)
Klemmt 2006	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Klemmt 2007	Index test outside inclusion criteria (functional study of endometrium; no direct measurement or comparison of biomarker expression between the groups)
Kobayashi 2012	Study design outside exclusion criteria (archived samples, poorly defined population); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Kocbek 2014a	Index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis)
Kocbek 2014b	Index test outside inclusion criteria (comparisons between endometrium in control group and endometriotic lesions)
Koshiba 2005	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm negative result)
Kreiner 1986	Insufficient information on study population (unclear if all the participants were of reproductive age), unable to contact the authors
Kyama 2006a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm negative findings)
Kyama 2006b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Kyama 2008	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm negative findings)
Kyama 2011	Insufficient description of the population recruitment and demographics (unclear if all the controls had pelvic surgery; unclear timing of sample collection in relation to surgery)
Laudanski 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Laudanski 2013a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Laudanski 2013b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Lawson 2007	Index test outside inclusion criteria (only ectopic endometrium assessed)
Lebovic 2002	Index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis)
Lee 2011	Index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis)
Lee 2014	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Leiva 1994	Insufficient description of study methods and/or population; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Lessey 1989	Insufficient description of study methods and/or population; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Lessey 1993	Insufficient description of study population; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Lessey 1994	Insufficient description of study population (unclear if all controls had pelvic surgery)
Li 2008	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Li 2010	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Li 2011	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Li 2012a	Insufficient description of study population (unclear if all controls had pelvic surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Li 2012b	Insufficient description of study population (unclear if all controls had pelvic surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Lima‐Couy 2004	Insufficient description of study methods and population (unclear if all controls had pelvic surgery, if prospective sample collection and time interval sampling to surgery)
Lin 2005	Reference standard outside inclusion criteria (controls did not have pelvic surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Lin 2010	Population outside inclusion criteria (women with genital tract malignancy were included); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Lin 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Liu 2002	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Liu 2005a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Lo Vasco 2012	Case report (1 participant in each group)
Locci 2013	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Luk 2005	Index test outside inclusion criteria (functional study of endometrium, focus on endometrial activity regulated by sex steroids)
Luk 2010	Index test outside inclusion criteria (functional study of endometrium)
Luo 2006a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Luo 2006b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Maia 2012	Insufficient description of study methods and population
Makri 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Malik 2006	Time interval outside inclusion criteria (exceeded 12 months)
Mathur 1990	Insufficient description of study population and sample collection methods; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Matsuzaki 2004a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Matsuzaki 2004b	Population outside inclusion criteria (only women with endometriosis were included, comparison eutopic vs ectopic endometrium)
Matsuzaki 2005a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or confirm negative result)
Matsuzaki 2005b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Matsuzaki 2006b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Matsuzaki 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Matsuzaki 2010a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Matsuzaki 2010b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Matsuzaki 2012	Index test outside inclusion criteria (functional study of endometrium;focus on localization of ABCG2þ cells and expression across menstrual cycle)
Matsuzaki 2013	Index test outside inclusion criteria (functional study of endometrium; no direct measurement or comparison of biomarker expression between the groups)
McBean 1993	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Mei 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Mei 2013	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Meola 2010	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Meola 2013a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Meresman 2000	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Meresman 2002	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Mettler 1996	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table confirm negative findings)
Mettler 1997	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or confirm negative findings)
Mihalich 2003	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Mikolajczyk 2007	Population overlapped with Mikolajczyk 2006
Mikolajczyk 2009	Insufficient description of study methods and/or population (unclear if controls had surgery and if prospective sample collection)
Minina 1989	Insufficient description of study methods and population (unclear if all controls had abdominal surgery and timing of sample collection)
Morsch 2009	Insufficient description of study methods and population
Mu 2008	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Newman 2013	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or confirm negative findings)
Nikoo 2014	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Noble 1996	Insufficient description of study methods and population
Nomiyama 1997	Insufficient description of study methods and population (unclear how the groups were selected from the entire cohort); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Novella‐Maestre 2010	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ordi 2003	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm negative findings)
Ota 1996	Insufficient description of study population (unclear if all controls had pelvic surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ota 1997a	Insufficient information on study population (unclear if all controls had pelvic surgery; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ota 1997b	Insufficient information on study population (unclear if all controls had pelvic surgery; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ota 2000	Insufficient information on study population (unclear if all controls had pelvic surgery; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ota 2001a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ota 2001b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ota 2002	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Pabona 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Pan 2007	Publication was retracted
Pan 2008	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Pan 2009	Insufficient description of study population (unclear if all controls had pelvic surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Park 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Pellegrini 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Penna 2008	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Petracco 2011	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Pillai 1996	Population outside inclusion criteria (women post Caesarean Section were included); insufficient description of study population (unclear if all the participants were of reproductive age)
Plante 2012	Insufficient description of study population (unclear if all controls had pelvic surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ponce 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Prentice 1992	Insufficient description of study methods and/or population; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Rai 1996	Insufficient description of study methods and/or population
Rai 2010	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ramon 2005	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ramon 2008	Population outside inclusion criteria (only women positive for certain genetic polymorphisms were tested)
Rey 1998	Insufficient description of study population (unclear if all controls had pelvic surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Rocha 2011	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Rocha 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Rombauts 2006	Insufficient description of study population (unclear if all controls had pelvic surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ruan 2010	Insufficient description of study population (unclear if all controls had pelvic surgery)
Ruan 2013	Insufficient description of the study population (unclear if all controls had abdominal surgery, unclear timing of sample collection); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Saracoglu 1985	Insufficient description of study methods and population (unclear timing of sample collection)
Sbracia 1997	Insufficient description of study population (unclear if all controls had pelvic surgery)
Schor 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Schulke 2009	Index test outside inclusion criteria (endometrial samples from full‐thickness uterine block, not curette); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Schutt 2015	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to confirm negative results)
Seo 2010a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Seo 2010b	Insufficient description of study population (unclear if all controls had pelvic surgery)
Sha 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Sharpe Timms 2000	Insufficient description of study population; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Sharpe‐Timms 1994	Insufficient description of study population (unclear age group and clinical/demographic characteristics)
Shen 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Shen 2013	Insufficient description of study population (unclear if all controls had pelvic surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Sherwin 2008	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Shi 2014a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Shi 2014b	Index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Silveira 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Stephens 2010	Insufficient description of study population; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Sun 2002	Index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis)
Szczepanska 2007	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table); population likely overlapped with Szczepanska 2010a and Szczepanska 2010b
Szczepanska 2010a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Szczepanska 2010b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Szymanowski 2007	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Szymanowski 2008	Unable to locate a full text
Takahashi 1988a	Population likely overlapped with Takahashi 1990, unable to clarify with the authors
Takahashi 1988b	Population likely overlapped with Takahashi 1990, unable to clarify with the authors
Takahashi 1989	Reference standard outside inclusion criteria (no pelvic surgery in controls); insufficient description of study population)
Takahashi 1991	Population likely overlapped with Takahashi 1990, unable to clarify with the authors
Takehara 2004	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Tan 2001	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Tan 2002	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ten Have 2007	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Tokushige 2006	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table, descriptive data for separate endometrial layers); ˜30% samples from full‐thickness uterine block, not curette
Tokushige 2007	Index test outside inclusion criteria (test of full‐thickness uterine block of endometrium and contiguous myometrium from hysterectomy samples); insufficient diagnostic accuracy information (data for separate layers of endometrium functional and basal); population overlapped with Tokushige 2006
Torres 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table; unable to confirm the reported negative result)
Trio 2007	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Tshishi 2010	Insufficient description of study population; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ulukus 2005	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Ulukus 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Uz 2011	Index test outside inclusion criteria (functional study of endometrium; no direct measurement or comparison of biomarker expression between the groups)
Uzan 2004	Insufficient description of population sampling and characteristics
Uzan 2005	Insufficient description of population sampling and characteristics
Velasco 2006	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table; negative assay in both control and study groups)
Vergetaki 2013	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Vernet‐Tomas 2006b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Vestergaard 2011	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table or to verify the reported negative result)
Wang 2005a	Insufficient description of the study population (unclear if all the participants were of reproductive age, unclear timing of sample collection)
Wang 2005b	Insufficient description of the study population (unclear if all controls had abdominal surgery, unclear timing of sample collection)
Wang 2010b	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Wang 2012	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Wei 2009	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Wei 2013	Insufficient description of the study population (unclear if all controls had abdominal surgery); differing reference standards between cases and controls; insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Wenzl 1998	Study design outside exclusion criteria (archived samples, poorly defined population)
Wolun‐Cholewa 2011	Index test outside inclusion criteria (focus on technical aspects of the test)
Wu 1999	Insufficient description of the study population (unclear if all controls had abdominal surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Wu 2003	Index test outside inclusion criteria (evaluation of ectopic endometrium vs eutopic endometrium in controls; no data on eutopic endometrium in women with endometriosis)
Wu 2007	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Xu 2010	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Yi 2003	Index test outside inclusion criteria (functional study of endometrium)
Yin 2006	Insufficient description of the study population (unclear if all controls had abdominal surgery); insufficient diagnostic accuracy information (unable to construct 2 × 2 table) for the positive tests
Yoo 2014	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Zeng 2012	Population outside inclusion criteria (postmenopausal women and women with genital tract malignancy were included); insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Zhang 2005	Insufficient description of study design and population characteristics; insufficient diagnostic accuracy information (unable to construct 2 × 2 tables)
Zhang 2006	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Zhang 2009a	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Zhang 2009b	Population outside inclusion criteria (control group included only women with adenomyosis and leiomyoma)
Zhang 2010	Population outside inclusion criteria (only women with endometriosis were included, comparison eutopic vs ectopic endometrium)
Zhang 2011a	Index test outside inclusion criteria (functional study of endometrium; no direct measurement or comparison of biomarker expression between the groups)
Zhang 2011b	Index test outside inclusion criteria (functional study of endometrium; no direct measurement or comparison of biomarker expression between the groups)
Zhao 2014	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)
Zong 2004	Insufficient diagnostic accuracy information (unable to construct 2 × 2 table)

Differences between protocol and review

General scope: this review is a part of the review series arising from the same generic protocol. We adjusted the following sections to the main topic of the review as follows.

• Background. We modified the section on the index test and removed all the irrelevant information on imaging. We updated the Rationale section to include a clearer definition of triage diagnostic tests.

• Objectives.

  • We updated the list of the sources of heterogeneity.

  • During the revision of the literature on the subject, we identified a substantial body of studies looking at the biomarkers that were not altered in the presence of endometriosis (we found no statistically significant difference between women with and without the disease). We believe that presenting this type of data, when obtained from adequately designed studies, is important for both clinicians and researchers. We explained our reasoning in the Background section under Rationale, in the Methods section under Criteria for considering studies for this review, Index tests and added a secondary objective to Objectives: "to assess the biomarkers that were not affected by endometriosis and hence are unlikely to discriminate between patients with and without the disease".

• Methods. We updated the criteria for considering studies for this review as follows.

  • Types of studies: We removed the 'cohort' and 'case control' classification and introduced the concept of single‐gate design and two‐gate' design, defined as the presence of a single or multiple set of inclusion criteria with regard to the clinical condition or the reference standard. We found this classification more informative to describe diagnostic studies, all of which are cross‐sectional in nature. We limited the inclusion criteria to the studies with a single set of inclusion criteria by reference standard (i.e. all women who underwent abdominal surgery), but included single or multiple sets of inclusion criteria by clinical presentation (i.e. women with suspected endometriosis or other indications for abdominal surgery), referring to these as single‐gate and two‐gate designs, respectively.

  • Likewise, we removed the terminology 'prospective studies' and introduced 'studies performed on prospectively collected samples'. This decision was guided by the fact that most diagnostic studies are retrospective in nature, as they aim to compare the results of the index test with the results of the reference standard in the same group of participants, where the groups are classified by the outcome of the reference standard. Also, the analysis of the index test could have been performed retrospectively in a single batch on stored samples after the prospective collection of samples. The timing of sample collection (before or after surgical treatment of the disease) from preoperatively recruited population has more impact on the test result than the timing of the laboratory assay. Therefore, we included only studies that collected endometrial tissue before the reference surgical procedure, i.e. 'prospectively collected', irrespective of the actual timing of test performance. In this way, we avoided the confusion of labelling studies as 'prospective' or 'retrospective'. This allowed us to include the studies from well‐established high quality tissue banks using well‐characterised archived samples, as omission of these studies could result in the loss of potentially valuable data. We discuss this choice in the Methods under Criteria for considering studies for this review.

  • We modified index tests to include only eutopic endometrial markers, updating the table listing the tests of interest (Table 4) accordingly.

  • Target conditions now also include deep pelvic endometriosis in view of the growing body of literature on this condition as a separate entity and its diagnostic importance to optimise the surgical approach.

  • Spectrum of disease: Following an ad hoc observation, we included the studies that involved only selected populations of women with endometriosis (i.e. specific rASRM stages) in view of the emerging evidence on poor correlation of this classification with infertility and pain symptoms. Exclusion of such studies could result in the loss of potentially important diagnostic information from otherwise eligible publications. Where possible we aimed to address the impact of the inclusion of these studies in investigations of heterogeneity.

• Search methods for identification of studies.

  • In the protocol, we stated that we would identify grey literature (unpublished studies including conference proceedings and reports) and also define specific search strategies. In practice, the paucity of relevant data that was available from abstracts made it impossible to apply the selection criteria and methodological quality judgement to these studies. We anticipated that identification of this type of study and attempts to obtain the necessary information directly from the study investigators would increase the already labour‐intensive work involved in preparation of this review. Therefore, by consensus among the key authors, we removed the unpublished studies we had already identified and did not complete an intended search for unpublished material.

  • We updated the search strings for all biomarkers excluding imaging (searched separately), applying the same principles as presented in the protocol.

• Assessment of methodological quality: We tailored the QUADAS‐2 tool for the topic of the review, outlining the differences between the original QUADAS‐2 tool and the tool designed for this review in the relevant section (see Assessment of methodological quality).

• Analysis.

  • We amended the section on statistical methods and tailored it to the types of tests included in the review.

  • We performed no sensitivity analyses and no assessment of heterogeneity due to insufficient data.

  • When we judged a test performance against the predetermined diagnostic criteria, we only considered the point estimates of sensitivity and specificity, as we believe that presenting these metrics of test performance is the most helpful and informative way to summarise the diagnostic data. We acknowledge that the choice of the most helpful summary is subjective. There are tests where the point estimates did not reach the predetermined criteria, but the confidence intervals (CIs) contain the values above the thresholds for replacement tests, triage tests or both. These tests could have diagnostic value if the point values underestimated their diagnostic potential. For the tests where the point estimates reached the criteria for a replacement or triage tests but the CIs contained values below the thresholds, point values could have overestimated the diagnostic performance of the test. If we use the range of the CIs rather than the point estimates of the data, the predetermined cut‐off becomes meaningless. We did not consider CIs in qualifying the test performance; however, we utilised the CIs in interpreting the reliability of the obtained data.

The authors' list and order changed to accurately reflect their contribution to the review.

Contributions of authors

Vicki Nisenblat and Louise Hull coordinated the production of the protocol, performed literature search and coordinated production of the review series. Devashana Gupta and Vicki Nisenblat undertook the review and took the primary role in writing the review. Laura Miller participated in data collection and quality appraisal of the included studies. Patrick Bossuyt provided advice on statistical methods for the review and performed the analyses. Ian Fraser and Neil Johnson contributed to the design of the review and critically appraised the discussion. All the authors contributed to the revision and drafting of the review.

Sources of support

Internal sources

• Cochrane Menstrual Disorders and Subfertility Group, University of Auckland, New Zealand.

  Technical support

• The Robinson Institute, University of Adelaide, Australia.

  Access to academic resources

External sources

• No sources of support supplied

Declarations of interest

Vicki Nisenblat: none known.

Louise Hull: Dr M. L. Hull obtained a grant of $10,000 to carry out a prevalence study of ultrasonographically diagnosed endometriosis in a fertility population.

Devashana Gupta: none known.

Laura Miller: none known.

Patrick Bossuyt: none known.

Ian Fraser: Ian Fraser is a member of Advisory Boards on Contraception and/or Abnormal Uterine Bleeding for Bayer Schering Pharma, Merck/MSD and Vifor Pharma; he has also received lecture fees, consultancy fees and expenses for contraception and abnormal uterine bleeding from Bayer Schering Pharma, Merck/MSD and Daiichi Sankyo. He has received a small grant for a collaborative research study with the University of Melbourne on an animal model of endometriosis.

Neil Johnson: Professor Neil Johnson is involved in research funded by Abb‐Vie. He has received support to attend conferences from MSD, Merck‐Serono and Bayer.

New