Abstract
Objective
Although a fraction of endometrial hyperplasia cases have concurrent endometrial carcinoma, patient characteristics associated with concurrent malignancy are not well described. The aim of our study was to identify predictive clinico-pathologic factors for concurrent endometrial carcinoma among patients with endometrial hyperplasia.
Methods
A case-control study was conducted to compare endometrial hyperplasia in both preoperative endometrial biopsy and hysterectomy specimens (n = 168) and endometrial carcinoma in hysterectomy specimen but endometrial hyperplasia in preoperative endometrial biopsy (n = 43). Clinico-pathologic factors were examined to identify independent risk factors of concurrent endometrial carcinoma in a multivariate logistic regression model.
Results
The most common histologic subtype in preoperative endometrial biopsy was complex hyperplasia with atypia [CAH] (n = 129) followed by complex hyperplasia without atypia (n = 58) and simple hyperplasia with or without atypia (n = 24). The majority of endometrial carcinomas were grade 1 (86.0%) and stage I (83.7%). In multivariate analysis, age 40–59 (odds ratio [OR] 3.07, p = 0.021), age ≥ 60 (OR 6.65, p = 0.005), BMI ≥ 35 kg/m2 (OR 2.32, p = 0.029), diabetes mellitus (OR 2.51, p = 0.019), and CAH (OR 9.01, p = 0.042) were independent predictors of concurrent endometrial carcinoma. The risk of concurrent endometrial carcinoma rose dramatically with increasing number of risk factors identified in multivariate model (none 0%, 1 risk factor 7.0%, 2 risk factors 17.6%, 3 risk factors 35.8%, and 4 risk factors 45.5%, p < 0.001). Hormonal treatment was associated with decreased risk of concurrent endometrial cancer in those with ≥ 3 risk factors.
Conclusions
Older age, obesity, diabetes mellitus, and CAH are predictive of concurrent endometrial carcinoma in endometrial hyperplasia patients.
Keywords: Endometrial hyperplasia, Endometrial cancer, Risk factor, Age, Diabetes, Obesity
1. Introduction
Endometrial hyperplasia is a benign lesion that can precede or coexist with endometrial carcinoma. Cancer of the uterus is the most prevalent gynecologic malignancy in the United States, with 54,870 projected cases in 2015 [1]. The majority of endometrial carcinomas will be type I, which are hormonally sensitive and most commonly have endometrioid histology. The rate of concurrent endometrial carcinoma in patients with untreated complex hyperplasia with atypia (CAH), the most severe category of the disease, is as high as 42.6% [2].
Endometrial hyperplasia is categorized using the World Health Organization (WHO) criteria based on the presence or absence of nuclear atypia and the degree of architectural crowding [3]. The standard of care for patients with atypical hyperplasia who are able and willing to undergo surgery, is hysterectomy [4]. Patients who desire future childbearing may choose to defer definitive management and opt for hormonal treatment of endometrial hyperplasia. The decision to forego surgical management is complicated by the lack of consensus regarding the non-surgical management of endometrial hyperplasia, and difficulty in identifying patients that are at increased risk of concurrent carcinoma.
Besides the presence of nuclear atypia, the characteristics that confer an increased risk of concurrent endometrial carcinoma in patients with endometrial hyperplasia are not well described. Identification of risk factors could allow better-informed counseling for patients considering non-surgical management of endometrial hyperplasia. It could also assist in the appropriate triage of patients desiring surgical management to a gynecologic oncologist, given that 1.2—2.1% of patients with CAH are estimated to have endometrial carcinoma with lymph node metastases [5]. The objective of this study was to identify clinico-pathologic factors for concurrent endometrial carcinoma in patients with endometrial hyperplasia.
2. Patients and methods
2.1. Eligibility
After Institutional Review Board (IRB) approval was obtained at the University of Southern California, an institutional surgical pathology database was utilized to identify cases. For the case group representing endometrial hyperplasia with concurrent carcinoma, the diagnostic code “endometrial cancer” was used to identify patients with the diagnosis of primary endometrial cancer in a hysterectomy specimen between March 2003 and March 2014. Among those identified cases, the results of the endometrial biopsies performed antecedent to the hysterectomy were reviewed. Eligibility criteria for the study were: women with a diagnosis of endometrial hyperplasia in an endometrial biopsy specimen who subsequently underwent hysterectomy for treatment. Exclusion criteria were those without endometrial hyperplasia in a preoperative endometrial biopsy (normal endometrium, atrophic endometrium, or endometrial cancer) and those with no endometrial biopsy prior to hysterectomy. For the control group representing endometrial hyperplasia without concurrent carcinoma, the diagnostic code “endometrial hyperplasia” was used to identify patients that had a diagnosis of any type of endometrial hyperplasia in a hysterectomy specimen during the same study period. The same eligibility and exclusion criteria were applied for the control group. The STROBE guidelines were consulted for reporting in a case-control study [6]. Some cases of the study population were within the context of our previous studies [7–11].
2.2. Clinical information
Among eligible cases, medical records were examined to abstract the following information: (i) patient demographics at the time of endometrial hyperplasia diagnosis, (ii) treatment patterns for endometrial hyperplasia, and (iii) histology results for endometrial biopsies and hysterectomy. Patient demographics included age, ethnicity, body mass index (BMI), pregnancy history, and presence of medical comorbidities (hypertension, diabetes mellitus, and hypercholesterolemia). Methods of endometrial sampling was also abstracted (pipelle, dilation and curettage, or vacuum aspiration with Vabra). Treatment patterns included whether there was hormonal treatment between endometrial hyperplasia diagnosis and hysterectomy. Among those who received treatment, the hormonal agent was recorded. In addition, the time interval between the initial endometrial hyperplasia diagnosis and the date of hysterectomy was abstracted. Histology results included type of endometrial hyperplasia in the initial endometrial biopsy. These were classified into: CAH, complex hyperplasia without atypia (CH), simple hyperplasia with atypia (SAH), and simple hyperplasia (SH). Among patients that were diagnosed with endometrial cancer on the hysterectomy specimen, histologic subtype, grade, cancer stage, depth of myometrial tumor invasion, and presence of lymphovascular space invasion were abstracted. In addition, survival outcomes (presence of disease recurrence or death due to endometrial cancer) were examined. Cancer stage was re-classified based on the International Federation of Gynecology and Obstetrics (FIGO) 2009 system [12].
2.3. Statistical analysis
The primary objective of this study was to identify independent risk factors for the presence of concurrent endometrial cancer in hysterectomy specimens among patients with endometrial biopsy-proven endometrial hyperplasia. Continuous variables were examined for normality by the Kolmogorov-Smirnov test expressed with mean (±SD) or median (range) as appropriate. Statistical significance of continuous variables was examined by the Student t-test or the Mann-Whitney U-test as appropriate. For continuous variables, clinically relevant age cutoff (<40, 40–59, and ≥60), mean value for BMI (<35 versus ≥35 kg/m2), and median for time interval between endometrial biopsy and hysterectomy (<105 versus ≥105 days) were used for the cutoff values. Pearson’s correlation coefficient value was determined among the continuous variables. Categorical or ordinal variables were expressed with number (%), and statistical significance was examined by chi-square test or the Fisher exact test as appropriate.
A binary logistic regression model for multivariate analysis was used to determine independent risk factors for concurrent endometrial cancer in a population of women with biopsy-proven endometrial hyperplasia. All variables with significance at p < 0.10 in univariate analysis were considered as candidates in the final model. This relatively liberal cutoff of the p-value was chosen due to the small sample size that may have a risk of type II error with a lower cutoff. These included age (<40 versus 40–59 versus ≥60), BMI (<35 versus ≥35 kg/m2), diabetes mellitus (yes versus no), hypertension (yes versus no), hypercholesterolemia (yes versus no), any hormonal treatment for endometrial hyperplasia (yes versus no), time interval between endometrial biopsy and hysterectomy (< 105 versus ≥105 days), and endometrial hyperplasia type (SH/SAH versus CH versus CAH). Then, a conditional backward method was used to determine the independently significant covariates in the final model given the degree of freedom in this sample size, expressed with odds ratio (OR) and 95% confidence interval (CI). Covariates entered in the final model were age, BMI, diabetes mellitus, and endometrial hyperplasia type. All statistical tests were two-tailed, and P value of less than 0.05 was considered as statistically significant. Statistical Package of Social Science (SPSS, Inc., version 12.0, Chicago, IL) was used for statistical analysis.
3. Results
There were 194 hysterectomy specimens with a preoperative diagnosis of endometrial hyperplasia identified during the study period. Of those, 20 cases lacking an endometrial biopsy prior to hysterectomy and 9 cases without endometrial hyperplasia on endometrial biopsy, were excluded. The remaining 168 cases had endometrial hyperplasia on endometrial biopsy that was subsequently treated with hysterectomy. There were 674 cases of primary endometrial cancer that underwent hysterectomy during the study period. Of those endometrial cancer cases, 43 hysterectomies were performed for the preoperative indication of endometrial hyperplasia. Taken together, 211 cases of endometrial biopsy-proven endometrial hyperplasia with subsequent hysterectomy comprised the study population.
Patient demographics are shown in Table 1. Mean age was 45.2, and the majority of the study population was Hispanic (73.9%). Mean BMI of the study patients was 35.6 kg/m2, and approximately three quarters (75.8%) were obese. Age and BMI had an inverse relationship (Pearson’s r = −0.24, p < 0.001). Approximately one fourth (25.1%) were nulligravid. Medical comorbidities were common in this study population and included hypertension (38.4%), diabetes mellitus (24.6%), and hypercholesterolemia (14.2%). Median time interval between the date of initial endometrial hyperplasia diagnosis and the date of hysterectomy was 105 days. There were 98 (46.4%) patients who received hormonal treatment for endometrial hyperplasia with megestrol acetate being the most common agent (28.4%) followed by medroxyprogesterone (9.0%) and the levonorgestrel-releasing intrauterine device (3.8%). The most common method of initial endometrial sampling was endometrial pipelle (67.4%) followed by vacuum aspiration with Vabra (25.1%). The most common histologic subtype of endometrial hyperplasia was CAH (61.1%) followed by CH (27.5%).
Table 1.
Risk factors for concurrent endometrial cancer in patients with endometrial hyperplasia.
| All |
Concurrent cancer |
P-value | ||
|---|---|---|---|---|
| N = 211 | Yes (n = 43) | No (n = 168) | ||
| Age | 45.2 (±9.1) | 49.4 (±9.3) | 44.1 (±8.7) | 0.001 |
| <40 | 62 (29.4%) | 7 (16.3%) | 55 (32.7%) | |
| 40–59 | 133 (63.0%) | 29 (67.4%) | 104 (61.9%) | |
| ≥60 | 16 (7.6%) | 7 (16.3%) | 9 (5.4%) | |
| Ethnicity | 0.21 | |||
| Caucasian | 22 (10.4%) | 1 (2.3%) | 21 (12.5%) | |
| African American | 20 (9.5%) | 3 (7.0%) | 17 (10.1%) | |
| Hispanic | 156 (73.9%) | 36 (83.7%) | 120 (71.4%) | |
| Asian | 13 (6.2%) | 3 (7.0%) | 10 (6.0%) | |
| BMI (kg/m2) | 35.6 (±7.4) | 36.9 (±8.0) | 35.3 (±7.3) | 0.06 |
| <35 kg/m2 | 116 (55.0%) | 17 (39.5%) | 93 (55.4%) | |
| ≥35 kg/m2 | 95 (45.0%) | 26 (60.5%) | 75 (44.6%) | |
| Gravidity | 2 (0–13) | 2.5 (0–13) | 2 (0–9) | 0.89 |
| Multigravida | 158 (74.9%) | 31 (72.1%) | 128 (76.2%) | |
| Nulligravida | 53 (25.1%) | 12 (27.9%) | 40 (23.8%) | |
| Hypertension | 0.078 | |||
| No | 130 (61.6%) | 21 (48.8%) | 109 (64.9%) | |
| Yes | 81 (38.4%) | 22 (51.2%) | 59 (35.1%) | |
| Diabetes mellitus | 0.017 | |||
| No | 159 (75.4%) | 26 (60.5%) | 133 (79.2%) | |
| Yes | 52 (24.6%) | 17 (39.5%) | 35 (20.8%) | |
| Hypercholesterolemia | 0.084 | |||
| No | 181 (85.8%) | 33 (76.7%) | 148 (88.1%) | |
| Yes | 30 (14.2%) | 10 (23.3%) | 20 (11.9%) | |
| Time interval (days)a | 105 (4–2514) | 96 (18–1503) | 109.5 (4–2514) | 0.07 |
| <105 days | 104 (49.3%) | 23 (53.5%) | 81 (48.2%) | |
| ≥105 days | 107 (50.7%) | 20 (46.5%) | 87 (51.8%) | |
| Hormonal therapy | 0.024c | |||
| No | 108 (51.2%) | 29 (69.0%) | 79 (48.2%) | |
| Yes (any) | 98 (46.4%) | 13 (31.0%) | 85 (51.8%) | |
| Megestrol acetate | 60 (28.4%) | 7 (53.8%) | 53 (62.4%) | |
| Medroxyprogesterone | 19 (9.0%) | 2 (15.4%) | 5 (5.9%) | |
| Levonorgestrel-releasing IUD | 8 (3.8%) | 0 | 8 (9.4%) | |
| Norethindrone acetate | 7 (3.3%) | 3 (23.1%) | 16 (18.8%) | |
| Otherb | 4 (1.9%) | 1 (7.7%) | 3 (3.5%) | |
| Unknown | 5 (2.4%) | 1 (2.3%) | 4 (2.4%) | |
| Initial hyperplasia type | 0.032 | |||
| SH/SAH | 24 (11.4%) | 1 (2.3%) | 23 (13.7%) | |
| CH | 58 (27.5%) | 9 (20.9%) | 49 (29.2%) | |
| CAH | 129 (61.1%) | 33 (76.7%) | 96 (57.1%) | |
| Methods of endometrial sampling | 0.76 | |||
| Pipelle | 142 (67.3%) | 29 (67.4%) | 113 (67.3%) | |
| Vacuum aspiration | 53 (25.1%) | 10 (23.3%) | 43 (25.6%) | |
| Dilation and curettage | 14 (6.6%) | 3 (7.0%) | 11 (6.5%) | |
| Unknown | 2 (0.9%) | 1 (2.3%) | 1 (0.6%) | |
Student t test (age), Mann–Whitney U test (time interval), chi-square test (initial hyperplasia type), or Fisher’s exact test (remaining variables) for p-values. Significant p-values are emboldened. Abbreviations: BMI, body mass index; IUD, intrauterine device; CAH, complex hyperplasia with atypia; CH, complex hyperplasia without atypia; SAH, simple hyperplasia with atypia; and SH, simple hyperplasia without atypia.
Time interval between endometrial biopsy for endometrial hyperplasia and hysterectomy.
Combination therapy.
Hormone therapy yes versus no for p-value. Mean (±SD), median (range), or number (%) is shown.
Among 211 cases of biopsy-proven endometrial hyperplasia, there were 43 (20.4%, 95% CI 19.0–32.2) cases of endometrial cancer diagnosed in the subsequent hysterectomy. The characteristics of women with concurrent endometrial cancer are shown in Table 2. The majority of concurrent endometrial cancer cases had endometrioid histology (90.7%), grade 1 tumors (86.0%), and stage IA disease (83.7%). These concurrent endometrial cancers in biopsy-proven endometrial hyperplasia cases tended not to express aggressive tumor behavior (median depth of myometrial tumor invasion, 8%; and prevalence of lymphovascular space invasion, 7%) and had a good overall prognosis. Complete surgical staging was infrequently performed among patients with concurrent endometrial carcinoma (pelvic and para-aortic lymphadenectomy, 11.6% and 4.7%, respectively, Table 2). With a median follow-up time of 48.6 months, there was no observed incidence of recurrence or death due to endometrial cancer.
Table 2.
Characteristics of endometrial cancer concurrent with endometrial hyperplasia.
| n = 43 | |
|---|---|
| Histology subtype | |
| Endometrioid | 39 (90.7%) |
| Endometrioid + mucinous | 3 (7.0%) |
| Endometrioid + serous | 1 (2.3%) |
| Grade | |
| 1 | 37 (86.0%) |
| 2 | 3 (7.0%) |
| 3 | 3 (7.0%) |
| Stage | |
| IA | 36 (83.7%) |
| II | 4 (9.3%) |
| IIIA | 1 (2.3%) |
| IIIC1 | 2 (4.7%) |
| Myometrial tumor invasion (%) | 8 (0–100) |
| None | 13 (30.2%) |
| Inner third | 24 (55.8%) |
| Middle third | 3 (7.0%) |
| Outer third | 3 (7.0%) |
| LVSI | |
| No | 40 (93.0%) |
| Yes | 3 (7.0%) |
| Lymphadenectomy | |
| Pelvic | 5 (11.6%) |
| Para-aortic | 2 (4.7%) |
| Any | 6 (14.0%) |
| Survival outcome | |
| Follow-up time (months) | 48.6 (1.8–141.9) |
| Recurrence | 0 |
| Die of cancer | 0 |
Number (%) or median (range) is shown. Abbreviations: LVSI, lymphovascular space invasion.
Clinico-pathological variables were compared between concurrent endometrial cancer cases and pure endometrial hyperplasia cases in univariate analysis (Table 1). When compared to patients who did not have concurrent endometrial cancer, patients with concurrent endometrial cancer were older (mean age, 49.4 versus 44.1, p = 0.001 ) and were more obese (Proportion of BMI ≥ 35 kg/m2, 60.5% versus 44.6%, p = 0.06) although it did not reach statistical significance. Diabetes mellitus was significantly associated with concurrent endometrial cancer (39.5% versus 20.8%, p = 0.017). Use of hormonal therapy (any type) for endometrial hyperplasia was significantly associated with decreased prevalence of concurrent endometrial cancer (31.0% versus 51.8%, p = 0.024). Among those who received hormonal treatment for endometrial hyperplasia (n = 98), the prevalence of concurrent endometrial cancer did not differ across the treatment agents (p = 0.50). Time interval between the date of initial endometrial biopsy showing endometrial hyperplasia and the date of hysterectomy was shorter in concurrent carcinoma group compared to pure hyperplasia group although it did not reach statistical significance (median time, 96 versus 109.5 days, p = 0.07). The type of endometrial hyperplasia in the initial endometrial biopsy was significantly associated with the risk of concurrent endometrial cancer: SH/SAH 2.3%, CH 20.9%, and CAH 76.7%, p = 0.032.
Multivariate analysis was performed to determine independent risk factors associated with concurrent endometrial cancer (Table 3). There were four independent risk factors identified in the analysis. These include older age (age 40–59, OR 3.07, 95% CI 1.18–7.97, p = 0.021; and age ≥60, OR 6.65,95% CI 1.75–25.3, p = 0.005), larger body habitus (BMI ≥ 35 kg/m2, OR 2.32, 95% CI 1.09–4.93, p = 0.029), presence of diabetes mellitus (OR 2.51, 95% CI 1.16–5.39, p = 0.019), and presence of CAH (OR 9.01, 95% CI 1.09–74.6, p = 0.042). The relative contribution of these four independent risk factors was examined to predict risk of concurrent endometrial cancer (Table 4). The risk of concurrent endometrial carcinoma rose dramatically with an increasing number of risk factors (none 0%, 1 risk factor 7.0%, 2 risk factors 17.6%, 3 risk factors 35.8%, and 4 risk factors 45.5%, p < 0.001). The combination of risk factors associated with the highest prevalence of concurrent endometrial cancer was the patient group with BMI ≥ 35 kg/m2, diabetes mellitus, and CAH (50.0%).
Table 3.
Independent risk factors for concurrent endometrial cancer in patients with endometrial hyperplasia.
| No. | Univariate |
Multivariatea |
P-value | |
|---|---|---|---|---|
| OR (95% CI) | OR (95% CI) | |||
| Age | ||||
| <40 | 62 | 1 | 1 | |
| 40–59 | 133 | 2.19 (0.90–5.32) | 3.07 (1.18–7.97) | 0.021 |
| ≥60 | 16 | 6.11 (1.73–21.6) | 6.65 (1.75–25.3) | 0.005 |
| BMI | ||||
| <35 kg/m2 | 110 | 1 | 1 | |
| ≥35 kg/m2 | 101 | 1.94 (0.99–3.84) | 2.32 (1.09–4.93) | 0.029 |
| Diabetes mellitus | ||||
| No | 159 | 1 | 1 | |
| Yes | 52 | 2.49 (1.21–5.08) | 2.51 (1.16–5.39) | 0.019 |
| Initial hyperplasia type | ||||
| SH/SAH | 24 | 1 | 1 | |
| CH | 58 | 4.22 (0.51–35.4) | 5.12 (0.57–46.2) | 0.15 |
| CAH | 129 | 7.91 (1.03–60.9) | 9.01 (1.09–74.6) | 0.042 |
Significant p-values are emboldened. Abbreviations: No., number; OR, odds ratio; CI, confidence interval; BMI, body mass index; SH, simple hyperplasia without atypia; SAH, simple hyperplasia with atypia; CH, complex hyperplasia without atypia; and CAH, complex hyperplasia with atypia.
Binary logistic regression model for multivariate analysis (entered age, BMI, diabetes mellitus, and initial hyperplasia type).
Table 4.
Extent and pattern of risk factors and prevalence of concurrent endometrial cancer in patients with endometrial hyperplasia.
| No. | Concurrent cancer (%) | P-value | |
|---|---|---|---|
| Total | 211 | 43 (20.4%) | |
| Any risk factora | <0.001 | ||
| None | 5 | 0 | |
| 1 risk factor | 57 | 4 (7.0%) | |
| 2 risk factors | 85 | 15 (17.6%) | |
| 3 risk factors | 53 | 19 (35.8%) | |
| 4 risk factors | 11 | 5 (45.5%) | |
| Combination patterns | 0.032 | ||
| Age alone | 34 | 2 (5.9%) | |
| BMI alone | 12 | 1 (8.3%) | |
| DM alone | 3 | 0 | |
| CAH alone | 8 | 1 (12.5%) | |
| Age + BMI | 13 | 4 (30.8%) | |
| Age + DM | 5 | 1 (20.0%) | |
| Age + CAH | 39 | 8 (20.5%) | |
| BMI + DM | 5 | 1 (20.0%) | |
| BMI + CAH | 20 | 1 (5.0%) | |
| DM + CAH | 3 | 0 | |
| Age + BMI + DM | 5 | 1 (20.0%) | |
| Age + BMI + CAH | 28 | 9 (32.1%) | |
| Age + DM + CAH | 14 | 6 (42.9%) | |
| BMI + DM + CAH | 6 | 3 (50.0%) | |
| Age + BMI + DM + CAH | 11 | 5 (45.5%) | |
Number (%) is shown. Chi-square test for p-values. Significant p-values are emboldened. Abbreviations: No., number; Age, age ≥ 40; BMI, body mass index ≥ 35 kg/m2; DM, diabetes mellitus; and CAH, complex hyperplasia with atypia.
Risk factor included: age ≥ 40, BMI ≥ 35, diabetes mellitus, and complex hyperplasia with atypia.
To examine the utility of these risk factors on hormonal treatment, the study population was stratified by the use of hormonal treatment (yes n = 108, and no n = 98) in post-hoc analysis (Table 5 and S1). Without hormonal treatment, these risk factors were significantly associated with increased risk of concurrent endometrial cancer (none 0%, 1 risk factor 8.3%, 2 risk factors 18.2%, 3 risk factors 45.8%, and 4 risk factors 80%, p < 0.001). However, among cases that received hormonal treatment, the number of risk factors was not associated with increased risk of concurrent endometrial cancer (none 0%, 1 risk factor 6.1%, 2 risk factors 18.4%, 3 risk factors 15.8%, and 4 risk factors 16.7%, p = 0.59). This effect was especially prominent among those with 3 or more risk factors (prevalence of concurrent endometrial cancer for hormonal treatment versus no hormonal treatment: 3 risk factors, 15.8% versus 45.8%; and 4 risk factors, 16.7% versus 80%).
Table 5.
Hormonal therapy and prevalence of concurrent endometrial cancer in patients with endometrial hyperplasia.
| Hormonal treatment (−) |
Hormonal treatment (+) |
|||||
|---|---|---|---|---|---|---|
| No. | Concurrent cancer (%) | P-value | No. | Concurrent cancer (%) | P-value | |
| Total | 108 | 29 (26.9%) | 98 | 13 (13.3%) | ||
| Any risk factora | 0.001 | 0.59 | ||||
| None | 2 | 0 | 2 | 0 | ||
| 1 risk factor | 24 | 2 (8.3%) | 33 | 2 (6.1%) | ||
| 2 risk factors | 44 | 8 (18.2%) | 38 | 7 (18.4%) | ||
| 3 risk factors | 33 | 15 (45.5%) | 19 | 3 (15.8%) | ||
| 4 risk factors | 5 | 4 (80%) | 6 | 1 (16.7%) | ||
| Combination patterns | 0.019 | 0.24 | ||||
| Age alone | 15 | 1 (6.7%) | 19 | 1 (5.3%) | ||
| BMI alone | 5 | 1 (20.0%) | 7 | 0 | ||
| DM alone | 1 | 0 | 2 | 0 | ||
| CAH alone | 3 | 0 | 5 | 1 (20.0%) | ||
| Age + BMI | 8 | 1 (12.5%) | 5 | 3 (60.0%) | ||
| Age + DM | 3 | 1 (33.3%) | 2 | 0 | ||
| Age + CAH | 24 | 5 (20.8%) | 13 | 3 (23.1%) | ||
| BMI + DM | 1 | 0 | 4 | 1 (25.0%) | ||
| BMI + CAH | 6 | 1 (16.7%) | 13 | 0 | ||
| DM + CAH | 2 | 0 | 1 | 0 | ||
| Age + BMI + DM | 2 | 0 | 3 | 1 (33.3%) | ||
| Age + BMI + CAH | 16 | 6 (37.5%) | 11 | 2 (18.2%) | ||
| Age + DM + CAH | 12 | 6 (50.0%) | 2 | 0 | ||
| BMI + DM + CAH | 3 | 3 (100%) | 3 | 0 | ||
| Age + BMI + DM + CAH | 5 | 4 (80.0%) | 6 | 1 (16.7%) | ||
Analysis of 206 patients with known information for hormonal treatment for endometrial hyperplasia. Number (%) is shown. Chi-square test for p-values. Significant p-values are emboldened. Abbreviations: No., number; Age, age ≥ 40; BMI, body mass index ≥ 35 kg/m2; DM, diabetes mellitus; and CAH, complex hyperplasia with atypia.
Risk factor included: age ≥ 40, BMI ≥ 35 kg/m2, diabetes mellitus, and complex hyperplasia with atypia.
Similarly, the study population was stratified by presence or absence of CAH (Table 6 and S2). When 129 patients with CAH in endometrial biopsy were examined, presence of additional risk factors were increasingly associated with prevalence of concurrent endometrial cancer (no additional risk factor 12.5%, 1 additional risk factor 14.5%, 2 additional risk factors 29.0%, and 3 additional risk factors 45.5%, p = 0.014). In specific combination patterns of additional risk factors, presence of ≥2 additional risk factors to CAH had significantly high concurrent endometrial carcinoma rates (32.1–50%). Conversely, there was no statistical correlation between the extent of additional risk factors and risk of concurrent carcinoma in non-CAH cases.
Table 6.
Complex hyperplasia with atypia and prevalence of concurrent endometrial cancer.
| CAH (−) in biopsy |
CAH (+) in biopsy |
|||||
|---|---|---|---|---|---|---|
| No. | Concurrent cancer (%) | P-value | No. | Concurrent cancer (%) | P-value | |
| Total | 82 | 10 (12.2%) | 129 | 33 (25.6%) | ||
| Any risk factora | 0.078 | 0.014 | ||||
| None | 5 | 0 | 8 | 1 (12.5%) | ||
| 1 risk factor | 49 | 3 (6.1%) | 62 | 9 (14.5%) | ||
| 2 risk factors | 23 | 6 (26.1%) | 48 | 18 (29.0%) | ||
| 3 risk factors | 5 | 1 (20%) | 11 | 5 (45.5%) | ||
| Combination patterns | 0.37 | 0.049 | ||||
| Age alone | 34 | 2 (5.9%) | 39 | 8 (25.8%) | ||
| BMI alone | 12 | 1 (8.3%) | 20 | 1 (5%) | ||
| DM alone | 3 | 0 | 3 | 0 | ||
| Age + BMI | 13 | 4 (30.8%) | 28 | 9 (32.1%) | ||
| Age + DM | 5 | 1 (20%) | 14 | 6 (42.9%) | ||
| BMI + DM | 5 | 1 (20%) | 6 | 3 (50%) | ||
| Age + BMI + DM | 5 | 1 (20%) | 11 | 5 (45.5%) | ||
Number (%) is shown. Chi-square test for p-values. Significant p-values are emboldened. Abbreviations: No., number; Age, age ≥ 40; BMI, body mass index ≥ 35 kg/m2; CAH, complex hyperplasia with atypia: and DM, diabetes mellitus.
Risk factor included: age ≥ 40, BMI ≥ 35 kg/m2, and diabetes mellitus.
4. Discussion
The key finding of this study is that older age, obesity, diabetes mellitus, and complex hyperplasia with atypia are all independent risk factors that can predict concurrent endometrial cancer in patients with endometrial hyperplasia. These factors proved to be additive, with the presence of an increasing number of risk factors conferring a higher risk of malignancy, such that there was a 45% risk of malignancy when all four factors were present. Because endometrial sampling cannot be relied upon to detect the presence of endometrial carcinoma in women with endometrial hyperplasia, the identification of risk factors that increase the risk of tumor progression from endometrial hyperplasia to carcinoma is of significant clinical value.
Cancer is primarily a disease of the elderly, with the overall incidence of malignancy in the United States increasing dramatically from 102 cases per 100,000 for those under 50 years of age, to 1374 cases per 100,000 for those 50 years and older [13]. In our study, older age was sequentially associated with increased risk of harboring concurrent endometrial cancer (Table 3). In endometrial cancer, older age is associated with more aggressive tumors and decreased survival [14,15]. The physiologic basis for worse outcomes with increasing age may be related to decreased immune function over time, otherwise known as immunosenescence. Compared to their younger counterparts, the elderly demonstrate immune system impairment in the process of apoptosis, and also an overexpression of pro-inflammatory cytokines (such as IL-10, IL-6 and TGF-β) [16].
Obesity is another characteristic that was associated with concurrent endometrial carcinoma in our study. The effect of obesity on the risk of endometrial cancer is well described. Each increase in BMI of 5 kg/m2 confers a 1.6-fold increased risk in developing endometrial cancer [17]. Severe and superobesity have also been shown to dramatically increase the risk of cancer-related death in patients with endometrial cancer [18]. The biologic mechanism for these effects is multimodal, and is especially pronounced in postmenopausal women, where adipose tissue represents the primary source of circulating estrogen [19]. Estrogen functions as a growth factor for endometrial tissue, causing the transcription of genes that lead to endometrial proliferation [20]. It also acts directly on endometrial tissue, leading to the activation of the PI3K/MAPK pathways [21]. In our study, age and BMI had an inverse correlation. This implies that younger patients were more likely to be obese. Because obesity has been strongly correlated with the risk of endometrial hyperplasia/carcinoma, our study population had a large number of young women (mean age, 45.2). A previous study has shown that lower BMI was associated with a higher pregnancy rate in women who underwent fertility-sparing management for endometrial hyperplasia/carcinoma [22]. This study endorses that weight reduction is an important consideration to maximize the effects of progestin therapy in obese women with endometrial hyperplasia/carcinoma.
Diabetes mellitus, another independent risk factor in our study, is often related to obesity due to insulin resistance. A hyperinsulinic state leads to greater insulin like growth factor 1 (IGF-1) bioavailability and resultant activation of pro-proliferative kinase pathways [23]. A large population-based study of over 40,000 patients with any type of cancer revealed significantly increased mortality rates for patients with pre-existing diabetes mellitus for at least two years prior to the cancer diagnosis [24]. In a study of 1644 patients with endometrial cancer, the 5-year survival rate for patients with diabetes mellitus was 16% lower than those without diabetes mellitus [25]. Therefore, maintaining adequate glycemic control may be a factor that would reduce the risk of endometrial hyperplasia progression to carcinoma.
Our study is among the first to demonstrate that there are risk factors for concurrent carcinoma besides the histologic severity of endometrial hyperplasia. In a study that was limited by the use of hormonal therapy by some subjects and wide variation in follow-up (1–26.7 years), progression to endometrial carcinoma was found in 29% and 8% of patients with CAH and SAH, respectively [3]. The risk for progression to malignancy in patients with non-atypical hyperplasia was 1.6%. Our findings are consistent with these results and demonstrate a histologic diagnosis of CAH additively increased the risk of malignancy and was the strongest risk factor for concurrent endometrial carcinoma.
Women with endometrial hyperplasia who desire future fertility and wish to defer surgical management can be treated with various formulations of progestin therapy. These patients are commonly instructed to attempt pregnancy shortly after the lesion has been cleared for a specified amount of time. In a recent systematic review, 66–81% of patients with CAH demonstrated resolution of nuclear atypia after progestin therapy [26–28]. Our results also reveal the effectiveness of hormonal treatment, with a 50% decreased rate of concurrent endometrial cancer among patients on progestin therapy, irrespective of the agent (26.9% versus 13.3%). A more dramatic protective effect against concurrent malignancy was seen when comparing patients with three or four of the risk factors identified in this study (Table 6). Treatment planning stratifying by these factors would be of future interest. In a view of the literature, megestrol acetate was found to have a higher response rate as opposed to other progestins in the treatment of endometrial hyperplasia/carcinoma, possibly due to higher oral bioavailability [28]. However, our study did not show a difference in the likelihood of concurrent endometrial carcinoma across the hormonal treatment types including megestrol acetate. Given that our study population was predominantly obese (75.8%), the association of excess adiposity and megestrol acetate effects merits further investigation to determine the most effective treatment choice of progestin therapy in obese women with endometrial hyperplasia/carcinoma.
The strengths of the study include the large number of consecutive subjects included in the study over a period of 11 years, over which the management strategy for endometrial hyperplasia remained relatively unchanged. In addition, the power of the study was adequate as >99% by using Cox and Snell pseudo R-square (0.12) in the multivariate model with 4 predictors in a sample size of 211 and an alpha level of 0.05. The limitation of this study is its retrospective nature that may miss confounding factors. For instance, our study did not answer if the tumors progressed from endometrial hyperplasia to carcinoma in the time interval between endometrial biopsy and hysterectomy. However, biopsy-to-hysterectomy time interval was not associated with concurrent carcinoma (median time, 3.5 months), and type I endometrial cancer generally progresses slowly [9]. Therefore, it is likely that carcinoma had co-existed with hyperplasia at the time of endometrial hyperplasia diagnosis. In addition, the exact indications for treatment allocation to progestin therapy as opposed to surgery could not be ascertained. Weaknesses of this study were that central pathology review for the biopsy and the hysterectomy specimens was not performed. Distinguishing low-grade endometrial carcinoma over CAH can be challenging with possible disagreement between the pathologists [29,30]. However, at our institution, all biopsy and hysterectomy slides were reviewed by a pathologist with special expertise in gynecologic pathology at the time of diagnosis. Cases that were deemed as unclear were reviewed by a second gynecologic pathologist. Finally, this was a single institution study with an ethnically dominant patient population to Hispanic that may lack generalizability in other populations.
In summary, prediction of concurrent endometrial carcinoma will be useful to both physicians and patients with endometrial hyperplasia, especially when considering fertility preservation and assessing the risks of conservative medical management with hormonal therapy. In particular, concrete response rates can be quoted to patients based on the extent of risk factors. Recognizing these risk factors will also assist general gynecologists in estimating the risk of malignancy and the possible need to refer or collaborate with a gynecologic oncologist in the surgical planning of endometrial hyperplasia. Generally, ovarian preservation is acceptable when performing hysterectomy for the indication of endometrial hyperplasia. However, if the probability of concurrent endometrial carcinoma is high, intraoperative frozen section should be considered so that an indication for comprehensive surgical staging can be assessed.
Supplementary Material
HIGHLIGHTS.
Older age, larger body habitus, diabetes mellitus, and complex hyperplasia with atypia were independent risk factors predicting concurrent endometrial cancer.
Up to 45.5% of endometrial hyperplasia patients can harbor concurrent endometrial cancer if multiple risk factors are present.
Hormonal treatment for endometrial hyperplasia may be beneficial among patients expressing 3 or more risk factors.
Acknowledgment
The study is supported by Ensign Endowment for Ovarian Cancer Research (KM and LDR). The authors thank Neisha R. Opper, MPH for the statistical consultation.
Footnotes
Disclosure
The authors did not report any potential conflicts of interest.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.ygyno.2015.07.108.
Part of the study was presented at the 46th Annual Meeting of Society of Gynecologic Oncologists, Chicago IL, March 28–31,2015.
References
- [1].Siegel RL, Miller KD, Jemal A, Cancer statistics, 2015, CA Cancer J. Clin. 65 (2015) 5–29 [DOI] [PubMed] [Google Scholar]
- [2].Trimble CL, Kauderer J, Zaino R, Silverberg S, Lim PC, Burke JJ 2nd, et al. , Concurrent endometrial carcinoma in women with a biopsy diagnosis of atypical endometrial hyperplasia: a Gynecologic Oncology Group study, Cancer 106 (2006) 812–819. [DOI] [PubMed] [Google Scholar]
- [3].Kurman RJ, Kaminski PF, Norris HJ, The behavior of endometrial hyperplasia. A long-term study of “untreated” hyperplasia in 170 patients, Cancer 56 (1985) 403–412. [DOI] [PubMed] [Google Scholar]
- [4].Trimble CL, Method M, Leitao M, Lu K, Ioffe O, Hampton M, Higgins R, Zaino R, Mutter GL, Management of endometrial precancers, Obstet. Gynecol. 120 (2012) 1160–1175. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Costales AB, Schmeler KM, Broaddus R, Soliman PT, Westin SN, Ramirez PT, et al. , Clinically significant endometrial cancer risk following a diagnosis of complex atypical hyperplasia, Gynecol. Oncol. 135 (2014) 451–454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].STROBE guideline at http://www.strobe-statement.org/ (accessed April 2015).
- [7].Matsuo K, Cahoon SS, Gualtieri M, Scannell CA, Jung CE, Takano T, et al. , Significance of adenomyosis on tumor progression and survival outcome of endometrial cancer, Ann. Surg. Oncol. 21 (2014) 4246–4255. [DOI] [PubMed] [Google Scholar]
- [8].Matsuo K, Gray MJ, Yang DY, Srivastava SA, Tripathi PB, Sonoda LA, et al. , The endoplasmic reticulum stress marker, glucose-regulated protein-78 (GRP78) in visceral adipocytes predicts endometrial cancer progression and patient survival, Gynecol. Oncol. 128 (2012) 552–559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Matsuo K, R Opper N, Ciccone MA, Garcia J, Tierney KE, Baba T, et al. , Time interval between endometrial biopsy and surgical staging for type I endometrial cancer: association between tumor characteristics and survival outcome, Obstet. Gynecol. 125 (2015)424–433. [DOI] [PubMed] [Google Scholar]
- [10].Matsuo K, Yessaian AA, Lin YG, Pham HQ, Muderspach LI, A Liebman H, et al. , Predictive model of venous thromboembolism in endometrial cancer, Gynecol. Oncol. 128 (2012) 544–551. [DOI] [PubMed] [Google Scholar]
- [11].Matsuo K, Gualtieri MR, Cahoon SS, Jung CE, Paulson RJ, Shoupe D, et al. , Surgical menopause and increased risk of nonalcoholic fatty liver disease in endometrial cancer, Menopause (2015) [Epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Pecorelli S, Revised FIGO staging for carcinoma of the vulva, cervix, and endometrium, Int.J. Gynaecol. Obstet. 105 (2009) 103–104. [DOI] [PubMed] [Google Scholar]
- [13].2011 SEER statistics, http://seer.cancer.gov/faststats/selections.php?series=age (accessed April 2015).
- [14].Goodwin JS, Samet JM, Key CR, Humble C, Kutvirt D, Hunt C, Stage at diagnosis of cancer varies with the age of the patient, J. Am. Geriatr. Soc. 34 (1986) 20–26. [DOI] [PubMed] [Google Scholar]
- [15].Lee NK, Cheung MK, Shin JY, Husain A, Teng NN, Berek JS, et al. , Prognostic factors for uterine cancer in reproductive-aged women, Obstet. Gynecol. 109 (2007) 655–662. [DOI] [PubMed] [Google Scholar]
- [16].Malaguarnera L, Cristaldi E, Malaguarnera M, The role of immunity in elderly cancer, Crit. Rev. Oncol. Hematol. 74 (2010) 40–60. [DOI] [PubMed] [Google Scholar]
- [17].Renehan AG, Tyson M, Egger M, Heller RF, Zwahlen M, Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies, Lancet 371 (2008) 569–578. [DOI] [PubMed] [Google Scholar]
- [18].Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ, Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults, N. Engl. J. Med. 348 (2003) 1625–1638 [DOI] [PubMed] [Google Scholar]
- [19].Cauley JA, Gutai JP, Kuller LH, LeDonne D, Powell JG, The epidemiology of serum sex hormones in postmenopausal women, Am. J. Epidemiol. 129 (1989) 1120–1131. [DOI] [PubMed] [Google Scholar]
- [20].Marino M, Galluzzo P, Ascenzi P, Estrogen signaling multiple pathways to impact gene transcription, Curr. Genomics 7 (2006) 497–508. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Marino M, Ascenzi P, Acconcia F, S-palmitoylation modulates estrogen receptor alpha localization and functions, Steroids 71 (2006) 298–303. [DOI] [PubMed] [Google Scholar]
- [22].Gonthier C, Walker F, Luton D, Yazbeck C, Madelenat P, Koskas M, Impact of obesity on the results of fertility-sparing management for atypical hyperplasia and grade 1 endometrial cancer, Gynecol. Oncol. 133 (2014) 33–37. [DOI] [PubMed] [Google Scholar]
- [23].Pollak M, Insulin and insulin-like growth factor signalling in neoplasia, Nat. Rev. Cancer 8 (2008) 915–928. [DOI] [PubMed] [Google Scholar]
- [24].Ranc K, Jorgensen ME, Friis S, Carstensen B, Mortality after cancer among patients with diabetes mellitus: effect of diabetes duration and treatment, Diabetologia 57 (2014) 927–934. [DOI] [PubMed] [Google Scholar]
- [25].Zanders MM, Boll D, van Steenbergen LN, van de Poll-Franse LV, Haak HR, Effect of diabetes on endometrial cancer recurrence and survival, Maturitas 74 (2013) 37–43. [DOI] [PubMed] [Google Scholar]
- [26].Gunderson CC, Fader AN, Carson KA, Bristow RE, Oncologic and reproductive outcomes with progestin therapy in women with endometrial hyperplasia and grade 1 adenocarcinoma: a systematic review, Gynecol. Oncol. 125 (2012) 477–482. [DOI] [PubMed] [Google Scholar]
- [27].Baker J, Obermair A, Gebski V, Janda M, Efficacy of oral or intrauterine device-delivered progestin in patients with complex endometrial hyperplasia with atypia or early endometrial adenocarcinoma: a meta-analysis and systematic review of the literature, Gynecol. Oncol. 125 (2012) 263–270. [DOI] [PubMed] [Google Scholar]
- [28].Koskas M, Uzan J, Luton D, Rouzier R, Darai E, Prognostic factors of oncologic and reproductive outcomes in fertility-sparing management of endometrial atypical hyperplasia and adenocarcinoma: systematic review and meta-analysis, Fertil. Steril. 101 (2014) 785–794. [DOI] [PubMed] [Google Scholar]
- [29].Silverberg SG, Problems in the differential diagnosis of endometrial hyperplasia and carcinoma, Mod. Pathol. 13 (2000) 309–327. [DOI] [PubMed] [Google Scholar]
- [30].Allison KH, Reed SD, Voigt LF, Jordan CD, Newton KM, Garcia RL, Diagnosing endometrial hyperplasia: why is it so difficult to agree? Am. J. Surg. Pathol. 32 (2008) 691–698. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.