Comparative effectiveness of hormonal therapies for preventing recurrence in endometriosis: a real-world retrospective cohort study with risk factor analysis

Abstract

Background

Endometriosis is a common chronic disease in women of reproductive age, and long-term postoperative medical management is a key strategy for preventing recurrence. Currently used clinical medications include dienogest (DNG), GnRH agonists (GnRH-a), combined oral contraceptives (COC), and the levonorgestrel-releasing intrauterine system (LNG-IUS). However, comparative effectiveness of different hormonal therapies for preventing recurrence in real-world clinical practice and the basis for individualised patient selection remain insufficient.

Objective

To systematically evaluate the efficacy and safety of DNG, GnRH-a, COC, and LNG-IUS in preventing postoperative recurrence of ovarian endometriomas; to analyse independent risk factors for postoperative recurrence, providing evidence-based support for individualised clinical treatment decisions.

Methods

A retrospective cohort study design was adopted. A total of 167 patients who underwent laparoscopic cystectomy at our hospital between January 2020 and January 2022, had a postoperative pathological diagnosis, and received sequential GnRH-a maintenance therapy were enrolled. According to the sequential maintenance regimen, patients were divided into three groups: GnRH-a + DNG group (n = 61), GnRH-a + COC group (n = 64), and GnRH-a + LNG-IUS group (n = 42). The primary outcome was the recurrence rate within 3 years after surgery. Secondary outcomes included menstrual bleeding profiles, recurrent cyst diameter, and adverse drug reactions. Cumulative recurrence rates were calculated using the Kaplan‑Meier method, and intergroup comparisons were performed using the log‑rank test. Multivariate logistic regression analysis was used to identify independent risk factors for postoperative recurrence.

Results

There were no statistically significant differences in baseline data among the three groups (P > 0.05), indicating comparability. The 3‑year cumulative recurrence rate in the GnRH-a + DNG group was 19.67% (12/61), significantly lower than that in the GnRH-a + LNG-IUS group (45.24%, 19/42; P = 0.003). The recurrence rate in the GnRH-a + DNG group was also lower than that in the GnRH-a + COC group (34.38%, 22/64), although this difference did not reach statistical significance (P = 0.053). No significant differences were observed among the three groups in mean daily menstrual blood loss, incidence of dysmenorrhoea, or menstrual cycle length (P > 0.05). However, the incidence of spotting in the LNG-IUS group (52.38%) was significantly higher than that in the DNG group (24.59%) and the COC group (12.50%, P < 0.001). There were no statistically significant differences in the total incidence of adverse drug reactions (13.11%, 14.06%, 11.90%) or recurrent cyst diameter among the groups (P > 0.05). Multivariate logistic regression analysis suggested that higher dysmenorrhea VAS score (OR = 1.376), history of pelvic procedures (OR = 1.483), and r-AFS stage IV (OR = 2.676) were independent risk factors for postoperative recurrence (all P < 0.05), while older age at surgery was a protective factor (OR = 0.891) (P < 0.05).

Conclusion

Among sequential GnRH-a maintenance regimens, DNG was associated with a lower recurrence rate than LNG-IUS in preventing 3‑year recurrence after laparoscopic cystectomy in this cohort. Although the recurrence rate in the DNG group was lower than that in the COC group, the difference did not reach statistical significance, indicating only a trend toward superiority. All three regimens have a favourable overall safety profile, but the LNG-IUS group has a higher incidence of spotting. Severe dysmenorrhoea, previous pelvic operation history, and r-AFS stage IV are independent risk factors for postoperative recurrence, whereas older age at surgery has a protective effect.

Introduction

Endometriosis is an oestrogen-dependent chronic inflammatory disease characterised by the presence of endometrial tissue (glands and stroma) with growth capacity outside the mucosal lining of the uterine cavity and the myometrium [1]. The disease predominantly affects women of reproductive age, with a global prevalence of approximately 10% and as high as 50% among infertile women [2]. Endometriosis has three recognised major pathological phenotypes, including superficial peritoneal lesions, ovarian endometriomas, and deep infiltrating endometriosis [3]. Among these, ovarian endometriosis is one of the most common pathological types, accounting for approximately 17%–44% of patients with endometriosis. The symptoms it causes, such as pelvic pain, dyspareunia, dysmenorrhoea, and infertility, seriously affect patients’ physical and mental health and quality of life [4].

Laparoscopic ovarian cystectomy is currently the preferred surgical approach for treating ovarian endometriosis, as it effectively removes lesions, relieves pain, and improves fertility prognosis [5]. However, surgery itself does not cure endometriosis, and postoperative recurrence remains a major challenge for both clinicians and patients. Existing studies have shown that the 5-year cumulative recurrence rate after conservative surgery for ovarian endometriosis can be as high as 50%, and each recurrence may lead to further ovarian tissue damage, exacerbation of decreased ovarian reserve, and even an increased risk of malignant transformation [6]. Therefore, exploring effective long-term postoperative medical management strategies to prevent recurrence has become a hot topic and a difficult issue in the field of endometriosis treatment. Emerging therapeutic options targeting inflammation, along with lifestyle and dietary modifications, are gaining attention as complementary strategies to hormonal therapy [7, 8]. Furthermore, translational animal models continue to provide critical insights into disease mechanisms and preclinical evaluation of new treatments [9].

The theoretical basis of postoperative medical therapy lies in the oestrogen-dependent pathophysiological mechanism of endometriosis. The survival and proliferation of ectopic endometrial tissue depend on continuous oestrogen stimulation. Inhibiting endogenous oestrogen production or blocking the effect of oestrogen on the ectopic endometrium through medication can induce atrophy of ectopic lesions, thereby achieving the goal of preventing recurrence [10, 11]. Currently, the commonly used postoperative hormonal therapies mainly include GnRH agonists (GnRH-a), dienogest (DNG), combined oral contraceptives (COC), and the levonorgestrel-releasing intrauterine system (LNG-IUS) [12]. These agents inhibit ovarian function, reduce oestrogen levels, or directly act on ectopic endometrial tissue through different mechanisms, theoretically possessing the potential to prevent postoperative recurrence [13]. GnRH-a downregulates pituitary GnRH receptors and inhibits the secretion of luteinising hormone (LH) and follicle-stimulating hormone (FSH), thereby reducing circulating oestradiol concentrations to postmenopausal levels within one month of treatment, creating a profound hypoestrogenic environment that suppresses the growth of ectopic lesions [14]. However, the severe hypoestrogenic state induced by long-term GnRH-a use can cause perimenopausal‑like adverse effects, including hot flushes, night sweats, mood swings, and bone loss, which limits the feasibility of its long-term use as monotherapy [15]. Therefore, a strategy of short-term GnRH-a treatment followed by sequential long-term maintenance therapy with other agents is often adopted in clinical practice, aiming to consolidate the surgical outcome while reducing adverse effects and prolonging the recurrence-free interval [16, 17].

DNG is a 19-nortestosterone-derived potent progestin with a unique dual central and peripheral mechanism of action. At the central level, DNG moderately inhibits ovulation by suppressing the hypothalamic-pituitary-ovarian axis, maintaining oestrogen levels within an “oestrogen threshold” that is sufficient to avoid oestrogen deficiency-related adverse effects while effectively inhibiting ectopic endometrial proliferation [18]. At the peripheral level, DNG acts directly on endometriotic lesions, inducing decidualisation and subsequent atrophy of ectopic endometrial tissue through progestin receptor-mediated anti-inflammatory, anti-proliferative, and pro-apoptotic effects [19, 20]. As a first-line medical therapy for endometriosis, COC offers advantages such as ease of use, high accessibility, relatively low cost, and the ability to regulate menstrual cycles, making it a preferred option for many patients [21]. However, the oestrogen component in COC may cause adverse effects including weight gain, breast tenderness, mood swings, and an increased risk of thrombosis. Moreover, COC use is contraindicated in patients with hypertension, smoking habits, or risk factors for thromboembolism, which limits its widespread application to some extent [22]. LNG-IUS continuously releases levonorgestrel locally within the uterine cavity, directly acting on the endometrium to exert a potent progestogenic effect, inhibiting endometrial proliferation and inducing endometrial atrophy [23]. The advantages of LNG-IUS include local drug delivery, few systemic adverse effects, a single placement that remains effective for up to 5 years, and good patient compliance [24]. However, the main limitation of LNG-IUS is its local mode of action, which may offer relatively limited systemic suppressive effects on deep pelvic lesions such as ovarian endometriomas [25].

Although all the above-mentioned drugs theoretically have the potential to prevent postoperative recurrence of endometriosis, comparative effectiveness data among different drugs in real-world clinical practice remain insufficient. Existing studies have largely focused on evaluating the efficacy of a single drug versus placebo or surgery, or direct comparisons between two drugs, whereas systematic comparisons of different sequential maintenance regimens following GnRH-a treatment are relatively scarce. In particular, dienogest, as a newly emerging agent that has attracted considerable attention for long-term management of endometriosis in recent years, lacks high-quality evidence‑based medical data derived from real-world studies regarding its recurrence prevention effectiveness and safety profile compared with COC and LNG-IUS within a GnRH-a sequential treatment framework. Furthermore, analysis of risk factors for postoperative recurrence of endometriosis is also a critical component for achieving precise individualised treatment. Using real-world clinical data and adopting a retrospective cohort study design, this study enrolled 167 patients who underwent laparoscopic cystectomy for ovarian endometriosis followed by different sequential maintenance regimens after GnRH‑a treatment. We systematically compared the efficacy of three regimens—GnRH‑a + DNG, GnRH‑a + COC, and GnRH‑a + LNG‑IUS—in preventing recurrence within three years after surgery and analysed independent risk factors for recurrence. This study aims to provide real-world evidence‑based support for clinicians in formulating individualised long‑term postoperative management strategies, ultimately improving the long‑term prognosis and quality of life of patients with endometriosis.

Methods

Study design and population

This study adopted a retrospective cohort study design and, based on real-world clinical data, systematically compared the efficacy of three sequential GnRH‑a maintenance regimens in preventing postoperative recurrence of ovarian endometriomas and analysed independent risk factors for postoperative recurrence. The study population was derived from patients who underwent laparoscopic cystectomy for ovarian endometriosis in the Department of Gynecology of our hospital between January 2020 and January 2022. The enrollment period for this retrospective study was from January 2023 to January 2025, during which we identified eligible patients from the electronic medical record system. By the time of data extraction, all enrolled patients had completed at least 36 months of postoperative follow‑up, ensuring that the 3‑year follow‑up was fully accomplished. Data were extracted and screened from the hospital’s electronic medical record system and outpatient follow‑up database. All enrolled patients received short‑term GnRH‑a treatment followed by sequential long‑term maintenance therapy with different drugs after surgery. This retrospective cohort study adhered to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guideline for reporting observational research. The detailed study flow chart is shown in Fig. 1. Briefly, a total of 312 patients who underwent laparoscopic cystectomy for ovarian endometriosis between January 2020 and January 2022 were initially screened from the electronic medical record system. After applying the inclusion criteria and exclusion criteria, 167 patients were finally enrolled. These patients were then divided into three groups according to the sequential maintenance regimen: GnRH‑a + DNG (n = 61), GnRH‑a + COC (n = 64), and GnRH‑a + LNG‑IUS (n = 42).

Fig. 1.

Study flow diagram

Inclusion criteria

• ① Patients with a preliminary diagnosis of ovarian endometriosis based on symptoms, signs, preoperative imaging, and serological examinations, confirmed by intraoperative laparoscopic exploration and postoperative histopathological diagnosis; ② All patients underwent laparoscopic ovarian endometrioma cystectomy performed by a surgeon with an associate senior or higher professional title; ③ Postoperative treatment with GnRH-a for 6 injections, followed by sequential long-term maintenance therapy with one of DNG, COC, or LNG-IUS, with medication duration ≥ 6 months; ④ Complete postoperative follow-up data, including gynaecological ultrasound examination and clinical symptom assessment at least every 3 months, with a total follow-up duration ≥ 36 months or until recurrence event occurs; ⑤ Age 21–45 years and of reproductive age.

Exclusion criteria

• ① Preoperative or postoperative pathological confirmation of concomitant ovarian malignancy, borderline tumour, benign ovarian tumour, or other histopathological types of endometriosis; ② Patients who underwent non-conservative surgery such as oophorectomy, hysterectomy, or extensive lesion resection during the operation; ③ Patients who discontinued medication or changed the drug regimen on their own during the follow-up period; ④ Short postoperative follow-up duration or missing key clinical data; ⑤ Pre-existing severe cardiac, hepatic, or renal dysfunction, uncontrolled hypertension, history of thromboembolic disease, or known contraindications to progestogen or oestrogen use; ⑥ Pregnancy or lactation during the postoperative follow-up period.

Ethics statement

This study was a retrospective cohort study, and all data were derived from electronic medical records and follow-up records collected during routine clinical diagnosis and treatment at our hospital. The study protocol was reviewed and approved by the Ethics Committee of our hospital, and the requirement for patient informed consent was waived. The study strictly adhered to the principles of the Declaration of Helsinki. To protect patient privacy, all extracted data were de-identified, with sensitive information such as patients’ real names and ID numbers replaced by codes. Data were stored on an encrypted server and were accessible only to authorised researchers. No intervention was made to the patients’ routine diagnosis and treatment protocols during the study, ensuring that patients’ rights and interests were not compromised.

Treatment regimens

Surgical treatment regimen

All patients underwent laparoscopic ovarian endometrioma cystectomy. Preoperative routine examinations included pelvic ultrasound, serum cancer antigen 125 (CA125) measurement, and ovarian reserve assessment. The surgery adhered to the principle of “thorough lesion removal, preservation of normal ovarian tissue, and restoration of pelvic anatomy.” The cyst wall was completely excised using hydrodissection or sharp dissection, with maximum preservation of normal ovarian cortex. Intraoperative staging was recorded according to the revised American Fertility Society (r-AFS) classification for endometriosis. At the end of the procedure, the pelvic cavity was thoroughly irrigated, and anti-adhesion materials were applied when necessary.

Postoperative pharmacological treatment regimen

All patients received a standardized sequential maintenance regimen with GnRH-a following surgery.

Phase 1: Short-term GnRH-a therapy. Drug selection: leuprorelin acetate microspheres for injection (3.75 mg/vial) or goserelin acetate sustained-release implant (3.6 mg/implant). Administration regimen: subcutaneous or intramuscular injection initiated on days 1–5 of the first postoperative menstrual period, once every 28 days, for a total of 6 consecutive injections. Add-back therapy: if pronounced hypoestrogenic symptoms (Kupperman score ≥ 15) occurred after 3 GnRH-a injections, add-back therapy (estradiol 0.5-1.0 mg/d + dydrogesterone 5–10 mg/d) was administered as appropriate.

Phase 2: Sequential maintenance therapy (initiated within 4 weeks after the last GnRH-a injection). GnRH-a + DNG group: dienogest tablets, 2 mg orally once daily at a fixed time, continued until the end of follow-up or recurrence. GnRH-a + COC group: combined oral contraceptives (desogestrel/ethinylestradiol or drospirenone/ethinylestradiol tablets), 1 tablet orally once daily for 21 consecutive days followed by a 7-day washout interval, continued until the end of follow-up or recurrence. GnRH-a + LNG-IUS group: levonorgestrel-releasing intrauterine system, inserted intrauterinely by an experienced physician, 52 mg/system with a release rate of 20 µg/day, left in place for 5 years or until recurrence.

Data collection

Clinical data were collected by reviewing inpatient and outpatient medical records and conducting telephone follow-ups with the patients. The collected data included the following: ① Body weight, menstrual, and reproductive data: body mass index (BMI), age at menarche, reproductive age, preoperative dysmenorrhea Visual Analog Scale (VAS) score, preoperative parity, and number of miscarriages; ② Medical history data: history of pelvic procedures; ③ Preoperative laboratory findings: fibrinogen (FIB), D-dimer (D-D), prothrombin time (PT), serum CA125, and serum CA199; ④ Comorbidities: uterine fibroids and adenomyosis; ⑤ Surgery-related data: age at surgery, laterality of the cyst, cyst diameter, intraoperative cyst rupture, and revised American Fertility Society (r-AFS) stage.

Clinical outcome measures

Primary outcome measure

Postoperative recurrence was defined as the de novo appearance of an ovarian cystic lesion meeting criteria ① through ③ below. Alternatively, recurrence could be confirmed by criterion ④, irrespective of ultrasound morphology.

① Cyst-free interval: No evidence of any ovarian cyst (particularly endometrioma) on pelvic ultrasound performed at 6 months after primary surgery, to distinguish true recurrence from persistent residual disease.

② Morphological criteria: A newly detected ovarian cyst with a maximum diameter ≥ 2.0 cm exhibiting ultrasound features highly suggestive of an endometrioma, including diffuse internal low-level (“ground-glass”) echoes and/or peripheral hyperechoic foci suggestive of old hemorrhage.

③ Persistence: The cystic lesion meeting criterion ② must persist on a follow-up ultrasound performed at least 3 months after its initial detection (i.e., the cyst does not resolve spontaneously).

④ Alternative confirmation: Histopathological confirmation of ovarian endometriosis at second-look surgery.

Secondary outcome measures

(1) Cumulative recurrence rate: Cumulative recurrence rates at 1, 2, and 3 years postoperatively, calculated using the Kaplan‑Meier method.

(2) Menstrual bleeding profiles: Used to assess the improvement of menstrual abnormalities during treatment. Specific indicators included: ① Spotting (%): the proportion of patients who experienced spotting during treatment, defined as episodes of unscheduled intrauterine bleeding lasting at least one day, not requiring the use of sanitary protection beyond a panty liner, as recorded in patient symptom diaries or outpatient interviews; ② Average daily flow: quantified amount of bleeding during menstruation; ③ Dysmenorrhea (%): the proportion of patients with dysmenorrhea symptoms; ④ menstrual cycle length: the mean number of days of the menstrual cycle.

(3) Endometrioma diameter at recurrence: The diameter of the cyst at the time of recurrence was recorded.

(4) Adverse drug reactions: Weight change: weight gain ≥ 3 kg from baseline was defined as weight increase. Mood disturbances: including anxiety, depression, and mood swings, screened using the Hospital Anxiety and Depression Scale (HADS). Breast tenderness: patient‑reported or confirmed by physical examination. Headache: patient‑reported new‑onset headache after excluding other definite causes (e.g., infection, hypertension, history of migraine). Change in bone mineral density: bone mineral density of the lumbar spine and femoral neck was measured by dual‑energy X‑ray absorptiometry (DXA) at 1 and 2 years postoperatively; a T‑score < -2.5 was defined as osteoporosis, and a T‑score between − 1.0 and − 2.5 was defined as osteopenia. Thrombotic events: including deep vein thrombosis, pulmonary embolism, and cerebral infarction, confirmed by imaging studies.

Statistical analysis

Statistical analysis was performed using SPSS 26.0, with two-tailed tests and a significance level of α = 0.05. Normality of continuous variables was assessed via the Shapiro–Wilk test. Normal variables are presented as mean ± SD and were compared using one-way ANOVA; non-normal variables as M (Q₁, Q₃) and compared using the Kruskal–Wallis H test. Categorical variables are presented as n (%) and compared using chi-square or Fisher’s exact test. Kaplan–Meier curves were plotted to estimate 1‑, 2‑, and 3‑year cumulative recurrence rates and median recurrence‑free survival. All recurrence endpoints were analysed as time‑to‑event data with right‑censoring accounted for. The reported cumulative recurrence rates are Kaplan–Meier estimates, not crude proportions. Between‑group comparisons of recurrence‑free survival distributions were performed using the log‑rank test (overall P < 0.05). Pairwise comparisons used the Bonferroni method (α’=0.017). Univariate logistic regression (entry P < 0.05) screened potential risk factors for postoperative recurrence; variables with P < 0.10 entered multivariate stepwise regression to identify independent risk factors, reporting OR and 95% CI. Model fit was assessed by the Hosmer–Lemeshow test.

Results

Comparison of general information

Table 1 compares the baseline clinical characteristics of patients in the GnRH-a + DNG group, GnRH-a + COC group, and GnRH-a + LNG-IUS group, including age, body mass index, severity of dysmenorrhea, cyst diameter, r-AFS stage, reproductive history, comorbidities, and surgery-related data. The results showed no statistically significant differences among the three groups for all variables included (P > 0.05), indicating good comparability between the groups and providing a reliable basis for baseline balance for subsequent comparisons of treatment efficacy and recurrence risk between groups.

Table 1.

Comparison of general information

Variable	GnRH-a + DNG group (n = 61)	GnRH-a + COC group (n = 64)	GnRH-a + LNG-IUS group (n = 42)	Statistic	P
BMI (kg/m²), mean ± SD	22.61 ± 2.63	22.68 ± 3.01	22.29 ± 3.52	F = 0.220	0.803
Dysmenorrhea VAS score, M (Q₁, Q₃)	4.00 (2.00, 6.00)	6.00 (3.00, 7.00)	5.50 (4.00, 7.00)	χ2 = 3.768	0.152
Number of miscarriages, n (%)				χ2 = 1.902	0.386
0	32 (52.46)	41 (64.06)	26 (61.90)
1	29 (47.54)	23 (35.94)	16 (38.10)
Preoperative parity, n (%)				χ2 = 0.829	0.661
0	13 (21.31)	11 (17.19)	7 (16.67)
1	40 (65.57)	44 (68.75)	27 (64.29)
2	8 (13.11)	9 (14.06)	8 (19.05)
History of pelvic operation, M (Q₁, Q₃)	2.00 (0.00, 2.00)	2.00 (0.00, 2.25)	2.00 (0.00, 2.00)	χ2 = 0.065	0.968
Preoperative coagulation parameters
FIB positive, n (%)	14 (22.95)	14 (21.88)	6 (14.29)	χ2 = 1.299	0.522
D-D positive, n (%)	24 (39.34)	25 (39.06)	14 (33.33)	χ2 = 0.462	0.794
PT positive, n (%)	6 (9.84)	5 (7.81)	7 (16.67)	-	0.375
Preoperative tumor markers
CA199 positive, n (%)	35 (57.38)	31 (48.44)	21 (50.00)	χ2 = 1.099	0.577
CA125 positive, n (%)	28 (45.90)	33 (51.56)	22 (52.38)	χ2 = 0.562	0.755
Comorbidities
Complicated with uterine fibroids, n (%)	16 (26.23)	12 (18.75)	9 (21.43)	χ2 = 1.030	0.597
Complicated with adenomyosis, n (%)	9 (14.75)	11 (17.19)	9 (21.43)	χ2 = 0.774	0.679
Surgery‑related data
Age at surgery (years), mean ± SD	36.20 ± 5.84	35.00 ± 4.57	34.07 ± 7.24	F = 1.737	0.179
Laterality of cyst, n (%)				χ2 = 0.036	0.982
Unilateral	41 (67.21)	42 (65.62)	28 (66.67)
Bilateral	20 (32.79)	22 (34.38)	14 (33.33)
Cyst diameter at surgery (cm), M (Q₁, Q₃)	6.00 (4.10, 7.40)	4.80 (3.88, 7.30)	4.85 (3.60, 7.15)	χ2 = 1.481	0.477
Cyst rupture, n (%)	15 (24.59)	14 (21.88)	14 (33.33)	χ2 = 1.809	0.405
r‑AFS stage, n (%)				χ2 = 1.324	0.516
Stage III	29 (47.54)	37 (57.81)	22 (52.38)
Stage IV	32 (52.46)	27 (42.19)	20 (47.62)

Comparison of postoperative cumulative recurrence rates

The Kaplan-Meier method was used to calculate the 1-year, 2-year, and 3-year cumulative recurrence rates for each group. The results are presented in Table 2; Fig. 2. Table 2; Fig. 2 show the 1-year, 2-year, and 3-year cumulative recurrence rates and the number of recurrence cases among the three groups. The Kaplan–Meier estimated 3‑year cumulative recurrence rate was 19.67% in the GnRH‑a + DNG group, compared with 34.38% in the GnRH‑a + COC group and 45.24% in the GnRH‑a + LNG‑IUS group. All estimates account for right‑censored data and varying follow‑up times. The log-rank test revealed a statistically significant difference in cumulative recurrence rates among the three groups (log-rank χ²=8.553, P = 0.014). Further pairwise comparisons suggested that the GnRH-a + DNG group had a significantly lower recurrence rate than the GnRH-a + LNG-IUS group (log-rank χ²=8.767, P = 0.003). The recurrence rate in the GnRH-a + DNG group was lower than that in the GnRH-a + COC group, but the difference did not reach statistical significance (log-rank χ²=3.750, P = 0.053). There was no statistically significant difference between the GnRH-a + COC group and the GnRH-a + LNG-IUS group (log-rank χ²=1.334, P = 0.248).

Table 2.

Comparison of postoperative cumulative recurrence rates

Recurrence cases, n (%)	GnRH-a + DNG group (n = 61)	GnRH-a + COC group (n = 64)	GnRH-a + LNG-IUS group (n = 42)
Non-recurrence cases, n (%)	49 (80.33)	42 (65.62)	23 (54.76)
Mean recurrence-free survival (months, 95% CI)	33.246 (31.542–34.950)	30.047 (27.739–32.355)	27.857 (24.669–31.045)
1-year cumulative recurrence rate, %	3 (4.92)	7 (10.94)	6 (14.29)
2-year cumulative recurrence rate, %	7 (11.48)	15 (23.44)	13 (30.95)
3-year cumulative recurrence rate, %	12 (19.67)	22 (34.38)	19 (45.24)

Fig. 2.

Comparison of cumulative recurrence rates among the three groups of patients after surgery

Comparison of menstrual bleeding profiles

Table 3 compares the menstrual bleeding-related indicators among the three groups of patients during sequential maintenance therapy. The results suggested that the GnRH-a + LNG-IUS group had the highest incidence of spotting (52.38%), which was significantly higher than that in the DNG group (24.59%) and the COC group (12.50%), with a statistically significant difference (P < 0.001). However, there were no statistically significant differences among the three groups in average daily flow, incidence of dysmenorrhea, or average menstrual cycle length (P > 0.05), suggesting that the different treatment regimens had similar effects on the above indicators.

Table 3.

Comparison of menstrual bleeding profiles

Indicator	GnRH-a + DNG group (n = 61)	GnRH-a + COC group (n = 64)	GnRH-a + LNG-IUS group (n = 42)	Statistic	P
Spotting (%)	15 (24.59)	8 (12.50)	22 (52.38)	χ2 = 20.760	< 0.001
Average daily flow (mL)	4.11 ± 0.32	4.09 ± 0.39	4.21 ± 0.42	F = 1.431	0.242
Dysmenorrhea (%)	10 (16.39)	12 (18.75)	8 (19.05)	χ2 = 0.162	0.922
menstrual cycle length	28.46 ± 4.23	28.02 ± 3.23	28.60 ± 3.88	F = 0.3602	0.699

Comparison of cyst diameter at recurrence

Table 4 presents the analysis of endometrioma diameter at recurrence in 53 patients who experienced recurrence. The results suggested that the recurrent cyst diameters in the GnRH-a + DNG group, GnRH-a + COC group, and GnRH-a + LNG-IUS group were 3.80 ± 0.96 cm, 3.83 ± 1.13 cm, and 4.24 ± 0.97 cm, respectively. There was no statistically significant difference among the three groups (P = 0.381), indicating that although the recurrence rates differed between groups, once recurrence occurred, the cyst sizes were not significantly different among the groups.

Table 4.

Comparison of recurrent cyst diameters

Group	Cyst diameter at recurrence (cm)
GnRH-a + DNG group (n = 12)	3.80 ± 0.96
GnRH-a + COC group (n = 22)	3.83 ± 1.13
GnRH-a + LNG-IUS group (n = 19)	4.24 ± 0.97
Statistic	F = 0.984
P	0.381

Comparison of adverse drug reactions

Table 5 summarizes the adverse drug reactions experienced by patients under the three treatment regimens, including weight gain, mood disturbances, breast tenderness, headache, etc. Although the total incidence of adverse drug reactions was similar among the three groups (13.11% in GnRH‑a + DNG, 14.06% in GnRH‑a + COC, and 11.90% in GnRH‑a + LNG‑IUS, P = 0.950), the pattern of specific adverse events differed. Headache was numerically more frequent in the LNG‑IUS group (7.14%) and DNG group (6.56%) than in the COC group (1.56%), while breast tenderness was slightly higher in the COC group (4.69%). No patient in any group discontinued treatment due to adverse events during the 3‑year follow‑up. Notably, no cases of decreased bone mineral density or thrombotic events were recorded in any group, which may be partly attributed to the short observation period and the exclusion of patients with baseline contraindications.

Table 5.

Comparison OF ADVERSE DRUG REACTIONS

Adverse reactions	GnRH-a + DNG group (n = 61)	GnRH-a + COC group (n = 64)	GnRH-a + LNG-IUS group (n = 42)	Statistic	P
Weight gain	1 (1.64)	3 (4.69)	1 (2.38)
Mood disturbance	1 (1.64)	2 (3.13)	0 (0.00)
Breast tenderness	2 (2.28)	3 (4.69)	1 (2.38)
Headache	4 (6.56)	1 (1.56)	3 (7.14)
Decreased bone mineral density	0 (0.00)	0 (0.00)	0 (0.00)
Thrombotic events	0 (0.00)	0 (0.00)	0 (0.00)
Total	8 (13.11)	9 (14.06)	5 (11.90)	χ2 = 0.104	0.950

Comparison of baseline characteristics between the recurrence and non-recurrence groups

Table 6 compares the differences in baseline characteristics between patients in the postoperative recurrence group and the non-recurrence group. The results suggested that the recurrence group had significantly more patients with higher dysmenorrhea VAS score, more frequent history of pelvic procedures, younger age at surgery, higher proportion of bilateral cysts, larger cyst diameter, and higher proportion of r-AFS stage IV compared to the non-recurrence group (all P < 0.05), suggesting that these factors may be associated with an increased risk of postoperative recurrence.

Table 6.

Comparison of baseline characteristics between the recurrence and non-recurrence groups

Variable	Total (n = 167)	Recurrence group (n = 53)	Non-recurrence group (n = 114)	Statistic	P
BMI (kg/m²), mean ± SD	22.55 ± 3.01	22.43 ± 3.11	22.62 ± 2.97	t = 0.379	0.705
Reproductive age, n (%)
Dysmenorrhea VAS score, M (Q₁, Q₃)	5.00 (3.00, 7.00)	6.00 (5.00, 8.00)	4.00 (3.00, 6.00)	Z=-4.063	< 0.001
Number of miscarriages, n (%)				χ2 = 0.231	0.631
0	99 (59.28)	30 (56.60)	69 (60.53)
1	68 (40.72)	23 (43.40)	45 (39.47)
Preoperative parity, n (%)				Z=-0.556	0.578
0	31 (18.56)	8 (15.09)	23 (20.18)
1	111 (66.47)	37 (69.81)	74 (64.91)
2	25 (14.97)	8 (15.09)	17 (14.91)
History of pelvic operation, M (Q₁, Q₃)	2.00 (0.00, 2.00)	2.00 (0.00, 3.00)	2.00 (0.00, 2.00)	Z=-3.638	< 0.001
Preoperative coagulation parameters
FIB positive, n (%)	34 (20.36)	8 (15.09)	26 (22.81)	χ2 = 1.327	0.249
D-D positive, n (%)	63 (37.72)	22 (41.51)	41 (35.96)	χ2 = 0.473	0.491
PT positive, n (%)	18 (10.78)	3 (5.66)	15 (13.16)	χ2 = 2.115	0.146
Preoperative tumor markers
CA199 positive, n (%)	87 (52.10)	25 (47.17)	62 (54.39)	χ2 = 0.755	0.385
CA125 positive, n (%)	83 (49.70)	32 (60.38)	51 (44.74)	χ2 = 3.540	0.060
Comorbidities
Complicated with uterine fibroids, n (%)	37 (22.16)	11 (20.75)	26 (22.81)	χ2 = 0.088	0.766
Complicated with adenomyosis, n (%)	29 (17.37)	12 (22.64)	17 (14.91)	χ2 = 1.506	0.220
Surgery‑related data
Age at surgery (years), mean ± SD	35.20 ± 5.82	33.11 ± 6.08	36.18 ± 5.46	t = 3.255	0.001
Laterality of cyst, n (%)
Unilateral	111 (66.47)	28 (52.83)	83 (72.81)	χ2 = 6.478	0.011
Bilateral	56 (33.53)	25 (47.17)	31 (27.19)
Cyst diameter at surgery (cm), mean ± SD	5.35 ± 1.74	5.86 ± 1.82	5.11 ± 1.67	t=-2.602	0.010
Cyst rupture, n (%)	43 (25.75)	16 (30.19)	27 (23.68)	χ2 = 0.801	0.371
r‑AFS stage, n (%)				χ2 = 15.776	< 0.001
Stage III	88 (52.69)	16 (30.19)	72 (63.16)
Stage IV	79 (47.31)	37 (69.81)	42 (36.84)

Logistic regression analysis results

Table 7 presents the independent risk factors for postoperative recurrence of ovarian endometrioma identified by univariate and multivariate logistic regression analysis. The multivariate analysis results suggested that higher dysmenorrhea VAS score (OR = 1.376), history of pelvic procedures (OR = 1.483), and r-AFS stage IV (OR = 2.676) were independent risk factors for postoperative recurrence (all P < 0.05), while older age at surgery was a protective factor (OR = 0.891). The model showed good fit (Hosmer-Lemeshow test: χ²=11.468, P = 0.177), suggesting that these factors can be used for clinical risk assessment and individualized treatment decisions.

Table 7.

Univariate and multivariate Logistic regression analysis of factors associated with postoperative recurrence of ovarian endometrioma

Variable	Univariate analysis				Multivariate analysis
	β	S.E	OR (95%CI)	P	β	S.E	OR (95%CI)	P
Dysmenorrhea VAS score	0.311	0.077	1.365 (1.173–1.589)	< 0.001	0.319	0.089	1.376 (1.154–1.639)	< 0.001
History of pelvic operation	0.521	0.151	1.684 (1.252–2.265)	< 0.001	0.394	0.167	1.483 (1.068–2.058)	0.019
Age at surgery (years)	-0.094	0.030	0.910 (0.858–0.966)	0.002	-0.116	0.037	0.891 (0.828–0.958)	0.002
Laterality of cyst
Unilateral			1.000 (Reference)				1.000 (Reference)
Bilateral	0.872	0.346	2.391 (1.212–4.714)	0.012	0.509	0.417	1.664 (0.735–3.766)	0.222
Cyst diameter at surgery (cm)	0.252	0.100	1.287 (1.059–1.565)	0.011	0.163	0.118	1.177 (0.935–1.483)	0.163
r‑AFS stage
Stage III			1.000 (Reference)				1.000 (Reference)
Stage IV	1.377	0.357	3.964 (1.970–7.976)	< 0.001	0.984	0.407	2.676 (1.206–5.937)	0.016

Discussion

The pathogenesis of endometriosis remains unclear. Symptoms may appear as early as adolescence, and the disease exhibits distant invasiveness involving multiple organs. Even after standardized surgical and medical treatment, postoperative recurrence rates remain high [26]. The high recurrence rate, low cure rate, and suboptimal treatment outcomes of ovarian endometrioma represent a challenging issue in the current management of ovarian endometrioma [27]. How to reduce the probability of recurrence and how to identify predictors of recurrence before it occurs are the key difficulties in treatment. Based on real-world clinical data, this study systematically compared the efficacy differences of three GnRH-a sequential maintenance regimens in preventing postoperative recurrence of ovarian endometrioma and identified independent risk factors affecting postoperative recurrence. The results suggested that sequential maintenance therapy with dienogest following short-term GnRH-a treatment resulted in a significantly lower 3-year cumulative recurrence rate compared to the sequential LNG-IUS regimen, and showed a numerical trend toward a lower recurrence rate compared with the sequential COC regimen, although this difference was not statistically significant. This finding provides important evidence-based support for clinicians in making individualized decisions regarding long-term postoperative medical management of endometriosis. Our results are highly consistent with the network meta-analysis by Chiu et al. [28], which included 11 clinical studies with a total of 2,394 patients and found that the GnRH-a combined with DNG regimen ranked highest in preventing postoperative recurrence of ovarian endometrioma, with an odds ratio of 0.04, significantly better than other combination regimens. The systematic review and meta-analysis by Liu et al. [29] similarly confirmed that DNG maintenance therapy significantly reduces the risk of postoperative recurrence, with a statistically significant difference compared to expectant management. Of note, although this study used retrospective real-world data, the three groups of patients had good comparability in baseline characteristics, which to some extent enhances the reliability of the study conclusions.

Regarding the secondary outcome measures in this study, there were no significant differences among the three groups in terms of average daily menstrual blood loss, incidence of dysmenorrhea, or mean menstrual cycle length. However, the incidence of spotting in the LNG-IUS group was significantly higher than that in the DNG group and the COC group. Although DNG maintenance therapy demonstrated excellent efficacy in preventing recurrence, it is indeed associated with an increased risk of vaginal bleeding. The meta-analysis by Liu et al. [29] clearly indicated that the incidence of vaginal bleeding was significantly higher in the DNG group. In the present study, the incidence of spotting in the LNG-IUS group was as high as 52.38%, which was significantly higher than that in the DNG group (24.59%) and the COC group (12.50%). This difference may be attributed to irregular endometrial shedding caused by the locally high concentration of progestogen in the uterine cavity during the early period after LNG-IUS placement. Although such spotting is generally not harmful to health, long-term abnormal bleeding patterns may negatively affect patient adherence to treatment and quality of life. Clinicians should take this into consideration when selecting a treatment regimen. Regarding adverse drug reactions, the overall incidence rates in the three groups were low, and no decreased bone mineral density or thrombotic events were observed, suggesting that all three regimens have favorable overall safety profiles.

This study identified four independent risk factors for postoperative recurrence through multivariate logistic regression analysis: higher dysmenorrhea VAS score, history of pelvic procedures, and r-AFS stage IV. Additionally, older age at surgery was found to be a protective factor. These findings provide important quantitative evidence for preoperative risk stratification and individualized treatment decisions in clinical practice. Jiang et al. [30] systematically summarized previous evidence on risk factors for postoperative recurrence of ovarian endometrioma, and their study clearly indicated that younger age at surgery, greater severity of preoperative dysmenorrhea, history of endometriosis-related surgery, and higher r-AFS score were all risk factors for recurrence, which is highly consistent with the results of the present study.

Endometriosis is a hormone-dependent disease, and endogenous estrogen levels tend to increase first and then decrease with advancing age. Under natural conditions, younger patients have relatively higher postoperative recurrence risk due to higher estrogen levels, whereas older patients have a gradually decreasing recurrence risk because of declining ovarian function and lower estrogen levels. Therefore, older age at surgery is generally considered a protective factor [31]. The r-AFS stage is one of the core indicators for assessing the severity of endometriosis. The present study confirmed that r-AFS stage IV is an important risk factor for postoperative recurrence of ovarian endometrioma. The underlying mechanisms can be explained from the following aspects. First, r-AFS stage IV typically indicates widespread distribution of endometriotic lesions accompanied by severe pelvic adhesions, which significantly increases the risk of residual ectopic endometrial cells during surgery [32]. Second, r-AFS stage IV induces a more intense local inflammatory response. Even after partial removal of the lesions, the persistent inflammatory microenvironment can promote the proliferation of residual ectopic endometrial cells and induce immune dysfunction, thereby reducing the immune recognition and clearance of ectopic endometrial cells and ultimately facilitating cyst recurrence [33]. Third, endometriotic lesions in r-AFS stage IV are often associated with marked neovascularization. These newly formed blood vessels provide adequate blood supply and nutritional support for ectopic endometrial cells. In such a vascular-rich microenvironment, even a small number of residual ectopic endometrial cells after surgery can rapidly obtain the necessary material basis for growth, ultimately leading to recurrence of ovarian endometrioma [34].

Endometriosis-associated pain includes various types such as dysmenorrhea, dyspareunia, chronic pelvic pain, and anorectal pain, among which dysmenorrhea is the most common and clinically indicative symptom. The results of this study show that severe dysmenorrhea is a high-risk factor for postoperative recurrence of ovarian endometrioma. Mechanistically, focal bleeding of ectopic lesions, the action of inflammatory factors in the peritoneal cavity, and stimulation or direct infiltration of nerve fibers may all contribute to the development of dysmenorrhea [35]. The density of nerve fibers is increased in ectopic lesions. These fibers not only transmit pain signals but also release neuropeptides such as substance P and calcitonin gene-related peptide (CGRP), which promote local inflammatory reactions, proliferation of ectopic endometrial cells, and angiogenesis, potentially creating a favorable microenvironment for the persistent growth and recurrence of residual postoperative lesions [36]. Therefore, patients with severe dysmenorrhea after ovarian endometrioma surgery should be considered a high-risk population for recurrence and receive enhanced follow-up. Combined with CA125 testing and ultrasound examinations, comprehensive assessment and early prediction of recurrence risk should be performed. The pathogenesis of ovarian endometrioma is associated with pelvic surgical procedures. After various pelvic surgeries, such as cesarean section or myomectomy, the inflammatory response triggered by the tissue wound surface leads to fibrinogen exudation and fibrin deposition, which in turn forms adhesions, alters the anatomical structure and physiological environment of the pelvic cavity, and affects the clearance of ectopic endometrial cells by immune cells [37]. At the same time, adhesions may encapsulate residual ectopic endometrial cells locally, providing them with a relatively independent growth environment that protects them from attack by the immune system, thereby facilitating the survival and proliferation of ectopic endometrial cells [38]. Moreover, pelvic surgery can interfere with the female endocrine system, causing fluctuations in hormone levels that stimulate the growth and proliferation of ectopic endometrial cells and increase the risk of postoperative recurrence of ovarian endometrioma [39].

The findings of this study have clear guiding value for clinical practice. First, sequential maintenance therapy with DNG following short-term GnRH-a treatment may be considered a preferred strategy, particularly in patients with high-risk factors, but its advantage over COC requires confirmation in larger prospective studies. Second, for patients with contraindications to LNG-IUS or intolerance to irregular vaginal bleeding, DNG or COC may be more appropriate alternative options. For high-risk patients with multiple independent risk factors, more intensive pharmacological intervention strategies and closer follow-up monitoring should be considered. The study by Capezzuoli et al. [40] emphasized that long-term hormonal therapy can significantly reduce the rate of repeated surgery for endometriosis and improve patients‘ quality of life, and that individualized treatment strategies should become the core concept of modern endometriosis management. The study by Wacharachawana et al. [41] also indicated that standardized postoperative medication use is an important protective factor against recurrence, further reinforcing the importance of postoperative medical management.

Beyond hormonal maintenance therapy, emerging evidence suggests that adjunctive strategies may further improve long-term outcomes in endometriosis patients, particularly those with infertility. Inositols, as insulin-sensitizing and anti-inflammatory nutraceuticals, have shown potential in reducing chronic pelvic inflammation and improving ovarian function in endometriosis-associated infertility [42]. However, their role in preventing postoperative recurrence remains unestablished. For young patients with high-risk factors (e.g., r-AFS stage IV, recurrent endometriomas, or bilateral cysts), fertility preservation through oocyte vitrification using an antagonist protocol with GnRH antagonist dual trigger may be considered before repeated surgery [43]. Both open and closed vitrification systems appear equally effective for blastocyst transfer, although closed systems theoretically reduce cross-contamination risks [44]. Long-term neonatal outcomes and follow-up of children born from frozen embryos are reassuring but warrant continued surveillance [45]. In pregnant patients with suspected endometriosis-related complications, contrast-enhanced MRI may be used cautiously when clinically necessary, as contrast agents cross the placenta but are not associated with proven teratogenic effects in human studies [46]. Advanced MRI techniques, including diffusion-weighted imaging and T2 mapping, are also improving non-invasive diagnosis of deep infiltrating endometriosis [47]. Collectively, these evolving strategies—nutraceuticals, fertility preservation, advanced imaging, and long-term offspring follow-up—complement hormonal maintenance therapy and should be integrated into individualized management plans for women with endometriosis.

Surgical approaches for advanced endometriosis with concurrent uterine pathology warrant updated consideration. Recent evidence suggests that minilaparoscopic hysterectomy is associated with comparable perioperative outcomes, shorter hospital stay, and improved cosmetic satisfaction compared with conventional laparoscopic hysterectomy, supporting its feasibility in selected patients with endometriosis and concomitant uterine fibroids or adenomyosis [48]. At the biological level, follicular fluid modulation of progesterone and IL-6 expression in granulosa cells may influence ovarian reserve and inflammatory status after repeated ovarian surgery, providing a potential mechanistic link between surgical trauma, local inflammation, and recurrence risk [49]. For overweight patients with endometriosis-related infertility, supplement strategies such as inositols and lifestyle modifications may improve metabolic and reproductive outcomes, although their direct role in preventing postoperative recurrence remains to be established [50]. Furthermore, immunological interventions in patients with recurrent implantation failure, including cytokine modulation and intrauterine immune profiling, offer insights into the immune microenvironment that may also be relevant to endometriosis recurrence [51]. Collectively, these evolving surgical, endocrine, metabolic, and immunological perspectives complement hormonal maintenance therapy and support a multidimensional, individualized approach to endometriosis care.

This study has several strengths. First, it is a real‑world retrospective cohort study with good baseline comparability among the three sequential maintenance therapy groups, which enhances the practical relevance of the findings. Second, to our knowledge, it is one of the few studies to directly compare dienogest, combined oral contraceptives, and the levonorgestrel‑releasing intrauterine system within a GnRH‑a sequential treatment framework. Third, we employed rigorous statistical methods, including Kaplan‑Meier survival analysis and multivariate logistic regression, to control for potential confounders. Fourth, the identification of independent risk factors (e.g., dysmenorrhea VAS score, history of pelvic procedures, r‑AFS stage IV, and age at surgery) provides clinically actionable information for individualised postoperative management.

Study limitations

This study employed a retrospective cohort design. This study is a retrospective cohort analysis without randomization; therefore, confounding by indication is a major concern. Despite statistical adjustment for multiple measured confounders, we cannot rule out the possibility that unmeasured or incompletely measured factors—such as patient pain perception, treatment adherence, socioeconomic status, or physician preference—influenced both the choice of maintenance therapy and the risk of recurrence. Consequently, the observed differences in recurrence rates among groups should be interpreted as hypothesis‑generating rather than causal. In addition, the follow-up duration was only 3 years. Although this is sufficient to assess intermediate-term recurrence risk, endometriosis is a chronic recurrent disease, and longer-term recurrence patterns require further observation. The study population was derived from a single medical center, and the sample size was relatively limited, which may affect the generalizability and external validity of the results.

Furthermore, while we used the Kaplan-Meier method to estimate cumulative recurrence rates and the log-rank test for between-group comparisons, we did not employ Cox proportional hazards regression to model time-to-recurrence as a function of multiple covariates simultaneously. Logistic regression was used for risk factor analysis by dichotomizing recurrence status at the end of follow-up, which does not fully account for variable follow-up durations across individuals. Future prospective studies with larger sample sizes should employ Cox regression to better handle censored data and variable follow-up times, thereby providing more precise estimates of treatment effects on recurrence risk.

It is worth noting that the finding of a higher recurrence rate in the LNG-IUS group, while consistent with the drug’s mechanism of action and some previous studies, may also be related to its smaller sample size, local rather than systemic action, possible selection bias (e.g., patients intolerant to oral medications), or other unmeasured confounders. Therefore, the comparison between the LNG-IUS group and the other two groups requires cautious interpretation and should be considered hypothesis-generating rather than definitive. In addition, several important clinical variables were not available in this real-world retrospective study, including detailed records on smoking status, alcohol consumption, body composition metrics (e.g., fat mass, waist-to-hip ratio), preoperative duration of pain symptoms, adherence to medication beyond prescription records, and socioeconomic or educational background. These factors may potentially influence both the choice of maintenance therapy and the risk of postoperative recurrence, and their absence limits the comprehensiveness of our risk factor analysis. Future prospective studies should systematically collect these variables to enable more robust confounding adjustment.

The treatment regimens within the same group were not entirely homogeneous. In the GnRH-a phase, both leuprorelin and goserelin were used; in the COC group, patients received either desogestrel/ethinylestradiol or drospirenone/ethinylestradiol. Although these agents belong to the same pharmacological class and are commonly considered interchangeable in clinical practice, subtle differences in their potency, pharmacokinetics, or side effect profiles may exist. This heterogeneity was not controlled for in our analysis and could have introduced confounding, potentially masking a more nuanced treatment effect. Therefore, our findings should be interpreted as representing the average effect of each class of maintenance therapy rather than the effect of any specific agent. Future prospective studies should adopt standardized protocols with a single agent per arm to eliminate this source of variability.

Conclusions

Based on real‑world data, this study systematically compared the effectiveness of three GnRH‑a sequential maintenance regimens. In this cohort, DNG was associated with a lower recurrence rate than LNG‑IUS and showed a trend toward a lower rate than COC. However, comparisons involving the LNG‑IUS group should be interpreted cautiously due to its smaller sample size, potential selection bias, and the local nature of LNG‑IUS action. Future prospective studies with larger sample sizes and balanced allocation are needed to confirm these associations.

Abbreviations

DNG

Dienogest

GnRH-a

Gonadotropin-releasing hormone agonist

COC

Combined oral contraceptives

LNG-IUS

Levonorgestrel-releasing intrauterine system

r-AFS

revised American Fertility Society

VAS

Visual Analog Scale

CA125

Cancer antigen 125

CA199

Cancer antigen 199

FIB

Fibrinogen

D-D

D-dimer

PT

Prothrombin time

BMI

Body mass index

OR

Odds ratio

CI

Confidence interval

HADS

Hospital Anxiety and Depression Scale

DXA

Dual-energy X-ray absorptiometry

Authors’ contributions

Zhuo Chen: Participated in collecting, assessing, and interpreting the date. Made significant contributions to date interpretation and manuscript preparation.Zhuo Chen, Yingying Zheng: Provided substantial intellectual input during the drafting and revision of the manuscript.All authors have read and approved the final manuscript.

Funding

None.

Data availability

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was reviewed and approved by the Ethics Committee of Jinhua Municipal Central Hospital (Approval No.: (2026) Ethics approval No.(295)), and the requirement for patient informed consent was waived by the same ethics committee.

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.