Natural History and Risk Factors for Adverse Outcomes of Ovarian Endometrioma in Pregnancy — A 9-Year Cohort with Histopathological Confirmation

Among 151 patients (median age, 31 years [IQR, 29– 34 years]), cyst behavior in Group 1 was heterogeneous: 17.4% decreased, 60.5% remained stable, and 22.1% increased in size. Revised American Society for Reproductive Medicine (rASRM) stage III–IV disease (61.6%) was significantly more prevalent in Group 1 (80.2% vs 36.9%; P < 0.001). Cyst rupture occurred in 20 patients (13.2%), decidualization in 26 (17.2%), and composite adverse outcomes in 22 (14.6%). rASRM stage III–IV disease (aOR = 5.784; 95% CI: 1.549– 21.596; P = 0.009) and decidualization (aOR = 4.530; 95% CI: 1.541– 13.316; P = 0.006) were independent risk factors for adverse outcomes, whereas absolute cyst size was not.

The gestational course of OE is characterized by considerable interindividual variability; most cysts remain stable, whereas approximately one-fifth enlarge. rASRM stage III–IV disease and decidualization were independent risk factors for adverse outcomes, whereas cyst size alone was not. Increased surveillance may be warranted for patients with advanced-stage disease or sonographic features suggestive of decidualization.

pregnancy, ovarian endometrioma, decidualization, adverse outcomes, rASRM stage

Endometriosis is an estrogen-dependent, chronic systemic disorder characterized by the presence of endometrial-like tissue (glands and stroma) outside the uterine cavity. The condition typically manifests as pelvic pain and infertility and affects an estimated 5–10% of women of reproductive age worldwide.1,2 Endometriosis is classified into four phenotypes based on lesion morphology and depth of infiltration: superficial peritoneal lesions, ovarian endometrioma (OE), deep infiltrating endometriosis (infiltration depth > 5 mm), and extra-pelvic lesions, among which OE represents one of the most prevalent presentations.3 The reported incidence of OE during pregnancy ranges from 0.05% to 0.5%.4 A single-center 12-year analysis from Japan demonstrated that the detection rate of OE during pregnancy increased from 0.14% to 0.52% (a 3.8-fold increase), suggesting a year-by-year upward trend that may be attributable to both the rising overall prevalence of endometriosis and the increasing age of women at childbirth.5 Despite this, our current understanding of the natural history of OE during pregnancy remains limited. As early as 1949, Beecham6 proposed that pregnancy could serve as an effective prophylactic and therapeutic measure against endometriosis, a concept that was subsequently widely cited. Nevertheless, a cautious perspective was already advancing at that time. In 1965, McArthur7 questioned the purported protective effect of pregnancy, noting that the supporting evidence was derived largely from case reports and small observational series without systematic validation. More recent studies have demonstrated that the gestational evolution of OE is highly heterogeneous; cysts may remain stable, regress, or enlarge, with some undergoing decidualization that may precipitate serious obstetric complications and adverse maternal–fetal outcomes.4,8 Furthermore, OE during pregnancy has been associated with a range of serious maternal complications, including cyst rupture, ovarian torsion, cyst infection, decidualization with or without malignant-like morphological features, suspected malignant transformation, and spontaneous hemoperitoneum, the latter being a rare but potentially life-threatening event.4 These complications, in turn, may translate into adverse pregnancy outcomes such as medically indicated (iatrogenic) preterm birth, low birth weight, placental disorders, and an increased overall risk of cesarean delivery.4 To date, systematic large-scale investigations of the natural history of OE during pregnancy and the associations between cyst characteristics and adverse maternal–fetal outcomes remain scarce. Therefore, the present study aimed to conduct a 9-year, single-center retrospective cohort analysis of 151 patients with histopathologically confirmed OE identified during cesarean section to (i) systematically describe the natural history of OE during pregnancy; (ii) evaluate the associations between cyst characteristics (size, stage, decidualization, and dynamic changes) and adverse maternal–fetal outcomes; and (iii) identify risk factors for adverse outcomes, thereby providing an evidence base for individualized antenatal management strategies.

Study Design and Participants This single-center retrospective cohort study was approved by the Ethics Committee of Hangzhou Women’s Hospital (approval No. [2026] Medical Ethics Review A No. 051) and was conducted in accordance with the principles of the Declaration of Helsinki (1964) and its subsequent revisions. Because patient consent to review medical records was not required by the Ethics Committee, individual informed consent was waived; all patient data were de-identified before analysis, assigned a study code, and stored on password-protected institutional servers accessible only to the study investigators, thereby ensuring patient data confidentiality. Consecutive women who underwent cesarean section at our institution between January 2017 and December 2025 and in whom OE was histopathologically confirmed based on intraoperative specimens were screened for eligibility. The inclusion criteria were as follows: (1) maternal age ≥18 years at delivery, (2) singleton or twin pregnancy with delivery by cesarean section, and (3) histopathological confirmation of OE based on intraoperative specimens. The exclusion criteria were as follows: (1) severe medical or surgical comorbidities; (2) known ovarian malignancy prior to conception; and (3) incomplete clinical data. Ultimately, 151 patients were enrolled. During the study period, 102,071 deliveries were recorded in the Department of Obstetrics at our hospital, among which 151 cases of pregnancy-associated OE were histopathologically confirmed at the time of cesarean section, corresponding to an incidence of 1 per 676 pregnancies. Patients were stratified into two groups according to the timing of OE detection: Group 1 (preconception/early-pregnancy detection group, n = 86), comprising patients in whom OE was identified by ultrasonography prior to conception or before 14 weeks of gestation; and Group 2 (intraoperative detection group, n = 65), comprising patients in whom OE was first identified and confirmed at cesarean section. Diagnosis and Assessment of OE Histopathological examination of surgical specimens obtained during the cesarean section served as the reference standard for the diagnosis of OE. For patients in Group 1, the most recent preconception ultrasound images were retrospectively retrieved; when these were unavailable, the earliest first-trimester ultrasound images documenting OE were used instead. The ultrasound diagnostic criteria were defined as a thick-walled cystic adnexal mass with well-defined borders and homogeneous low-level internal echoes (a “ground-glass” appearance).9 To minimize measurement variability, all sonographic examinations were performed by experienced sonographers at a single tertiary referral center using standardized transvaginal or transabdominal protocols, and the maximum cyst diameter was recorded; the same sonographer performed serial examinations whenever possible. Because the analysis relied on archived images and reports, formal interobserver and intraobserver reproducibility testing could not be performed, and this is acknowledged as a limitation. Ultrasound measurement is influenced by abdominal wall thickness, transducer pressure and operator technique, whereas intraoperative measurement is affected by specimen decompression after rupture, adhesions and reduced adnexal exposure owing to uterine enlargement. The two are therefore not strictly interchangeable, and the absolute size differences reported here should be interpreted accordingly. In the present study, a change in maximum cyst diameter of ≥1 cm was considered clinically significant, based on previously reported assessments of ultrasound measurement error.5 Baseline size was defined as the maximum cyst diameter measured on the most recent ultrasound scan performed prior to conception or before 14 weeks of gestation, whereas the endpoint size was defined as the maximum diameter measured under direct visualization during cesarean section. Based on changes in cyst size, patients in Group 1 were categorized into three subgroups: enlargement (increase ≥1 cm), stability (change between −1 cm and +1 cm), and regression (decrease ≥1 cm). The severity of endometriosis was classified according to the revised American Society for Reproductive Medicine (rASRM) staging system,10 with stages I–II defined as mild disease and stages III–IV as moderate-to-severe disease. The present study therefore prioritized a comprehensive surgical assessment of the disease over cyst diameter alone. Each patient’s stage was assigned from the summed score for the number, location, size and depth of the endometriotic lesions together with the degree of pelvic adhesions, a total of 1–15 points corresponding to stage I–II and a total of 16 points or more to stage III–IV. In the present cohort, the operative reports of the 58 patients classified as stage I–II described small endometriotic lesions lying on the ovarian surface, with adhesions documented in only one of these 58 reports and dense adhesions or obliteration of the posterior cul-de-sac in none, whereas adhesions were documented in 41 of the 93 reports of patients classified as stage III–IV (dense adhesions in 21 and involvement of the posterior cul-de-sac in six). These stage I–II lesions were therefore scored as superficial ovarian lesions (2 points for a lesion measuring 1–3 cm) rather than as deep ovarian lesions (16 points), and the summed score remained within stage I–II. In Group 2 specifically, the 41 patients classified as stage I–II had a median maximum intraoperative cyst diameter of 2.0 cm (IQR 2.0–3.0 cm; range 0.5–4.0 cm), whereas the 24 patients classified as stage III–IV had a median diameter of 4.0 cm (range 2.0–10.0 cm) with documented adhesions. It should be acknowledged that cesarean delivery does not permit systematic inspection of the posterior compartment or formal assessment of infiltration depth; some degree of under-staging therefore cannot be excluded. Outcome Measures The primary outcomes were as follows: (1) changes in cyst size during pregnancy among patients in Group 1; and (2) a composite endpoint of adverse maternal and neonatal outcomes, including cyst rupture, suspected malignant transformation, and cyst-related iatrogenic preterm birth. Secondary outcomes included the incidence of decidualization, gestational age at delivery, birth weight, and rASRM stage. Cyst-related iatrogenic preterm birth was defined as delivery before 37 weeks of gestation initiated specifically because of an ovarian endometrioma-related complication, as documented in the operative indication record. The qualifying complications were (i) cyst rupture presenting with acute abdominal pain, hemoperitoneum, or signs of peritoneal irritation; (ii) adnexal torsion of the affected ovary; (iii) rapid cyst enlargement or sonographic features suspicious for malignant transformation requiring expedited delivery; and (iv) pelvic mass effect causing obstruction of labor. Preterm births performed for unrelated obstetric indications—such as placenta previa, preeclampsia, multifetal gestation, or intrauterine infection—were not counted as cyst-related iatrogenic preterm births. In the present cohort, one patient met this definition. Suspected malignant transformation was defined by combined assessment of ultrasound morphology, serum tumor markers (interpretation limited by pregnancy physiology), serial clinical and sonographic surveillance for rapid enlargement, pelvic magnetic resonance imaging where uncertain, and multidisciplinary team discussion; definitive diagnosis rested on histopathology. Statistical Analysis All statistical analyses were performed using SPSS software (version 27.0; IBM Corp., Armonk, NY, USA). The normality of the distribution of continuous variables was assessed using the Kolmogorov–Smirnov test. Continuous variables with a non-normal distribution are expressed as median (interquartile range [IQR]) and were compared between groups using the Mann–Whitney U-test. Normally distributed continuous variables are expressed as mean ± standard deviation (mean ± SD) and were compared using the independent-samples t-test. Categorical variables are expressed as frequencies (percentages) and were compared using the chi-square test or Fisher’s exact test, as appropriate. The strength of the association between the candidate risk factors and outcomes was quantified using odds ratios (ORs) and 95% confidence intervals (CIs). Statistical significance was defined as a two-sided P-value < 0.05. Candidate risk factors were initially screened using univariable logistic regression analysis (inclusion criterion: P < 0.10) and subsequently entered into a multivariable logistic regression model, from which adjusted odds ratios (aORs) and 95% CIs were derived. The number of covariates included in the multivariable model was determined according to the principle of a minimum of 10 events per variable (EPV).11

Baseline Demographic Characteristics Baseline demographic and obstetric characteristics of the 151 patients are presented in Table 1. The overall cohort ranged in age from 23 to 41 years, with a median age of 31 years (IQR, 29–34 years). The median pre-pregnancy body mass index (BMI) was 20.3 kg/m2 (IQR 19.2–21.9 kg/m2). Nulliparity was observed in 103 patients (68.2%), conception via assisted reproductive technology in 14 (9.3%), multiple gestations in 9 (6.0%), multiple cysts in 16 (10.6%), and a university degree or higher in 136 (90.1%). | Table 1 Baseline Demographic and Obstetric Characteristics of the Study Population | Eighty-six patients (57.0%) were identified as having OE before conception or during early pregnancy and were assigned to Group 1, whereas the remaining 65 patients (43.0%) were first diagnosed with OE at the time of cesarean section and were assigned to Group 2. No statistically significant differences were observed between the two groups in age (31 [29–33] vs 31 [29–35] years; P = 0.397), pre-pregnancy BMI (20.4 [19.3–22.0] vs 20.2 [19.1–21.7] kg/m2; P = 0.801), nulliparity (72.1% vs 63.1%; P = 0.317), multiple cysts (11.6% vs 9.2%; P = 0.636) or educational attainment (university degree or higher: 90.7% vs 89.2%; P = 0.765). However, the proportion of patients with rASRM stage III–IV disease was significantly higher in Group 1 than in Group 2 (80.2% vs 36.9%; P < 0.001), and the median maximum cyst diameter at cesarean section was also greater in Group 1 (4.0 [IQR 3.0–5.0] cm vs 3.0 [IQR 2.0–4.0] cm; P < 0.001). These findings suggest a greater disease burden in Group 1, potentially reflecting a longer disease duration and persistent lesion presence. Changes in OE During Pregnancy in Group 1 In Group 1 (n = 86), the initial maximum diameter of OE identified prior to conception or during early pregnancy ranged from 1.2 to 9.3 cm, with a median of 3.7 cm (IQR 2.9–4.3). Cysts measuring ≥5 cm were observed in 17 patients (19.8%), whereas 69 (80.2%) had cysts measuring <5 cm. Intraoperative measurements at the time of cesarean section revealed a median maximum cyst diameter of 4.0 cm (IQR 3.0–5.0). The difference between the pre-pregnancy and intraoperative measurements ranged from −4.0 to +4.0 cm, with a median change of 0 cm (IQR −0.4 to 0.8 cm), indicating an overall trend toward stability. Based on changes in cyst size, 52 (60.5%) cysts remained stable (change <1 cm), 19 (22.1%) increased in size (≥1 cm increase), and 15 (17.4%) decreased in size (≥1 cm decrease). These findings indicate that the majority of OEs remained stable throughout pregnancy, whereas approximately one-fifth exhibited a clinically significant increase. Maternal and Neonatal Outcomes Delivery Outcomes All 151 patients delivered via cesarean section. The median gestational age at delivery was 39 weeks (IQR 37.5–39.0), with a mean of 38.15 ± 1.59 weeks. Preterm birth (<37 weeks) occurred in 14 patients (9.3%), including seven in Group 1 (8.1%) and seven in Group 2 (10.8%), with no significant intergroup difference (P = 0.581). Of the 14 preterm births, only one was directly attributable to OE. This patient underwent emergency cesarean section at 36 weeks of gestation because of cyst rupture and indeterminate cyst characteristics, and intraoperative histopathology confirmed a benign endometrioma. The remaining 13 cases of preterm birth were unrelated to OE and were attributed to placenta previa, twin pregnancy, preeclampsia, intrauterine infections, and other obstetric causes. A total of 160 neonates (including nine second-born twins) were delivered, with a mean birth weight of 3145 ± 521 g. No significant differences were observed between Group 1 and Group 2 in terms of birth weight (3189 ± 515 g vs 3091 ± 526 g; P = 0.238) or gestational age at delivery (39 [38–39] vs 38 [37–39] weeks; P = 0.252), suggesting that pre- or early-pregnancy identification of OE was not associated with adverse effects on neonatal birth weight or gestational age at delivery. In seven patients (4.6%), the indication for cesarean section was cyst-related (rupture, suspected malignant transformation, or patient request owing to cyst-related concerns), while the remaining 144 cesarean sections were performed for standard obstetric indications. Adverse Outcomes Using the composite endpoint of cyst rupture, suspected malignant transformation, and cyst-related iatrogenic preterm birth to define adverse outcomes, 22 patients (14.6%) were identified. Cyst rupture occurred in 20 patients (13.2%) and was the most common cyst-related complication, with 13 cases in Group 1 (15.1%) and seven in Group 2 (10.8%); the intergroup difference was not statistically significant (P = 0.435). Among the 20 patients with rupture, eight (40.0%) demonstrated stromal decidualization on postoperative histopathology, suggesting that decidualization may increase cyst wall fragility. Malignant cysts during pregnancy were highly suspected in three patients (2.0%). In Case 1, rapid cyst enlargement and abnormal ultrasound findings (multilocular septations, hyperechoic nodules, and papillary projections) during pregnancy raised suspicion of malignant transformation, prompting cesarean section. The final histopathological examination confirmed the presence of a borderline ovarian tumor coexisting with endometriosis. In Case 2, a patient with bilateral ovarian cysts, an endometrioma was detected on one side and an ovarian carcinoid tumor arising from a mature teratoma on the contralateral side. In Case 3, cyst rupture occurred at 6 months of gestation following defecation, which resolved with conservative management. The cyst subsequently enlarged rapidly, and a second rupture occurred at 36 weeks of gestation following defecation. Emergency cesarean section was performed because of cyst rupture and because malignancy could not be excluded, intraoperative pathological examination was performed, and OE was confirmed. This case represented the only preterm birth directly attributable to OE and constituted a composite of three adverse outcomes, including rupture, suspected malignancy, and preterm birth. No cases of malignant OE transformations were observed in this cohort. Decidualization Postoperative histopathology revealed stromal decidualization of the cyst in 26 patients (17.2%), representing a pregnancy-specific transformation of the ectopic endometrial stroma. Decidualization was identified in 14 patients in Group 1 (16.3%) and 12 in Group 2 (18.5%), with no significant difference in incidence between the two groups (P = 0.725). A significant association was observed between decidualization and cyst rupture. Rupture occurred in eight of 26 patients (30.8%) in the decidualization-positive group, compared with only 12 of 125 (9.6%) in the decidualization-negative group. Fisher’s exact test demonstrated that decidualization was significantly associated with an increased risk of rupture (OR = 4.185; 95% CI: 1.504–11.648; P = 0.008). No statistically significant difference in maximum intraoperative cyst diameter was observed between the decidualization-positive and decidualization-negative groups (P > 0.05), indicating that decidualization was not significantly correlated with cyst size. Risk Factor Analysis for Adverse Outcomes Associated with OE in Pregnancy Adverse outcomes (cyst rupture, suspected malignant transformation, and cyst-related iatrogenic preterm birth) were used as the dependent variable (occurrence = 1, nonoccurrence = 0). A total of 22 patients (14.6%) experienced adverse outcomes. Univariable logistic regression analysis was performed for 11 candidate variables, including age (years, continuous), pre-pregnancy BMI (kg/m2, continuous), gravidity (continuous), nulliparity (yes vs no), mode of conception (in vitro fertilization [IVF] vs spontaneous conception), timing of cyst detection (Group 1 vs Group 2), laterality (bilateral vs unilateral), maximum intraoperative cyst diameter (≥5 cm vs <5 cm), rASRM stage (III–IV vs I–II), decidualization (positive vs negative), and pre-pregnancy treatment history (yes vs no). In addition, cyst enlargement during pregnancy (yes vs no) was analyzed separately in the Group 1 subgroup (n = 86) to assess its association with adverse outcomes. Univariable analysis revealed that rASRM stage III–IV (OR = 4.707, 95% CI: 1.326–16.706, P = 0.017) and decidualization (OR = 3.524, 95% CI: 1.295–9.591, P = 0.014) were significantly associated with adverse outcomes (P 0.10). The complete univariable analysis is presented in Table 2. | Table 2 Univariable Logistic Regression Analysis of Candidate Risk Factors for Adverse Outcomes Associated with Ovarian Endometrioma During Pregnancy | | Table 3 Multivariable Logistic Regression Analysis of Independent Risk Factors for Adverse Outcomes Associated with Ovarian Endometrioma During Pregnancy | Two variables that met the screening threshold were entered into a multivariable logistic regression model. Given that the number of composite adverse outcome events was 22, and in accordance with the principle of ≥10 EPV,11 the multivariable model could accommodate a maximum of two independent variables (EPV = 22/2 = 11) to ensure stability of regression coefficient estimates. Multivariable logistic regression analysis demonstrated that rASRM stage III–IV (aOR = 5.784, 95% CI: 1.549–21.596, P = 0.009) and decidualization (aOR = 4.530, 95% CI: 1.541–13.316, P = 0.006) were independent risk factors for adverse outcomes associated with OE during pregnancy (Table 3). Supplementary Analysis An intraoperative cyst diameter ≥5 cm was not significantly associated with adverse outcomes in the univariable analysis (OR = 1.353, P = 0.564) and was therefore not included in the multivariable model. In the Group 1 subgroup (n = 86), cyst enlargement during pregnancy was not significantly associated with adverse outcomes (OR = 2.036, 95% CI, 0.600–6.912; P = 0.254), which may be attributable to the limited sample size.

Through a 9-year retrospective analysis of 151 histopathologically confirmed cases of OE during pregnancy, this study demonstrated that the majority (60.5%) of OEs remained stable in size throughout gestation, whereas approximately one-fifth exhibited significant enlargement. rASRM stage III–IV disease and decidualization were identified as independent risk factors for adverse outcomes, whereas absolute cyst size was not a reliable predictor. These findings suggest that risk stratification for OE during pregnancy should extend beyond cyst dimensions to incorporate disease stage and histological status. Regarding the natural evolution of OE during pregnancy, the enlargement rate observed in the present study (22.1%) was comparable to the 20% rate reported by Ueda et al,5 whereas the stability rate (60.5%) exceeded the 28% unchanged proportion in their series. In contrast, the proportion of cysts that decreased in size (17.4%) was substantially lower than that reported by Bean et al12 (85% shrinkage) and Ueda et al5 (52% shrinkage). These discrepancies may be explained by three factors. First, both Bean et al12 and Ueda et al5 extended the follow-up into the postpartum period, during which withdrawal of estrogen and progesterone may promote cyst regression. In contrast, the present study used intraoperative measurements during cesarean section as the endpoint, potentially underestimating the proportion of cysts that would ultimately regress postpartum. Second, Bean et al12 did not define a numerical threshold for shrinkage, whereas both the present study and Ueda et al5 applied a threshold of ≥1 cm. Third, 80.2% of patients in Group 1 had rASRM stage III–IV disease; the more extensive fibrosis associated with advanced disease may reduce responsiveness to hormonal withdrawal. Differences in sample size and study populations may also have contributed to the discrepant findings. Notably, 43.0% of patients (Group 2) were not diagnosed until cesarean section, consistent with the well-documented diagnostic delay associated with endometriosis.2 The inclusion of Group 2 adds four specific contributions. First, it quantifies the real-world gap in antenatal detection: 65 of the 151 histopathologically confirmed cases (43.0%) had no antenatal sonographic diagnosis of OE, showing that obstetric ultrasound focused on fetal biometry does not reliably survey the adnexa and that a negative scan does not exclude OE — consistent with the diagnostic delay well documented in endometriosis.2 Second, the group comparison shows that antenatal detection is not random but tracks disease severity: Group 1 had a markedly higher prevalence of rASRM stage III–IV disease (80.2% vs 36.9%) and a larger median cyst diameter (4.0 vs 3.0 cm), so larger, more advanced lesions entered surveillance whereas smaller lesions whose advanced stage was determined by adhesions escaped detection. Third, Group 2 serves as a natural comparison group that offsets the selection bias inherent in cohorts restricted to antenatally diagnosed cases, which by design exclude precisely the clinically silent cysts. Fourth, under a uniform policy of cesarean delivery with intraoperative assessment, the timing of diagnosis made no measurable difference to neonatal outcomes, birth weight and gestational age at delivery being comparable between groups. Together, these observations argue for more systematic adnexal assessment during obstetric ultrasound and suggest that prospective identification of OE may refine risk stratification even where it does not alter perinatal outcomes. Several conclusions follow from the follow-up structure of this study, and its limitations in this respect should be stated plainly. First, follow-up ended at delivery, the intraoperative measurement at cesarean section being the terminal observation; because postpartum involution was not assessed, the 22.1% enlargement rate must not be extrapolated to long-term cyst trajectory, existing evidence indicating that many endometriomas regress after delivery once hormonal support is withdrawn.5,12 The observed enlargement is therefore an intra-pregnancy phenomenon rather than a durable change in lesion behavior. Second, with no serial imaging, the transitions between stability, enlargement and regression could not be timed and the gestational window in which they occur remains unknown, precluding inference about the optimal timing of surveillance. Third, decidualization was ascertained solely from postoperative histopathology, with no systematically captured antenatal imaging correlate; when decidualization develops, whether it evolves sonographically in a recognizable sequence, and whether it is reversible therefore remain undetermined — a clinically consequential gap, since decidualization proved an independent risk factor for adverse outcomes. Fourth, the absence of postpartum imaging likewise precluded examining its relationship to subsequent cyst resolution. These gaps define the agenda for future work: a prospective cohort with standardized serial ultrasonography at defined gestational intervals and systematic postpartum imaging is needed to establish the true natural history of OE in pregnancy and to test whether sonographically defined decidualization can serve as an actionable antenatal marker. Decidualization represents an important phenomenon in the gestational evolution of OE and may constitute a critical link between morphological changes and clinical complications. On imaging, decidualized OE typically manifests as round, smooth-surfaced, hypervascular papillary projections with ground-glass cyst fluid, features that may overlap with those of malignant neoplasms. The smooth, rounded contour of decidualized projections, as opposed to the irregular surface of malignant excrescences, may represent the principal differentiating feature.13,14 In the present study, the incidence of decidualization (17.2%) was comparable to the 11.8% reported by Pateman et al,15 whereas the higher rate (50%) reported by Bean et al12 may have been influenced by the concurrent inclusion of deep nodular endometriosis. Decidualization was significantly associated with an increased risk of complications. The rupture rate was 30.8% in the decidualization-positive group vs 9.6% in the decidualization-negative group, while the odds of adverse outcomes were approximately 3.5-fold higher. The overall rupture rate of 13.2% was higher than the 4% reported by Ueda et al.5 Beyond the predominance of stage III–IV disease, decidualization may contribute to structural remodeling of the cyst wall and increased tissue fragility. Endometriosis during pregnancy is also associated with the risk of spontaneous hemoperitoneum,16 with Brosens et al17 identifying endometriosis as a major risk factor. No cases of spontaneous hemoperitoneum were observed in our cohort; however, the potentially severe consequences of this complication underscore the importance of clinical vigilance. Importantly, decidualization-associated papillary projections predominantly develop during the first trimester and resolve spontaneously postpartum,12 suggesting that intramural nodules detected during pregnancy should be interpreted cautiously, with decidualization considered in the differential diagnosis before assuming malignant transformation.13 Magnetic resonance imaging may serve as a useful adjunct when the diagnosis remains uncertain.18 With respect to delivery outcomes, the preterm birth rate was 9.3% (14/151), and only one case was directly attributable to OE. A meta-analysis by Busnelli et al19 demonstrated that only severe endometriosis (rASRM stage III–IV) was significantly associated with increased risk of placenta previa (OR = 6.61) and preterm birth (OR = 2.76), whereas Breintoft et al20 reported a more modest association with preterm birth (OR = 1.46). Similarly, Farland et al,21 in a cohort of 116,429 women, reported a small but significant increase in risk (risk ratio = 1.16). The direction of these associations is consistent with our findings and supports the view of Vercellini et al22 that severe endometriosis may be more strongly associated with adverse pregnancy outcomes than milder disease. No significant intergroup differences were observed in birth weight or gestational age at delivery, suggesting that the timing of OE diagnosis did not appreciably influence neonatal outcomes under standardized cesarean-based management. These findings may be interpreted from several perspectives. First, Taylor et al2 highlighted substantial progesterone resistance in endometriosis, with approximately one-third of patients exhibiting no response to progestogen therapy. This hormonal resistance may partly explain why 60.5% of OEs remained stable rather than undergoing universal regression during pregnancy. The 22.1% rate of cyst enlargement may be related to incomplete decidualization, potentially involving aberrant PI3K/AKT signaling.4,8 Some authors have proposed that this initial enlargement may represent an intermediate stage preceding subsequent regression.8 Second, patients with rASRM stage III–IV disease typically harbor extensive pelvic adhesions, which may render the cyst more susceptible to traction as the uterus enlarges during pregnancy. This may partly explain why disease stage was a stronger predictor than cyst size per se. Finally, the absence of confirmed malignant transformation despite three episodes of clinical suspicion has four plausible explanations. First, the observation window was confined to pregnancy and spanned less than one year, whereas endometriosis-associated ovarian cancer typically develops over decades. Second, the absence of malignant transformation must be interpreted within the same time frame. Malignant transformation of ovarian endometrioma is a long-latency event, and a gestational window of observation cannot exclude it; the three patients classified as having suspected malignant transformation were subsequently shown to have a borderline tumor coexisting with endometriosis, a carcinoid arising from a mature teratoma, and a ruptured benign endometrioma, respectively. The term “suspected” in our composite endpoint therefore reflects diagnostic uncertainty at the time of surgery rather than confirmed malignant change. Third, 151 patients are underpowered for an event of this rarity, so the absence of cases indicates a low risk rather than a nonexistent one. Fourth, although serum tumor markers and pelvic magnetic resonance imaging were used as ancillary investigations in this study, the physiological changes of pregnancy reduce the specificity of tumor markers, so subtle features of early transformation may still have gone unrecognized, particularly in Group 2, whose cysts were never imaged before delivery. This is nonetheless no reassurance: the endometrioma subtype confers the highest risk of endometriosis-associated ovarian cancer,23–25 Barnard et al26 reported a nearly 19-fold elevation in type I ovarian cancer risk, and the clonal progression pathway from benign endometriosis to malignancy is increasingly elucidated.27–29 The converse error weighs equally, since decidualized endometriomas often acquire papillary projections and hypervascularity that mimic malignancy and regress spontaneously after delivery,12–14 so over-investigation and radical intervention carry their own risk to the pregnancy. Suspicion of malignancy in pregnancy therefore warrants structured confirmatory assessment rather than immediate radical surgery, with surveillance continuing beyond the puerperium. Based on these findings, we propose a risk-stratified management framework that warrants prospective validation, outlined as follows: (1) Low-risk (rASRM stage I–II, no sonographic signs of decidualization, cyst <5 cm): routine antenatal ultrasound surveillance is recommended with cyst assessment performed once in each trimester; (2) Intermediate risk (rASRM stage I–II with decidualization or rASRM stage III–IV without decidualization): increased ultrasound surveillance frequency is recommended, with particular attention to signs of cyst rupture (acute abdominal pain, changes in cyst morphology); and (3) High-risk (rASRM stage III–IV with decidualization or rapid cyst enlargement over a short interval): close ultrasound follow-up is recommended, with multidisciplinary team management involving obstetricians and gynecologic oncologists, formulation of a planned early delivery strategy, and preparation for complete cyst excision at the time of cesarean section. This stratified framework represents expert opinions informed by the present data and has not been prospectively validated. It is intended to generate hypotheses for future clinical trials rather than serve as an evidence-based guideline. The strengths of this study are as follows: (1) a relatively large sample size (n = 151) derived from a complete dataset of 102,071 deliveries over 9 years, representing one of the larger cohorts in this field to date; (2) histopathological confirmation of all cases, thereby minimizing the diagnostic bias inherent to imaging-based studies; and (3) systematic multidimensional analysis encompassing dynamic cyst changes, decidualization, and rASRM staging, providing comprehensive data to support clinical risk assessment. Some limitations should also be acknowledged. First, the retrospective design may have introduced selection bias because all included patients delivered by cesarean section, and data on patients with OE who delivered vaginally were unavailable. Consequently, milder cases that permitted vaginal delivery may have been systematically underrepresented, potentially resulting in overestimation of the observed rates of adverse outcomes (rupture, 13.2%; composite endpoint, 14.6%) and rASRM stage III–IV disease (61.6%). Accordingly, the absolute risks reported here should be interpreted as potential upper-bound estimates for the general OE-in-pregnancy population. The direction and magnitude of potential bias in the estimated associations (aORs) are less certain because the relationships between rASRM stage or decidualization and adverse outcomes may operate similarly within the cesarean delivery subpopulation. Therefore, external validation in an unselected cohort is essential. Second, the absence of long-term postpartum follow-up precluded assessment of postpartum OE resolution and recurrence rates, and no serial antenatal imaging was available to define the trajectory of individual cysts. Third, cyst size was ascertained by two methodologically distinct approaches, sonographic measurement before or during early pregnancy and direct intraoperative measurement at cesarean section, which are not strictly equivalent: ultrasound is influenced by abdominal wall thickness, transducer pressure, probe orientation and operator experience, whereas intraoperative measurement is affected by specimen decompression, adhesiolysis and the altered pelvic anatomy of the gravid uterus. A systematic tendency for intraoperative values to exceed sonographic ones cannot be excluded and may have inflated the apparent enlargement rate, and in ruptured cysts the decompressed specimen no longer allows the original diameter to be reconstructed, so these size data are approximate. No validation of the two modalities against each other was performed and inter-observer reproducibility was not assessed, the analysis relying on archived images and reports. Fourth, the single-center design limits generalizability of the findings, necessitating validation in other populations. Fifth, the sample size and relatively small number of adverse outcome events (n = 22) limited the statistical power of the multivariable logistic regression model. Although the model adhered to the EPV ≥10 principle, certain subgroup analyses, such as those involving the association between dynamic cyst changes and adverse outcomes, may have been underpowered to detect clinically meaningful differences. Sixth, the rASRM staging system does not adequately capture deep infiltrating endometriosis and has recognized limitations in its associations with pain, fertility outcomes, and prognosis. The #Enzian classification may provide complementary information; however, a comparative analysis was not feasible in the present study.30–32

In this 9-year cohort of 151 histopathologically confirmed cases, the gestational course of OE was heterogeneous: most cysts remained stable, whereas approximately one-fifth enlarged. rASRM stage III–IV disease and decidualization were independent risk factors for adverse outcomes, whereas absolute cyst size was not. Enhanced antenatal surveillance, with early identification of complications and multidisciplinary management when indicated, is therefore recommended for patients with rASRM stage III–IV disease or sonographic features suggestive of decidualization. Abbreviations 95% CI, 95% confidence interval; aOR, adjusted odds ratio; BMI, body mass index; EPV, events-per-variable; IQR, interquartile range; OE, ovarian endometrioma; OR, odds ratio; rASRM, revised American Society for Reproductive Medicine staging system; SD, standard deviation. Data Sharing Statement The data supporting the findings of this study are available from the corresponding author upon reasonable request; the dataset contains potentially identifiable clinical information and is not publicly shared in accordance with institutional policy. Acknowledgments The authors thank the Department of Obstetrics at Hangzhou Women’s Hospital for supporting data collection and the patients who participated in this study. Author Contributions All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Disclosure The authors report no conflicts of interest in this work.

1. As-Sanie S, Mackenzie SC, Morrison L, et al. Endometriosis: a review. JAMA. 2025;334:64–12. 2. Taylor HS, Kotlyar AM, Flores VA. Endometriosis is a chronic systemic disease: clinical challenges and novel innovations. Lancet. 2021;397:839–852. 3. Zondervan KT, Becker CM, Missmer SA. Endometriosis. N Engl J Med. 2020;382:1244–1256. doi:10.1056/NEJMra1810764 4. Leone Roberti Maggiore U, Ferrero S, Mangili G, et al. A systematic review on endometriosis during pregnancy: diagnosis, misdiagnosis, complications and outcomes. Hum Reprod Update. 2016;22:70–103. 5. Ueda Y, Enomoto T, Miyatake T, et al. A retrospective analysis of ovarian endometriosis during pregnancy. Fertil Steril. 2010;94:78–84. 6. Beecham CT. Surgical treatment of endometriosis with special reference to conservative surgery in young women. J Am Med Assoc. 1949;139:971. 7. McArthur JW, Ulfelder H. The effect of pregnancy upon endometriosis. Obstet Gynecol Surv. 1965;20:709–733. 8. Leeners B, Damaso F, Ochsenbein-Kölble N, et al. The effect of pregnancy on endometriosis-facts or fiction? Hum Reprod Update. 2018;24:290–299. 9. Van Holsbeke C, Van Calster B, Guerriero S, et al. Endometriomas: their ultrasound characteristics. Ultrasound Obstet Gynecol. 2010;35:730–740. 10. Canis M, Donnez JG, Guzick DS, et al. Revised American Society for Reproductive Medicine classification of endometriosis: 1996. Fertil Steril. 1997;67:817–821. 11. Peduzzi P, Concato J, Kemper E, et al. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49:1373–1379. 12. Bean E, Knez J, Setty T, et al. Natural history of endometriosis in pregnancy: ultrasound study of morphology of deep endometriosis and ovarian endometrioma. Ultrasound Obstet Gynecol. 2023;62:585–593. 13. Mascilini F, Moruzzi C, Giansiracusa C, et al. Imaging in gynecological disease. 10: clinical and ultrasound characteristics of decidualized endometriomas surgically removed during pregnancy. Ultrasound Obstet Gynecol. 2014;44:354–360. 14. Yin M, Wang T, Li S, et al. Decidualized ovarian endometrioma mimicking malignancy in pregnancy: a case report and literature review. J Ovarian Res. 2022;15:33. 15. Pateman K, Moro F, Mavrelos D, et al. Natural history of ovarian endometrioma in pregnancy. BMC Womens Health. 2014;14:128. 16. Schreurs AMF, Overtoom EM, de Boer MA, et al. Spontaneous haemoperitoneum in pregnancy: nationwide surveillance and Delphi audit system. BJOG. 2023;130:1620–1628. 17. Brosens IA, Lier MC, Mijatovic V, et al. Severe spontaneous hemoperitoneum in pregnancy may be linked to in vitro fertilization in patients with endometriosis: a systematic review. Fertil Steril. 2016;106:692–703. 18. Thomassin-Naggara I, Dolciami M, Chamie LP, et al. ESUR consensus MRI for endometriosis: protocol, lexicon, and compartment-based analysis. Eur Radiol. 2025;35:7272–7286. 19. Busnelli A, Di Simone N, Somigliana E, et al. Untangling the independent effect of endometriosis, adenomyosis, and ART-related factors on maternal, placental, fetal, and neonatal adverse outcomes: results from a systematic review and meta-analysis. Hum Reprod Update. 2024;30:751–788. 20. Breintoft K, Pinnerup R, Henriksen TB, et al. Endometriosis and risk of adverse pregnancy outcome: a systematic review and meta-analysis. J Clin Med. 2021;10:667. 21. Farland LV, Prescott J, Sasamoto N, et al. Endometriosis and risk of adverse pregnancy outcomes. Obstet Gynecol. 2019;134:527–536. 22. Vercellini P, Viganò P, Bandini V, et al. Association of endometriosis and adenomyosis with pregnancy and infertility. Fertil Steril. 2023;119:727–740. 23. Younis JS. Endometriosis-associated ovarian cancer: what are the implications for women with intact endometrioma planning for a future pregnancy? A reproductive clinical outlook. Biomolecules. 2022;12:1721. doi:10.3390/biom12111721 24. Navarro R, Poder L, Sun D, et al. Endometriosis in pregnancy. Abdom Radiol. 2020;45:1741–1753. 25. Kvaskoff M, Mahamat-Saleh Y, Farland LV, et al. Endometriosis and cancer: a systematic review and meta-analysis. Hum Reprod Update. 2021;27:393–420. 26. Barnard ME, Farland LV, Yan B, et al. Endometriosis typology and ovarian cancer risk. JAMA. 2024;332:482–489. 27. Sorrentino F, Nappi L, Vona L, et al. From endometriosis to endometriosis-associated ovarian cancer: molecular mechanisms, risk stratification and clinical implications. Cancers. 2026;18:1233. 28. Steinbuch SC, Lüß AM, Eltrop S, et al. Endometriosis-associated ovarian cancer: from molecular pathologies to clinical relevance. Int J Mol Sci. 2024;25:4306. 29. Kłodnicka K, Michalska A, Januszewski J, et al. From inflammation to malignancy: the link between endometriosis and gynecological cancers. Int J Mol Sci. 2025;26:11816. 30. Hudelist G, Valentin L, Saridogan E, et al. What to choose and why to use - a critical review on the clinical relevance of rASRM, EFI and Enzian classifications of endometriosis. Facts Views Vis Obgyn. 2021;13:331–338. 31. Keckstein J, Saridogan E, Ulrich UA, et al. The #Enzian classification: a comprehensive non-invasive and surgical description system for endometriosis. Acta Obstet Gynecol Scand. 2021;100:1165–1175. 32. Eichinger T, Oppelt P, Lastinger J, et al. Current endometriosis classifications (rASRM, #Enzian, AAGL2021) and their correlation with operative time. Geburtshilfe Frauenheilkd. 2025;86:201–207. German. © 2026 The Author(s). This work is published and licensed by Dove Medical Press Limited. The full terms of this license are available at https://www.dovepress.com/terms and incorporate the Creative Commons Attribution - Non Commercial (unported, 4.0) License. By accessing the work you hereby accept the Terms. Non-commercial uses of the work are permitted without any further permission from Dove Medical Press Limited, provided the work is properly attributed. For permission for commercial use of this work, please see paragraphs 4.2 and 5 of our Terms.