Analysis of factors influencing ovarian reserve in patients with endometriosis

For clinicians managing endometriosis: AFC outperforms AMH in detecting localized ovarian damage, with cyst size and sliding signs serving as key predictors of diminished ovarian reserve. This supports AFC as the preferred biomarker for fertility preservation counseling in untreated endometriosis patients.

This study was a retrospective, single-center clinical cohort study conducted in accordance with the principles of the Declaration of Helsinki. Ethical approval was obtained from the Institutional Review Board of the Third Affiliated Hospital of Sun Yat-sen University (Approval NO: EY AF/SC-02-06-01/03.0). Data were collected between January and March 2024. As this was a retrospective study, the requirement for informed consent was waived. The study was reported following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for observational studies. The sample included patients diagnosed with EMS who were admitted to the Department of Gynecology at the Third Affiliated Hospital of Sun Yat-sen University, either as outpatients or inpatients. Data were collected from cases diagnosed between January 2022 and December 2023, in which patients underwent either DIE ultrasound assessment or surgical procedures. This study aimed to provide a descriptive overview of the clinical characteristics and did not involve any statistical inference. Consequently, formal sample size calculation was not required, aligning with the guidelines for descriptive epidemiological studies [ 6 ]. The surgical procedures were carried out in accordance with the standardized plan [ 7 ]. The ultrasound evaluation followed [ 8 ] the Chinese expert consensus on sonographic evaluation of endometriosis . The form of ultrasound examination should be selected according to the patient's will. Transvaginal ultrasound is recommended for patients as much as possible. For those who have no sexual life or do not receive a transvaginal ultrasound, transrectal ultrasound should be used instead. The following parameters were collected: patient age and body mass index (BMI). For medical history, information on pregnancy history, recurrence history, and consistency of previous medication use was recorded. Uterine comorbidities, including endometrial polyps, adenomyosis, and uterine fibroids, were documented. During ultrasound examinations, we recorded whether patients had DIE, involvement of unilateral or bilateral ovaries, laterality of unilateral ovarian endometriomas (left or right), cyst size, and AFC. The ultrasound sliding sign included assessments of uterine posterior wall sliding, rectouterine pouch sliding, uterine anterior wall sliding, and bladder peritoneal reflection sliding. Serum AMH levels were also collected. For surgical findings, the degree of obliteration of the pouch of Douglas was classified as no obliteration, partial obliteration, or complete obliteration. Additionally, operative time, surgical procedures performed, endometriosis fertility index (EFI) scores, and ASRM scores and staging were recorded. Data were analyzed using SPSS software version 27.0. Continuous variables were tested for normality. Normally distributed variables were presented as mean ± standard deviation (SD), while non-normally distributed variables were expressed as median (interquartile range, IQR). Categorical variables were presented as frequencies and percentages. Differences in continuous variables were analyzed using the t -test for normally distributed data and the Mann–Whitney U test for non-normally distributed data. Wilcoxon signed-rank tests were performed for paired variables. Categorical variables with cell counts > 5 were analyzed using the chi-square test, while Fisher’s exact test was applied for variables that did not meet this criterion. Spearman’s rank correlation coefficient was used to evaluate correlations between variables. Statistical significance was defined as p  < 0.05. In clinical practice, AMH and AFC are often detected asynchronously. To reflect the real scenarios and solve the problem of selection bias at the same time, this study adopted a dual-track strategy. The initial cohort was retained for preliminary analysis, and sensitivity analysis was conducted on the subgroups with complete data simultaneously.

Between January 2002 and December 2023, a cohort of 1,199 patients diagnosed with EMS was identified, all presenting with OEM. Notably, 267 cases (22.3%) exhibited concomitant DIE (Fig.  1 A). Among them, 864 underwent surgical exploration, 524 underwent ultrasound evaluation of DIE, 337 underwent AFC measurement, and 676 underwent preoperative AMH detection (Fig.  1 B). Fig. 1 A Population distribution of OEM. B Distribution map of patient examination items A Population distribution of OEM. B Distribution map of patient examination items The ages of all patients ranged from 18 to 54 years. Among them, 838 had unilateral ovarian endometriomas, with a median age of 30 years (IQR, 26–35), and 361 had bilateral ovarian endometriomas, with a median age of 30 years (IQR, 27–36). A total of 921 patients provided height and weight data, with a median BMI of 20.70 kg/m 2 (IQR, 18.82–22.66). The BMI levels of patients with EMS were generally lower than average. Pearson correlation analyses were performed to evaluate the relationship between age and AFC (left, right, and total) as well as AMH levels. The results showed significant negative correlations between age and the following parameters: left AFC ( R  = −0.210, P  < 0.001), right AFC ( R  = −0.162, P  = 0.003), total AFC ( R  = −0.242, P  < 0.001), and AMH ( R  = −0.432, P  < 0.001) (Table  1 ). Table 1 Relationship between age and AFC/AMH Items Number of effective cases Age R P Left AFC 337 −0.210 < 0.001* Right AFC 337 −0.162 0.003* Total AFC 337 −0.242 < 0.001* AMH 676 −0.432 < 0.001* * P  < 0.05 Relationship between age and AFC/AMH * P  < 0.05 Nonparametric tests were conducted to compare left AFC, right AFC, total AFC, and AMH between patients with unilateral and bilateral ovarian cysts (Table  2 ). The results showed no significant difference in left AFC between unilateral and bilateral cyst groups ( P  = 0.254). However, cyst laterality significantly influenced right AFC ( P  < 0.001) and total AFC ( P  < 0.001), with unilateral cysts demonstrating significantly higher right AFC and total AFC compared to bilateral cysts. No significant difference was observed in AMH levels between unilateral and bilateral cysts ( P  = 0.942). Table 2 Relationship between unilateral/bilateral cysts and AFC/AMH Items No Unilateral Bilateral Z P No Description No Description Left AFC 337 218 6 (4, 12) 119 7 (4, 9) −1.141 0.254 Right AFC 337 218 9 (6, 13) 119 5 (3, 9) −4.869 < 0.001* Total AFC 337 218 16 (12, 21) 119 12 (8.25, 16) −4.094 < 0.001* AMH 676 444 2.79 (1.76, 4.62) 232 2.97 (2.07, 4.43) −0.072 0.942 * P  < 0.05 Relationship between unilateral/bilateral cysts and AFC/AMH * P  < 0.05 In patients with unilateral ovarian cysts, AFC in the healthy ovary was significantly higher than in the affected ovary ( Z  = −9.786, P  < 0.001) (Table  3 ). These findings suggest that AFC can serve as a reliable marker of fertility potential in patients with unilateral ovarian cysts. Table 3 Relationship between healthy and affected ovaries in AFC Item Healthy Affected Z P No Description No Description AFC 218 12 (8, 13) 218 5 (3, 9) −9.786 < 0.001* * P  < 0.05 Relationship between healthy and affected ovaries in AFC * P  < 0.05 Pearson correlation analyses were performed to assess the relationship between the largest cyst diameter on the left and right sides and AFC (left, right, and total) as well as AMH levels (Table  4 ). The results revealed the following: Left AFC was significantly negatively correlated with the largest cyst diameter on the left side ( R  = −0.448, P  < 0.001). Right AFC was significantly negatively correlated with the largest cyst diameter on the right side ( R  = −0.486, P  < 0.001). Total AFC showed a significant negative correlation with the largest cyst diameter on both sides (left: R  = −0.181, P  = 0.004; right: R  = −0.286, P  < 0.001). No significant correlation was found between AMH levels and the largest cyst diameter on either side. Table 4 Relationship between cyst size and AFC/AMH Items Largest cyst diameter on the left side Largest cyst diameter on the right side No R P No R P Left AFC 250 −0.448 < 0.001* 206 0.008 0.910 Right AFC 250 0.103 0.105 206 −0.486 < 0.001* Total AFC 250 −0.181 0.004* 206 −0.286 < 0.001* AMH 492 −0.028 0.536 416 −0.020 0.682 * P  < 0.05 Relationship between cyst size and AFC/AMH * P  < 0.05 Pearson correlation analyses were conducted to assess the relationship between AMH levels and AFC (left, right, and total) (Table  5 ). The results demonstrated significant positive correlations between AMH and left AFC ( P  < 0.001), right AFC ( P  < 0.001), and total AFC ( P  < 0.001), with a moderate correlation between AMH and total AFC ( R  = 0.543). Table 5 Relationship between AMH and AFC Items Number of effective cases AMH R P Left AFC 240 0.430 < 0.001* Right AFC 240 0.394 < 0.001* Total AFC 240 0.543 < 0.001* * P  < 0.05 Relationship between AMH and AFC * P  < 0.05 Relevant analysis showed that the total AFC was correlated with five factors: postoperative recurrence ( P  = 0.038), whether there was adenomyosis ( P  = 0.007), posterior uterine wall sliding sign ( P  < 0.001), hysterorectal depression sliding sign ( P  = 0.036), and the maximum diameter of deep nodules ( P  = 0.023). However, AMH was only correlated with the first four. It was not related to the maximum diameter of deep nodules ( P  = 0.058), and was also related to pregnancy history ( P  < 0.001), combined uterine fibroids ( P  < 0.001), and BMI ( P  < 0.001). The influencing factors of AMH are more mixed (Table  6 ). Table 6 Relationships between other factors and AFC/AMH Items Left AFC Right AFC Total AFC AMH No Z/R P No Z/R P No Z/R P No Z/R P Pregnancy history No 262 −1.101 0.271 262 −0.582 0.56 262 −0.959 0.338 484 −0.795 < 0.001 * Yes 75 75 75 192 Postoperative recurrence No 321 −1.273 0.203 321 −2.247 0.025 * 321 −2.077 0.038 * 642 −2.203 0.028 * Yes 16 16 16 34 Consistent medication use No 319 −1.493 0.135 319 −0.627 0.53 319 −0.579 0.563 654 −0.82 0.412 Yes 18 18 18 22 Associated endometrial polyps No 270 −0.383 0.702 270 −0.494 0.621 270 −0.086 0.931 557 −1.745 0.081 Yes 67 67 67 119 Associated adenomyosis No 267 −1.313 0.189 267 −3.053 0.002 * 267 −2.677 0.007 * 267 −4.518 < 0.001 * Yes 70 70 70 70 Associated uterine fibroids No 264 −0.199 0.842 264 −1.075 0.282 264 −0.504 0.614 500 −3.504 < 0.001 * Yes 73 73 73 176 Posterior uterine wall sliding sign Positive 152 −3.009 0.003 * 152 −4.126  < 0.001 * 152 −4.324 < 0.001 * 171 −2.927 0.003 * Negative 171 171 171 178 Rectouterine pouch sliding sign Positive 184 −1.727 0.084 184 −1.775 0.076 184 −2.098 0.036 * 200 −1.995 0.046 * Negative 152 152 152 157 Anterior uterine wall sliding sign Positive 314 −0.648 0.517 314 −1.333 0.183 314 −1.289 0.198 337 −0.996 0.319 Negative 9 9 9 12 Bladder peritoneal reflection sliding sign Positive 326 −1.722 0.085 326 −0.847 0.397 326 −1.631 0.103 326 −1.081 0.28 Negative 10 10 10 10 Deep nodules No 167 −0.045 0.964 167 −1.187 0.235 167 −0.542 0.588 181 −1.865 0.062 Yes 169 169 169 176 Largest diameter of deep nodules 169 −0.222 0.004 * 169 −0.051 0.507 169 −0.175 0.023 * 176 −0.143 0.058 BMI 160 −0.153 0.053 160 0.01 0.9 160 −0.095 0.234 509 −0.203 < 0.001 * * P  < 0.05 Relationships between other factors and AFC/AMH * P  < 0.05 The univariate analyses identified several significant factors influencing AFC and AMH: left AFC shared correlations with age, largest cyst diameter on the left side, AMH, posterior uterine wall sliding sign, and largest diameter of deep nodules. Right AFC was correlated with age, laterality (unilateral/bilateral), largest cyst diameter on the right side, AMH, recurrence, adenomyosis, and posterior uterine wall sliding sign. Total AFC was correlated with age, laterality, the largest cyst diameter on both sides, AMH, recurrence, adenomyosis, posterior uterine wall sliding sign, the rectouterine pouch sliding sign, and the largest diameter of deep nodules. AMH shared associations with age, BMI, left AFC, right AFC, total AFC, gravidity, recurrence, adenomyosis, uterine fibroids, posterior uterine wall sliding sign, and rectouterine pouch sliding sign (Table  7 ). Table 7 Multivariate analysis Items Factors OR Standard error P VIF Residual P D-W Adjusted R 2 Left AFC Age −0.024 0.072 0.741 1.142 < 0.001 1.735 0.428 Largest cyst diameter on the left side −0.547 0.132 < 0.001 * 1.076 AMH 0.775 0.14 < 0.001 * 1.212 Posterior uterine wall sliding sign −0.53 0.681 0.439 1.1 Largest diameter of deep nodules −0.802 0.507 0.117 1.146 Right AFC Age 0.008 0.077 0.913 1.216 < 0.001 2.108 0.388 Unilateral/bilateral −0.354 0.593 0.551 1.122 Largest cyst diameter on the right side −0.601 0.109 < 0.001 * 1.048 AMH 0.788 0.126 < 0.001 * 1.216 Recurrence −1.235 1.153 0.286 1.126 Associated adenomyosis −0.171 0.661 0.796 1.151 Posterior uterine wall sliding sign −0.568 0.592 0.339 1.181 Total AFC Age 0.045 0.105 0.673 1.285 < 0.001 2.087 0.448 Largest cyst diameter on the left side 0.407 0.168 0.020 * 1.238 Largest cyst diameter on the right side −0.422 0.145 0.005 * 1.177 AMH 0.667 0.174 < 0.001 * 1.391 Recurrence −1.716 1.577 0.282 1.199 Associated adenomyosis −1.916 0.992 0.06 1.453 Posterior uterine wall sliding sign −1.62 0.985 0.107 1.253 Rectouterine pouch sliding sign 2.29 0.996 0.026 * 1.408 Largest diameter of deep nodules −0.86 0.686 0.217 1.273 AMH Age −0.054 0.05 0.281 1.645 < 0.001 1.977 0.236 BMI −0.136 0.082 0.102 1.327 Pregnancy history −0.134 0.492 0.786 1.427 Recurrence 1.436 0.715 0.047 * 1.077 Left AFC −0.322 0.655 0.624 254.041 Right AFC −0.304 0.648 0.64 269.574 Total AFC 0.465 0.651 0.476 538.527 Associated adenomyosis −0.452 0.501 0.37 1.226 Associated uterine fibroids 0.42 0.518 0.419 1.31 Posterior uterine wall sliding sign 0.3 0.481 0.534 1.897 Rectouterine pouch sliding sign −0.334 0.47 0.478 1.832 * P  < 0.05 Multivariate analysis * P  < 0.05 To account for potential confounding factors, multivariate linear regression analyses were performed, including all significant variables identified in the univariate analyses, while excluding multicollinearity. The results of the multivariate analyses showed the following: Larger cyst diameters on the left side were associated with lower left AFC. Larger cyst diameters on the right side were associated with lower right AFC. Higher AMH levels were associated with higher left AFC, right AFC, and total AFC. AMH levels were influenced only by a history of recurrence. We conducted sensitivity analyses in 240 cases with complete AMH/AFC data. The correlation of the original data analysis enhances the correlation of the Bia group analysis. The consistency of key findings across analytical approaches supports the validity of our conclusions (Table S1 ).

EMS is a common gynecological disorder affecting women of reproductive age, characterized by the presence of endometrial glands and stroma outside the uterine cavity. It is estimated that approximately 10% of women of reproductive age are affected by EMS, equating to around 200 million women globally. EMS is closely associated with infertility, with 20–50% of patients experiencing infertility[ 1 ]. Currently, AMH and antral AFC are widely used clinical biomarkers to assess ovarian reserve, representing hormonal and ultrasound indicators of follicle quantity, respectively. AMH, a dimeric glycoprotein belonging to the transforming growth factor-β superfamily, is produced by granulosa cells of preantral and early antral follicles in women of reproductive age and decreases with age. Importantly, AMH expression remains stable throughout the menstrual cycle. AFC, on the other hand, refers to the number of follicles measuring 2–5 mm or 2–10 mm in diameter observed in the ovaries at the start of the menstrual cycle and is highly correlated with the number of oocytes retrieved during controlled ovarian stimulation. Studies have shown that AMH levels are lower in women with EMS compared to healthy individuals [ 2 ]. Surgical treatment of EMS has been reported to negatively impact AMH levels both in the short and long term [ 3 ], likely due to surgical treatment damage to normal ovarian tissue and aggravate inflammation of the ovarian cortex, leading to diminished ovarian reserve. However, even in patients who do not undergo surgical intervention, AMH levels decline more rapidly compared to the healthy population [ 4 ]. Additionally, it has been observed that AFC is consistently lower in the affected ovary compared to the contralateral ovary, both preoperatively and postoperatively [ 5 ]. Early ovarian insufficiency not only results in reduced fertility among women of reproductive age but also leads to insufficient estrogen production, which can have long-term health consequences. Therefore, early determination of whether the ovarian reserve is reduced is crucial for addressing these issues.

This study focused on the ovarian reserve of patients with EMS prior to treatment and found that the disease itself negatively impacted ovarian reserve. For the first time, we identified that, in OEM, compared to AMH, AFC may serve as a more suitable marker for assessing ovarian reserve.

As demonstrated by the findings of this study, EMS predominantly affects women of reproductive age, underscoring the importance of considering ovarian reserve when managing this condition. There has long been debate regarding whether endometriotic cysts themselves reduce ovarian reserve or whether the ovarian reserve is diminished due to surgical intervention for endometriotic cysts. Our study found that, prior to treatment, patients with unilateral cysts had significantly higher right AFC and total AFC compared to those with bilateral cysts. Furthermore, for patients with unilateral cysts, the AFC of the affected ovary was significantly lower than that of the contralateral healthy ovary. These results indicate that endometriotic cysts themselves can reduce ovarian reserve even before treatment. To further validate this, we examined the relationship between ovarian cyst size and AFC. The results revealed a significant inverse correlation: the larger the maximal cyst diameter, the lower the AFC of the corresponding ovary. We also analyzed the relationships between AFC and other factors, including BMI, gravidity, recurrence, consistency of medication use, comorbidities, ultrasound sliding signs, and the largest diameter of deep nodules. AMH reflects global ovarian reserve, thus unaffected by unilateral cysts, whereas AFC is ovary-specific. Compared with the AFC, it has a lag. Factors such as recurrence, rectouterine pouch sliding sign, and the largest diameter of deep nodules were found to significantly affect AFC. This suggests that the severity of EMS is associated with a reduction in AFC. Preoperative AFC measurement not only assesses ovarian reserve but may also provide insights into the complexity of surgical intervention. Further multivariate regression analysis, after accounting for multicollinearity, reaffirmed that larger maximal cyst diameters were associated with lower AFC on the corresponding side. This finding suggests that as endometriotic cysts progress, the damage to the ovary becomes more pronounced. Previous studies have similarly reported that patients with EMS exhibit reduced AFC, indicating that the disease itself diminishes ovarian reserve [ 9 ]. Additionally, research has shown that follicle density around endometriotic cysts is significantly lower than in the normal contralateral ovarian cortex [ 10 ], which aligns with our findings. A 2015 study observed [ 10 ] that AFC was reduced in the affected ovary of patients with unilateral endometriotic cysts. However, it also reported no reduction in the number of oocytes retrieved from the affected ovary, suggesting that AFC in the affected ovary might be underestimated. Other studies [ 11 ] have proposed that although AFC reflects ovarian reserve more accurately in patients with unilateral EMSs, factors such as ovarian stretching and distortion of pelvic anatomy may increase the distance between the transvaginal probe and the ovary, reducing ultrasound resolution and leading to an underestimation of AFC. AFC may serve as a more suitable marker and appears more reflective of localized ovarian damage than AMH in endometriotic cysts. Currently, there is a lack of large-scale prospective studies to validate these findings, and further research is needed to confirm the observed associations. In this study, we also measured pre-treatment AMH levels in patients and found no significant differences between unilateral and bilateral cyst groups. Similarly, there was no significant correlation between AMH levels and cyst size. However, AMH was positively correlated with AFC. Further analysis revealed that AMH levels were negatively correlated with adenomyosis, rectouterine pouch sliding sign, and posterior uterine wall sliding sign. These findings indicate that endometriotic cysts themselves reduce ovarian reserve and that this reduction is associated with disease severity. Several studies [ 12 – 15 ] have shown that patients with DIE exhibit lower AMH levels. Similarly, another study reported [ 16 ] significantly lower AMH levels in patients with stage IV EMS. Moreover, AMH levels were also found to be influenced by factors such as pregnancy history, BMI, and uterine fibroids. These results suggest that AMH, compared to AFC, is more affected by extraneous factors, making AFC a more reliable indicator of ovarian reserve. As previously discussed, AFC measurement can also be influenced by the disease itself and technical factors during ultrasound evaluation. Therefore, a comprehensive assessment of ovarian reserve should be conducted prior to treatment. However, this study has certain limitations. The absence of a healthy control group prevented a comparative analysis of AMH levels between healthy individuals, patients with benign ovarian cysts, and those with endometriotic cysts. The mechanisms underlying the reduction in ovarian reserve remain unclear. Studies [ 14 ] have suggested that endometriotic ovaries exhibit significant fibrosis, which may contribute to the decline in ovarian reserve. One study [ 17 ] found that early follicular development in endometriotic ovaries may be prematurely activated, leading to increased follicular atresia and localized depletion of the primordial follicle pool, which constitutes ovarian reserve. The lower detection rate of left-sided cysts may be attributable to bowel gas interference during transvaginal ultrasound examinations. Despite the fact that the Enzian classification was not formally utilized in our study, our documentation of DIE-specific markers, including nodule size and sliding signs, captured the critical aspects of deep infiltrative endometriosis (DIE) burden. The retrospective design of the study resulted in incomplete data for certain parameters, potentially introducing a degree of selection bias. Furthermore, the absence of healthy or benign cyst controls hindered our ability to make direct comparisons. Additionally, the heterogeneity in assessments, with some patients undergoing surgery and others only receiving ultrasound evaluations, may have confounded the study results.

Below is the link to the electronic supplementary material. Supplementary file 1 (DOCX 18 KB) Supplementary file 1 (DOCX 18 KB)