Incident endometriosis diagnosis and anti-müllerian hormone (AMH): how surgical staging and typology relate to serum AMH levels

Endometriosis is a chronic, gynecologic condition in which tissue similar to the lining of the uterus implants throughout the body. Women with endometriosis have a higher prevalence of infertility and a greater risk of early natural menopause compared to those without endometriosis.

This study aimed to evaluate preoperative serum AMH levels among women with and without incident endometriosis and to assess whether levels differ by surgical staging and typology.

The ENDO (Endometriosis: Natural History, Diagnosis, and Outcomes) study was conducted between 2007 and 2009. The ENDO study consisted of an operative and population cohort (n=600). Only those in the ENDO operative cohort from the Utah site were used for this analysis, and included women aged 18 to 44 years who were scheduled for gynecologic surgery, irrespective of clinical indication (n=476). AMH levels were measured from stored serum collected before surgery using a quantitative enzyme-linked immunosorbent assay. After excluding participants with missing outcome data (n=51), unilateral oophorectomy (n=8), or those within the population cohort (n=69), 348 participants remained for the analysis. Surgically confirmed endometriosis diagnosis, staging (American Society for Reproductive Medicine I-IV), and typology (superficial, deep, ovarian) were ascertained by the operative report. Outliers for AMH (>14.0 ng/mL) were excluded from the analyses and AMH values were log-transformed. Multivariable linear regression models adjusted for age (squared and continuous), body mass index, serum cotinine levels, and exogenous hormonal contraceptive use were conducted. Percentage differences in AMH were calculated as (exp[β]-1)×100, and 95% confidence intervals were reported.

Compared with no endometriosis, incident endometriosis diagnosis was associated with lower AMH levels (-19.8%; 95% confidence interval, -37.0 to 1.0); however, this association was not statistically significant. Stage III to IV disease was associated with 40.1% lower AMH levels (95% confidence interval, -58.9 to -12.7). Ovarian endometriomas were most strongly associated with lower AMH levels (-54.3%; 95% confidence interval, -69.4 to -31.8), with a more pronounced association among those with infertility (-72.6%; 95% confidence interval, -85.4 to -48.5). Deep (-24.1%; 95% confidence interval, -48.2 to 11.0) and superficial (-15.5%; 95% confidence interval, -34.6 to 9.3) endometriosis also showed a trend toward lower AMH levels, but these findings were not statistically significant. Compared with a postoperative diagnosis of a normal pelvis, incident endometriosis was associated with 26.8% lower AMH levels (95% confidence interval, -44.6 to -3.4). Stage III to IV disease was associated with 47.8% lower AMH levels (95% confidence interval, -65.8 to -23.2), and all subtypes of endometriosis were statistically significantly associated with lower levels of AMH compared with a postoperative diagnosis of a normal pelvis (ovarian: -60.8%; 95% confidence interval, -74.4 to -39.9; deep: -34.3%; 95% confidence interval, -56.2 to -1.4; superficial: -24.8%; 95% confidence interval, -43.9 to -0.8).

Ovarian and moderate to severe (stage III-IV) endometriosis were associated with markedly lower AMH levels compared with no endometriosis. Compared with a postoperative diagnosis of a normal pelvis, incident endometriosis and moderate to severe stages (stage III-IV) were associated with statistically significantly lower AMH levels. Additionally, typology (deep, ovarian, or superficial) was associated with statistically significantly lower AMH levels. However, this association was likely driven by the presence of ovarian endometriomas across all subtypes. These findings are consistent with previous studies and demonstrate that endometriosis lesions themselves, independent of surgical intervention, influence AMH levels.