Introduction
Endometriosis is a chronic estrogen-dependent inflammatory disorder characterized by the presence of endometrial tissue outside the uterine cavity. It affects approximately 10% of women of reproductive age worldwide and is associated with severe symptoms, such as chronic pelvic pain, painful menstruation (dysmenorrhea), infertility, and irregular menstrual cycles. Beyond causing physical damage, endometriosis significantly reduces the quality of life and imposes a considerable psychological and economic burden on both patients and healthcare systems [1-3]. Despite its high prevalence, the diagnostic delay for endometriosis remains unacceptably long, averaging between 7 years and 10 years from symptom onset [4,5]. This delay is largely due to the reliance on laparoscopy as the gold standard for diagnosis; however, this procedure is invasive, costly, and not always accessible. There is a pressing need for non-invasive biomarkers that can enable earlier detection and guide clinical management.
Recent studies have focused on identifying circulating biomarkers involved in inflammation, angiogenesis, and ovarian function. Interleukin-6 (IL-6) is a central proinflammatory cytokine that contributes to ectopic tissue proliferation, immune dysregulation, and pain sensitization in endometriosis [6]. A 2024 systematic review and meta-analysis by Dolińska et al. [7] underscored the diagnostic promise of multi-cytokine panels including IL-6, IL-8, tumor necrosis factor alpha (TNF-α), and vascular endothelial growth gactor, hence the importance of further research in this area.
In addition to inflammation, there is growing recognition of the link between chronic immune activation and impaired ovarian reserve. Anti-Müllerian hormone (AMH), a widely accepted marker of ovarian function, is expressed at lower levels in women with endometriosis. This hormone reduction is directly linked to cytokine-mediated effects, follicular injury, oxidative stress, and surgical intervention complications [8]. Cancer antigen 125 (CA125), a glycoprotein known primarily as a tumor marker, especially for ovarian cancers, has also been associated with endometriosis in moderate-to-severe cases. These markers may be useful, multifaceted diagnostic tools [9,10].
To date, most biomarker studies have focused on European, Asian, or North American populations, with limited representation of North African women despite the potential influence of genetic, immunological, environmental, and healthcare access-related factors on disease presentation [11]. In this context, the present prospective study was conducted at a gynecology center in Morocco to evaluate the diagnostic value of IL-6, CA125, AMH, and other hormonal parameters in a cohort of North African women.
Materials and methods
1. Study population
This prospective case-control study was conducted between August 2024 and May 2025 at a fertility center in Casablanca, Morocco. In total, 224 women aged 19-52 years participated in this study, comprising 112 patients with pathologically confirmed endometriosis and 112 healthy controls. A 1:1 ratio was chosen to ensure balanced group comparisons, minimize sampling bias, and optimize the statistical power. Participants were recruited from gynecology clinics and referrals for laparoscopic evaluation. This study was approved by the Institutional Ethics Committee (approval no. [CE/UM6SS/25/24]), and written informed consent was obtained from all participants.
The inclusion criteria were women diagnosed with endometriosis via laparoscopy with no prior ovarian surgery or endocrine disorders. It also included women of Moroccan origin from a single ethnic background. Moroccan women with suspected or confirmed malignancy, pregnancy, postmenopausal status, or refusal to participate were excluded.
Endometriosis severity was classified according to the revised American Society for Reproductive Medicine classification system. Based on the laparoscopic findings, patients were categorized into four stages: stage I (minimal): 1-5 points; stage II (mild): 6-15 points; stage III (moderate): 16-40 points; and stage IV (severe): >40 points. This classification was used to assess disease progression.
2. Sample collection and processing
Peripheral venous blood samples were collected under sterile conditions immediately before laparoscopy, between 07:30 AM and 09:30 AM, to minimize circadian variation in cytokine concentrations during the follicular phase (from day 3 to day 10 of the follicular phase of the menstrual cycle) and reduce hormonal variability, particularly for sensitive markers such as IL-6 and CA-125. Plasma was separated within 30 minutes of collection by centrifugation at 1,800 rpm for 15 minutes and stored at −20°C in ethylenediaminetetraacetic acid-coated tubes for subsequent analysis. Serum samples were obtained from dry tubes using the same centrifugation procedure.
3. Biomarker measurement
Quantitative detection of IL-6 levels was performed using sandwich enzyme-linked immunosorbent assay kits, Elecsys® IL-6 (Roche Diagnostics, Penzberg, Germany), with a sensitivity of 1.5 pg/mL and specificity of 84%.
CA125 levels were measured using an Elecsys® electrochemiluminescence immunoassay system, with a diagnostic threshold of 35 IU/mL. Hormonal markers, including AMH, estradiol (E2), follicle-stimulating hormone (FSH), luteinizing hormone (LH), and thyroid hormones (thyroid-stimulating hormone [TSH] thyroxine [T4]), were analyzed using a Roche Cobas 6000 analyzer, module e601 (Roche Diagnostics). All assays were conducted in triplicate to ensure precision, with intra- and inter-assay coefficients of variation maintained at <5% and <10%, respectively.
4. Statistical analysiss
Data were analyzed using Microsoft Excel (Microsoft Corporation, Redmond, WA, USA), DATAtab (DATAtab, Graz, Austria), and GraphPad Prism version 10 (GraphPad Software, San Diego, CA, USA). The distribution of all variables was assessed using the Shapiro-Wilk normality test. Because the biomarkers showed non-normal distributions, the results were reported as median values with interquartile ranges (IQR). Comparisons between two groups were performed using the Mann-Whitney U-test, while comparisons across more than two groups were conducted using the Kruskal-Wallis test, followed by Dunn’s post-hoc multiple comparison test when applicable. Correlations between biomarkers and clinical or hormonal parameters were assessed using Spearman’s rank correlation coefficient (ρ). Statistical significance was set at P<0.05.
Results
1. Clinical characteristics of participants
In total, 112 Moroccan women with endometriosis and 112 Moroccan control participants were included in this study. The mean age was 35.48±5.64 years in the endometriosis group and 34.08±7.82 years in the control group; P=0.0499, indicating a statistically significant difference between the two groups. Body mass index was significantly higher in women with endometriosis (25.74±4.71 vs. 23.29±4.75; P<0.001). The mean duration of infertility was 4.92±3.04 years, while all women in the control group had preserved fertility. Advanced stages of endometriosis (stage III-IV) were predominant (n=83), with a substantial proportion of patients presenting as deep infiltrating endometriosis (DIE) (Table 1).
2. Biomarkers analysis
As summarized in Table 2, plasma concentrations of IL-6 (40.42±25.07 pg/mL vs. 2.82±1.23 pg/mL; 37.3 [20.68-54.38] vs. 2.7 [1.9-3.5]; P<0.001) and serum concentration of CA125 (44.93±33.41 U/mL vs. 16.54±7.03 U/mL; 36 [18.0-65.6] vs. 14.75 [11.58-19.85]; P<0.001) were significantly higher in women with endometriosis compared to control patients, reflecting a persistent inflammatory response and increased lesion activity characteristic of the disease. From an endocrine perspective, patients showed a marked reduction in AMH levels (1.28±0.97 ng/mL vs. 2.44±1.58 ng/mL; 1.2 [0.53-2.5] vs. 2.3 [1.08-3.0]; P<0.001), suggesting impaired ovarian reserve. E2 concentrations were also significantly lower in the endometriotic group (70.99±47.99 pg/mL vs. 118.18±46.45 pg/mL; 57.5 [39.0-85.25] vs. 127.8 [83.98-150.38]; P<0.001), indicating ovarian hypofunction that may result from an inflammatory microenvironment or follicular damage. LH levels were also significantly reduced in women with endometriosis compared to controls (5.57±3.53 IU/mL vs. 9.18±3.56 IU/mL; 4.95 [3.38-7.38] vs. 8.55 [6.5-10.9]; P<0.001), indicating a broader dysregulation of the hypothalamic-pituitary-ovarian (HPO) axis. In contrast, FSH levels were significantly elevated in patients compared to controls (8.36±4.66 IU/mL vs. 5.34±1.52 IU/mL; 8 [5.4-10.0] vs. 5.65 [4.3-6.7]; P<0.001), reflecting compensatory stimulation in response to declining ovarian function. Regarding thyroid function, TSH showed no significant difference between groups (2.60±1.95 vs. 2.56±1.25; 2 [1.02-3.3] vs. 2.6 [1.5-3.58]; P=0.2443). Although T4 was slightly reduced in patients (11.74±5.82 vs. 12.94±2.07; 12 [9.0-14.88] vs. 12.9 [11.23-14.48]; P=0.0125), the difference was of limited clinical magnitude. Overall, these findings underscore the combined inflammatory and endocrine disruptions associated with endometriosis (Table 2).
1) Biomarker profiles across American Society for Reproductive Medicine stages of endometriosis
Moroccan women with endometriosis demonstrated markedly elevated plasma IL-6 concentrations during the disease stages. Median IL-6 levels increased from 15.6 pg/mL in stage (I-II) to 38.2 pg/mL in stage (III-IV), compared with 2.7 pg/mL in controls (both P<0.0001 vs. controls). Although IL-6 concentrations were higher in advanced disease, the difference between stage (I-II) and stage (III-IV) remained statistically significant (P=0.0159) (Fig. 1A). Similarly, serum CA125 levels were significantly increased from 27.7 U/mL in stage (I-II) to 46.0 U/mL in stage (III-IV); P=0.0259, while controls exhibited markedly lower levels (14.75 U/mL; P=0.0154 and P<0.0001, respectively) (Fig. 1B). In contrast, AMH levels, reflecting ovarian reserve, were reduced in stage (III-IV) (0.9 ng/mL) relative to stage (I-II) (1.9 ng/mL; P=0.0053) and control (2.3 ng/mL; P<0.0001). However, the difference between stage (I-II) and controls was not statistically significant (P=0.6109) (Fig. 1C). These findings show that IL-6 and CA125 levels increase in parallel with endometriosis severity, whereas the reduction in AMH levels appears to reflect an early impairment of ovarian function that is dependent on the disease stage.
2) Biomarker levels according to endometriosis types
Plasma IL-6 concentrations were highest in patients with DIE associated with ovarian endometrioma (OMA) lesions (median, 44 pg/mL), followed by patients with isolated DIE (43.07 pg/mL), OMA (34.07 pg/mL), and superficial peritoneal lesion (SUP; 15.08 pg/mL) (Fig. 2A). The Kruskal-Wallis test followed by Dunn’s post-hoc test showed that IL-6 was significantly higher in DIE than in SUP (P<0.0001), and in DIE&OMA than in SUP (P=0.0001). Although the IL-6 concentrations in the OMA did not differ significantly from those in the SUP (P=0.0695), the DIE phenotype showed the highest IL-6 levels, reflecting the increased inflammatory burden characteristic of deep lesions.
Similarly, serum CA125 concentrations increased proportionally with lesion complexity: DIE&OMA (51.15 U/mL), isolated DIE (40.1 U/mL), OMA (21.9 U/mL), and SUP (25.1 U/mL). CA125 was significantly elevated in DIE compared to SUP (P=0.0022), whereas OMA did not differ significantly from SUP (P>0.99). The DIE&OMA group showed the highest concentrations, clearly exceeding those in SUP (P=0.0086), highlighting the cumulative effect of deep and ovarian lesions (Fig. 2B).
In contrast, serum AMH levels were markedly reduced in the most severe phenotypes, particularly in women with combined DIE&OMA (0.78 ng/mL), followed by OMA (0.69 ng/mL), isolated DIE (0.90 ng/mL), and notably SUP (2.11 ng/mL). AMH levels were significantly lower in DIE than in SUP (P<0.0001), with an even more pronounced reduction in OMA (P=0.0002). The DIE&OMA group exhibited the most significant decline, which differed significantly from that of the SUP (P<0.0001) (Fig. 2C). Overall, these results highlight an increasingly pronounced inflammatory burden and endocrine impairment in the presence of deep, ovarian, or combined lesions, underlining the unfavorable biological profile associated with the most complex forms of endometriosis.
3) Correlation of IL-6 with hormonal and clinical biomarkers
Spearman correlation (ρ) revealed that IL-6 levels showed selective correlations with biomarkers of inflammation and ovarian reserve, but not with gonadotropins or E2. A moderate positive association was found between IL-6 and CA125 (ρ=0.3784; P<0.0001), suggesting that systemic inflammation may mirror peritoneal immune activation in endometriosis (Fig. 3A). Conversely, IL-6 exhibited a negative correlation with AMH (ρ=−0.3389; P=0.0003), supporting the hypothesis that inflammatory processes may contribute to a decline in ovarian reserve (Fig. 3B). No significant associations were observed between IL-6 and gonadotropins such as FSH (ρ= −0.07991; P=0.4023) or LH (ρ=0.2794; P=0.0131), nor with circulating E2 (ρ=0.06391; P=0.5032) (Fig. 3C-E). Overall, these results highlight the specificity of IL-6 as an inflammatory marker, distinct from classical endocrine regulators, and underline its potential role at the interface between inflammation, ovarian dysfunction, and clinical expression of endometriosis.
4) Correlation of IL-6, CA125 and AMH with age
The relationship between age and IL-6, CA125, and AMH levels was investigated in both the control and endometriosis groups. The IL-6 levels remained persistently elevated in patients across all age groups, with no significant correlation with age. Regression analysis confirmed the absence of association between IL-6 and age in the endometriosis group (ρ=0.02624), while controls exhibited a weak, non-significant positive correlation (ρ=0.02214) (Fig. 4A). Similarly, CA125 levels were also consistently higher in patients compared to controls, with minimal age-related correlation in either group (ρ=0.1843 in endometriosis, ρ=−0.03115 in controls) (Fig. 4B). These findings suggest that IL-6 and CA125 are stable disease-specific biomarkers that are largely unaffected by chronological aging, reinforcing their diagnostic utility. In contrast, AMH levels exhibited a strong inverse correlation with age in controls (ρ=−0.9312), consistent with the expected physiological decline in ovarian reserve. However, in endometriosis patients, this relationship was substantially weaker (ρ=−0.3424), indicating considerable variability in AMH levels independent of age (Fig. 4C). This suggests that additional pathological factors, such as lesion type and severity of ovarian involvement, may influence ovarian reserve in this population.
5) Evaluation of the biomarker sensitivity and specificity by receiver operating characteristic (ROC) analysis
ROC analysis demonstrated excellent diagnostic performance for IL-6, with an area under the curve (AUC) of 0.952 and sensitivity and specificity approaching 90% at the optimal cutoff value of 7.0 pg/mL, reflecting its strong association with inflammatory processes in Moroccan women with endometriosis (Fig. 5A). CA125 also showed excellent diagnostic accuracy (AUC=0.79) (Fig. 5B). Hormonal markers, including E2 and LH, demonstrated good discriminatory performance with AUCs of 0.78 and 0.798, respectively, supporting the involvement of ovarian hypofunction and HPO axis disruption in the disease (Fig. 5E, F). AMH and FSH showed fair diagnostic values, with AUCs of 0.721 and 0.752, respectively (Fig. 5C, D). These values reflect a reduction in the ovarian reserve and compensatory endocrine mechanisms, which have already been highlighted in previous analyses.
Discussion
Diagnosing endometriosis remains a major challenge in gynecology despite the use of laparoscopy as the most validated diagnostic approach. Although accurate, this technique is invasive, costly, and limited in detecting certain forms such as retroperitoneal lesions or DIE [12]. These limitations have driven researchers toward non-invasive approaches, notably the identification of reliable blood biomarkers, which could improve early diagnosis, patient stratification, and follow-up. Proinflammatory cytokines, lesion activity markers, hormones of ovarian function, and other plasma and serum factors are involved in immunity and angiogenesis [13,14].
This study is the first conducted on a North African (Moroccan) cohort to simultaneously evaluate a panel combining IL-6, CA125, AMH, and an extended set of reproductive and thyroid hormones (including FSH, LH, E2, T4, and TSH). The objective of this study was to determine their diagnostic relevance and association with the disease stage and phenotype. Among these biomarkers, IL-6 showed the highest diagnostic accuracy, indicating its potential as a clinical tool for endometriosis diagnosis. As a pleiotropic cytokine, IL-6 is produced in response to tissue injury by immune and endometrial cells and plays a central role in inflammation, angiogenesis, and modulation of the endometrial microenvironment [15]. Several studies have reported elevated IL-6 levels in plasma, serum, peritoneal fluid, and follicular fluid of women with endometriosis [16-18]. This increase reflects the chronic inflammatory state of the disease.
In our cohort of 112 patients, IL-6 concentrations were markedly higher than those in the controls (37.3 vs. 2.7 pg/mL; P<0.001), reflecting a pronounced systemic inflammatory state. Levels increased in parallel with disease severity, reaching 38.2 pg/mL in stage (III-IV) and peaking in patients with combined DIE&OMA (44 pg/mL), followed by isolated DIE (43.07 pg/mL), OMA (34.07 pg/mL), and SUP (15.08 pg/mL).
These findings are consistent with those of several studies reporting elevated IL-6 concentrations in women with endometriosis compared with those in healthy controls. Desdicioglu et al. [19] observed significantly higher serum IL-6 levels in patients with endometriosis (1.31±12.10 pg/mL) compared to controls (4.31±1.82 pg/mL; P<0.05). Similarly, Kashanian et al. [10] also demonstrated significantly elevated IL-6 levels in women with endometriosis (30.4±6.43 pg/mL) relative to controls (13.9±3.17 pg/mL), which closely approximates the values observed in our cohort. Conversely, several studies did not detect significant differences in IL-6 levels between patients with endometriosis and healthy controls. For example, Somigliana et al. [20] reported comparable IL-6 levels in affected and control participants (0.6 pg/mL [undetectable-1.4] vs. 1.0 pg/mL [0.4-1.9]; P=0.09), possibly due to their relatively small sample size, which included only 45 endometriosis cases.
Regarding disease stage, our results showing a progressive increase in IL-6 concentrations with advancing endometriosis severity are in line with Li et al. [15], who found that serum IL-6 levels increased from 16.96±5.95 pg/mL in stage I-II to 22.77±11.02 pg/mL in stage III-IV, compared to 17.15±7.29 pg/mL in controls [17]. However, Martínez et al. [21] found that while serum IL-6 was elevated in women with minimal-mild endometriosis (29.4±9.0 pg/mL), levels were paradoxically lower in advanced stages (17.6±10.3 pg/mL) compared with controls (15.7±9.3 pg/mL; P=0.002). The progressive increase in IL-6 with disease severity observed in our study supports its potential use as a marker of disease progression [21].
IL-6 levels were also independent of age and did not correlate with gonadotropin or E2 levels, suggesting its specificity as an inflammatory marker rather than an endocrine surrogate. Its robust AUC (0.952), high sensitivity (91.1%), and high specificity (80.4%) further emphasize its diagnostic potential in clinical practice. These results align with those of Jiang et al. [22], who reported an AUC of 0.905 for IL-6, with 90.0% sensitivity and 93.7% specificity. However, Somigliana et al. [20] suggested that the ROC analysis of serum IL-6 levels was inadequate, and they observed low performance in detecting endometriosis.
CA125, a historical biomarker of lesion activity, was also significantly elevated in patients (44.93 vs. 16.54 U/mL). Its progressive increase according to disease stage and lesion type, particularly in the DIE and DIE&OMA phenotypes, is consistent with its role in assessing lesion burden. These findings align with those of Jing et al. [23] and Kokot et al. [9], confirming the relevance of this marker in advanced forms, although its performance remains limited in early stages. We also observed a positive correlation between IL-6 and CA125 (ρ=0.3784; P<0.0001), consistent with Velasco et al. [24], who described a potential synergy between inflammation (IL-6) and lesion activity (CA125) in severe cases. Recent data have confirmed that panels combining IL-6, CA125, and angiogenic or neuroinflammatory markers significantly improve diagnostic accuracy [25,26].
Regarding AMH, our results reveal a significant reduction in its concentration among women with endometriosis (1.28±0.97 ng/mL vs. 2.44±1.58 ng/mL; P<0.001), with a more pronounced decline in deep-infiltrating and cystic forms. These findings corroborate those of Garavaglia et al. [11] and Kasapoglu et al. [27], who demonstrated diminished ovarian reserves in severely diseased forms. The negative correlation between IL-6 and AMH (ρ=−0.3389; P=0.0003) suggests a possible role of inflammation in follicular impairment, a hypothesis further supported by a recent study showing that IL-6 modulates follicular viability through the JAK/STAT3 and NF κB pathways [28].
ROC analyses demonstrated superior diagnostic performance of IL-6 (AUC=0.952), whereas CA125 (0.79), AMH (0.721), LH (0.798), and E2 (0.78) showed intermediate performance. These results, which outperform those reported by Mihalyi et al. [29] (AUC=0.75 for IL-6) and are consistent with the performance observed by Rokhgireh et al. [30] (AUC=0.90), reinforce the value of a multi-marker approach, particularly relevant in resource-limited settings. In parallel with Burghaus et al. [14], integrating IL 6 into artificial intelligence algorithms could enhance diagnostic performance.
The hormonal pattern identified in our cohort-significantly elevated FSH (8.36±4.66 IU/mL), markedly reduced LH (5.57±3.53 IU/mL), and lower E2 levels (70.99±47.99 pg/mL) measured in the early-mid follicular phase (from day 3 to day 10 of the menstrual cycle)-is clinically and biologically plausible and aligns with current evidence on inflammation-driven ovarian dysfunction in endometriosis. Recent multi-omics and mechanistic studies demonstrate that chronic exposure to IL-6, TNF-α, and oxidative stress can disrupt granulosa cell steroidogenesis, accelerate primordial follicle depletion, induce cortical fibrosis, and alter autophagy/ubiquitination pathways in ovarian endometriosis, as demonstrated by Nahdi et al. [31] and Ou et al. [32]. These alterations reduce E2 synthesis from developing follicles, weaken negative feedback to the hypothalamic-pituitary axis, and explain the compensatory rise in FSH observed in our patients. Impaired E2-mediated gonadotropin-releasing hormone pulsatility may attenuate LH secretion, consistent with the reduced LH levels found in our cohort. Evidence from animal models and ovarian tissue studies, such as those by Ge et al. [33] and Kanellopoulos et al. [34], further confirms that endometriosis can impair folliculogenesis, reduce AMH, and disrupt gonadotropin responsiveness even without prior surgery, with the most pronounced effects occurring in endometrioma and DIE phenotypes.
Importantly, not all previous studies have reported these hormonal changes. Somigliana et al. [20] and Mihalyi et al. [29] found no significant differences in gonadotropin or E2 levels between women with endometriosis and controls, highlighting the heterogeneity that is likely related to differences in disease stage, phenotype distribution, or inadequate cycle-phase control. Large contemporary cohorts, including Ramezani Tehrani et al. [35] and Chung et al. [36], further emphasize that ovarian reserve decline is not universal, with preserved AMH in younger women and in those without sizeable or bilateral endometriomas, and underscore the confounding impact of prior cystectomy on ovarian reserve assessments. Therefore, the endocrine signature detected in our population likely reflects a high-risk inflammatory phenotype with cortical ovarian involvement, rather than a uniform hormonal pattern across all patients with endometriosis. These findings suggest that our study is a valuable contributor to clarifying the hormonal dynamics of endometriosis and emphasizes the clinical relevance of identifying distinct high-risk endocrine phenotypes. By integrating robust mechanistic and clinical evidence, our results provide a strong foundation for future phenotype-specific and longitudinal research aimed at validating this hormonal signature as an early indicator of ovarian vulnerability and as a useful tool for individualized fertility management.
Collectively, these findings suggest that endometriosis is characterized by a distinct biological signature encompassing systemic inflammation (IL-6), lesion activity (CA125), reduced ovarian reserve (AMH), and altered gonadotropin dynamics (FSH, LH, and E2). This integrated biomarker profile may be a powerful tool for non-invasive diagnosis, disease stratification, and monitoring.
A distinctive feature of our cohort was that the levels of several biomarkers, particularly IL-6 and CA125, were higher than those reported in other populations. This difference may reflect genetic predisposition, lifestyle factors, environmental exposures such as pollution or endocrine disruptors, and ethnicity-specific immune-inflammatory responses, all of which could amplify systemic inflammation in this population.
Altogether, these results underscore the diagnostic potential of a multi-marker panel comprising IL-6, CA125, AMH, and reproductive hormones, particularly in moderate-to-severe disease forms. The marked elevation of IL-6 in advanced stages, its high discriminatory power (AUC=0.952), and its inverse association with the ovarian reserve highlight its clinical relevance. CA125 similarly contributed to phenotypic differentiation despite its known limitations in early-stage disease. Although highly encouraging, these results warrant validation in larger, ethnically diverse cohorts to ensure robustness and generalizability.
Integrating these biomarkers into predictive models, potentially supported by artificial intelligence, may transform endometriosis management by enabling earlier detection, improved risk stratification, and personalized therapeutic strategies.
