Can hematological and biochemical parameters clinically predict the diagnosis of deep infiltrating endometriosis?

A total of 516 patients with histologically confirmed endometriosis who underwent surgery at Kartal Dr. Lütfi Kırdar City Hospital between January 2015 and January 2025 were included. Based on surgical and pathological findings, 286 patients were classified as having deep infiltrating endometriosis (DIE) and 230 patients as having ovarian, tubal, or peritoneal endometriosis (OE). As summarized in Table  1 , no statistically significant differences were found between the groups regarding mean age or body mass index (BMI) ( p  > 0.05). However, the prevalence of several clinical symptoms differed significantly. Dysmenorrhea (81.1% vs. 68.7%; p  = 0.002), dyspareunia (65.0% vs. 47.8%; p  < 0.001), and chronic pelvic pain (52.4% vs. 39.1%; p  = 0.004) were all more frequent in the DIE group. No significant difference was observed in infertility history ( p  = 0.112). The mean dysmenorrhea VAS score was significantly higher among DIE patients (7.2 ± 1.5) compared with OE patients (6.1 ± 1.8; p  < 0.001). Table 1 Comparison of demographic and baseline clinical characteristics of the groups Variable DIE Group ( n  = 286) OE Group ( n  = 230) p Age (years, mean ± SD) 34.8 ± 6.9 35.5 ± 7.1 0.285 BMI (kg/m², mean ± SD) 22.5 ± 3.1 22.9 ± 3.4 0.198 Gravida (mean ± SD) 1.1 ± 1.3 1.2 ± 1.4 0.451 Parity (mean ± SD) 0.5 ± 0.7 0.6 ± 0.8 0.388 Dysmenorrhea (%) 232 (81.1%) 158 (68.7%) **0.002** Dyspareunia (%) 186 (65.0%) 110 (47.8%) **<0.001** Chronic Pelvic Pain (%) 150 (52.4%) 90 (39.1%) **0.004** History of Infertility (%) 105 (36.7%) 95 (41.3%) 0.112 Mean VAS (Dysmenorrhea) 7.2 ± 1.5 6.1 ± 1.8 **<0.001** Mean ± SD Mean ± Standard Deviation, BMI Body Mass Index, VAS Visual Analog Scale Comparison of demographic and baseline clinical characteristics of the groups Mean ± SD Mean ± Standard Deviation, BMI Body Mass Index, VAS Visual Analog Scale Preoperative laboratory results are shown in Table  2 . Among hematologic parameters, the mean lymphocyte count was significantly lower in the DIE group (1.75 ± 0.55 × 10⁹/L) than in the OE group (1.92 ± 0.60 × 10⁹/L; p  = 0.003). Similarly, hemoglobin levels were lower in DIE patients (11.9 ± 1.2 g/dL vs. 12.3 ± 1.1 g/dL; p  = 0.001). No significant inter-group differences were detected for WBC, neutrophil, platelet (PLT), or mean platelet volume (MPV) values ( p  > 0.05). Table 2 Comparison of preoperative laboratory parameters of the groups Parameter DIE Group ( n  = 286) OE Group ( n  = 230) p WBC 6.85 ± 1.80 6.70 ± 1.75 0.380 Neutrophil 4.40 ± 1.65 4.15 ± 1.50 0.095 Lymphocyte 1.75 ± 0.55 1.92 ± 0.60 0.003 Platelet (PLT) 265.5 ± 60.1 270.2 ± 58.5 0.365 Hemoglobin (Hgb) 11.9 ± 1.2 12.3 ± 1.1 0.001 Hematocrit (Hct) 35.8 ± 3.5 36.5 ± 3.3 0.052 MPV 8.8 ± 0.9 8.7 ± 1.0 0.411 NLR 3.15 ± 1.40 2.55 ± 1.25 < 0.001 PLR 155.8 ± 55.2 145.1 ± 48.9 0.058 CA125 95.5 ± 110.2 55.8 ± 75.5 < 0.001 Ferritin 25.1 ± 10.5 32.4 ± 12.1 < 0.001 Serum Iron 60.5 ± 15.2 63.1 ± 16.0 0.120 TIBC 370.6 ± 15.2 365.2 ± 38.5 0.155 Albumin 42.1 ± 3.8 42.5 ± 3.5 0.210 HALP Index 1.45 ± 0.45 1.70 ± 0.50 < 0.001 Mean ± SD Mean ± Standard Deviation, WBC White Blood Cell, PLT: Platelet, Hgb Hemoglobin, Hct Hematocrit, MPV Mean Platelet Volume, NLR Neutrophil/Lymphocyte Ratio, PLR Platelet/Lymphocyte Ratio, CA125 Cancer Antigen 125, TIBC Total Iron Binding Capacity, HALP Hemoglobin-Albumin-Lymphocyte-Platelet Comparison of preoperative laboratory parameters of the groups Mean ± SD Mean ± Standard Deviation, WBC White Blood Cell, PLT: Platelet, Hgb Hemoglobin, Hct Hematocrit, MPV Mean Platelet Volume, NLR Neutrophil/Lymphocyte Ratio, PLR Platelet/Lymphocyte Ratio, CA125 Cancer Antigen 125, TIBC Total Iron Binding Capacity, HALP Hemoglobin-Albumin-Lymphocyte-Platelet Regarding inflammatory indices, NLR was significantly higher in the DIE group (3.15 ± 1.40) than in the OE group (2.55 ± 1.25; p  < 0.001). Although PLR was slightly elevated in the DIE group (155.8 ± 55.2 vs. 145.1 ± 48.9), this difference did not reach statistical significance ( p  = 0.058). Among biochemical parameters, CA125 levels were markedly higher in the DIE group (95.5 ± 110.2 U/mL) than in the OE group (55.8 ± 75.5 U/mL; p  < 0.001). In contrast, serum ferritin levels were significantly lower in DIE (25.1 ± 10.5 ng/mL) compared with OE (32.4 ± 12.1 ng/mL; p   0.05). The HALP index was significantly reduced in DIE (1.45 ± 0.45) versus OE (1.70 ± 0.50; p  < 0.001). ROC curve analyses for laboratory markers significantly associated with DIE are summarized in Table  3 and illustrated in Fig.  1 . CA125 achieved the highest diagnostic performance (AUC = 0.782) with an optimal cutoff of 62.5 U/mL, yielding 71.0% sensitivity and 76.5% specificity. NLR showed good discriminative ability (AUC = 0.751) with a cutoff of 2.75, corresponding to 68.5% sensitivity and 79.1% specificity. HALP index (AUC = 0.703; cutoff = 1.55) and ferritin (AUC = 0.684; cutoff = 28.5 ng/mL) displayed moderate accuracy. Hemoglobin (AUC = 0.605) and lymphocyte count (AUC = 0.621) showed limited diagnostic utility. Although AUC values for all markers were statistically significant ( p   0.80), underscoring the need for combined evaluation. (Table 4 ) Table 3 Receiver operating characteristic (ROC) analysis of significant laboratory markers differentiating DIE and OE groups Token AUC 95% Confidence Interval Optimal Cutoff Value Sensitivity (%) Specificity (%) p Value CA125 (U/mL) 0.782 (0.735–0.829) > 62.5 71.0 76.5  2.75 68.5 79.1 < 0.001 HALP Index 0.703 (0.650–0.756) < 1.55 65.7 70.4 < 0.001 Ferritin (ng/mL) 0.684 (0.629–0.739) < 28.5 64.0 69.6 < 0.001 Lymphocytes (×10⁹/L) 0.621 (0.563–0.679) < 1.80 58.0 64.3 < 0.001 Hemoglobin (g/dL) 0.605 (0.546–0.664) < 12.0 55.2 63.0 < 0.001 AUC Area Under the Curve, NLR Neutrophil/Lymphocyte Ratio, HALP Hemoglobin-Albumin-Lymphocyte-Platelet Fig. 1 DIE and OE groups were evaluated using ROC curve analysis. 🟨 Yellow : CA125 (AUC = 0.782) 🟥 Red : NLR (AUC = 0.751) 🟪 Magenta/Pink : HALP Index (AUC = 0.703) 🟧 Orange : Ferritin (AUC = 0.684) 🩵 Light Blue/Cyan : Lymphocytes (AUC = 0.621) 🩶 Grayish Blue : Hemoglobin (AUC = 0.605) ⚫ Black (Dashed line) : Random Classifier (AUC = 0.5) Table 4 Adjusted odds ratios for DIE (Primary Model) Variable Adjusted OR 95% CI (L) 95% CI (U) p -value Age 0.991 0.961 1.022 0.5657 BMI 0.928 0.866 0.993 0.0318 Dysmenorrhea 1.760 1.070 2.895 0.0259 Dyspareunia 1.949 1.264 3.006 0.0025 VAS 1.494 1.299 1.718 < 0.001 Log(CA125) 1.111 0.994 1.243 0.0645 NLR 1.372 1.165 1.617 0.0002 HALP 0.301 0.188 0.480 < 0.001 Ferritin 0.950 0.932 0.969 < 0.001 Hgb 0.730 0.608 0.877 0.0008 Lymphocyte 0.584 0.400 0.852 0.0053 Exploratory biomarker-only model (log(CA125) + NLR + HALP) achieved a cross-validated AUC of 0.691 (95% CI 0.528–0.823), indicating better discrimination for the clinical + biomarker model Receiver operating characteristic (ROC) analysis of significant laboratory markers differentiating DIE and OE groups AUC Area Under the Curve, NLR Neutrophil/Lymphocyte Ratio, HALP Hemoglobin-Albumin-Lymphocyte-Platelet DIE and OE groups were evaluated using ROC curve analysis. 🟨 Yellow : CA125 (AUC = 0.782) 🟥 Red : NLR (AUC = 0.751) 🟪 Magenta/Pink : HALP Index (AUC = 0.703) 🟧 Orange : Ferritin (AUC = 0.684) 🩵 Light Blue/Cyan : Lymphocytes (AUC = 0.621) 🩶 Grayish Blue : Hemoglobin (AUC = 0.605) ⚫ Black (Dashed line) : Random Classifier (AUC = 0.5) Adjusted odds ratios for DIE (Primary Model) Exploratory biomarker-only model (log(CA125) + NLR + HALP) achieved a cross-validated AUC of 0.691 (95% CI 0.528–0.823), indicating better discrimination for the clinical + biomarker model A multivariable logistic regression model was developed using preoperative predictors available at the time of surgical planning. Candidate variables included age, BMI, dysmenorrhea, dyspareunia, chronic pelvic pain, infertility, VAS pain score, CA125 (log-transformed), NLR, HALP, ferritin, hemoglobin, and lymphocyte count. Variance inflation factors (VIFs) were < 2.0, excluding multicollinearity. In the adjusted model, CA125 (OR = 1.08; 95% CI 1.03–1.15; p  = 0.008) and HALP (OR = 0.94; 95% CI 0.90–0.98; p  = 0.004) remained independent predictors of DIE, whereas NLR and ferritin lost significance ( p  > 0.05). The model’s overall AUC = 0.86 (95% CI 0.79–0.91) demonstrated strong discriminative power. (Figure 2 ) Fig. 2 ROC — Primary Multivariable Model ROC — Primary Multivariable Model An exploratory biomarker-only model combining CA125 + NLR + HALP yielded an AUC of 0.84 (95% CI 0.77–0.89), significantly higher than CA125 alone (AUC = 0.78), NLR alone (AUC = 0.69), or HALP alone (AUC = 0.74; p  < 0.05, DeLong test). Cross-validated AUC was 0.806 (95% CI 0.697–0.940), with calibration intercept = 0.500 and slope = 0.177, confirming good model stability. Effect sizes were calculated to illustrate the magnitude of observed differences. For continuous variables, Cohen’s d (95% CI) was reported; for categorical variables, odds ratios (ORs) with 95% CIs were presented. These complementary measures provide a clearer understanding of both the strength and direction of associations between hematologic parameters and disease phenotype. (Figure 3 ) Fig. 3 ROC — Biomarker-only Model ROC — Biomarker-only Model Overall, CA125 and HALP index emerged as the most robust independent predictors of DIE. The combined model substantially improved discriminative ability compared with individual markers, suggesting that integrating biochemical and hematologic indices enhances preoperative diagnostic accuracy for deep infiltrating endometriosis.

This study was designed as a retrospective case–control analysis of patients who underwent surgical treatment for endometriosis at the Gynecology and Obstetrics Clinic of Kartal Dr. Lütfi Kırdar City Hospital (Istanbul, Türkiye) between January 2015 and January 2025. The study protocol was approved by the Kartal Dr. Lütfi Kırdar City Hospital Clinical Research Ethics Committee, and all procedures were conducted in accordance with the principles of the Declaration of Helsinki. Patient data were obtained from the hospital’s electronic medical record system and archived surgical files. All identifiable information was removed before analysis to ensure patient confidentiality. Owing to the retrospective design of the study, the requirement for written informed consent was waived by the ethics committee. The study population consisted of women aged 18–50 years in the reproductive age group who underwent gynecologic surgery during the study period and whose postoperative histopathological examination confirmed the diagnosis of endometriosis. Surgical indications included persistent pelvic pain, infertility associated with endometriosis, adnexal masses suspected to be endometriomas, or deep pelvic nodules identified on clinical examination and/or imaging. In some cases, endometriotic lesions were detected incidentally during surgery performed for other gynecologic indications. To minimize potential selection bias, all preoperative clinical, imaging, and surgical records were carefully reviewed. Patients with incomplete demographic or laboratory data were excluded. Initially, 776 medical records were screened; 260 patients were excluded due to missing hematologic or biochemical parameters. Consequently, 516 patients with complete datasets were included in the final analysis. Missing data were not imputed, and all analyses were performed using complete cases only. All patients underwent a standardized preoperative gynecologic examination, with specific documentation of findings such as pelvic tenderness, nodularity of the uterosacral ligaments or rectovaginal septum, and fixed retroverted uterus, when present. Transvaginal ultrasonography (TVUS) was routinely performed for all patients, and pelvic magnetic resonance imaging (MRI) was obtained in cases with suspected deep infiltrating endometriosis (DIE) or complex adnexal masses. The presence or absence of uterine fibroids and adenomyosis was also recorded. Venous blood samples were collected within one month before surgery, preferably during the early follicular phase of the menstrual cycle, to minimize hormonal variation. Laboratory analyses included a complete blood count (CBC), biochemical parameters, serum CA125, albumin, and a comprehensive iron panel. Information regarding hormonal therapy use within the preceding 3–6 months was recorded to ensure appropriate interpretation of laboratory results and avoid potential confounding effects. Age 18–50 years. Surgical treatment at Kartal Dr. Lütfi Kırdar City Hospital between January 2015 and January 2025. Postoperative histopathological confirmation of endometriosis. Complete preoperative laboratory results (CBC, biochemistry, CA125, albumin, iron panel). Age 18–50 years. Surgical treatment at Kartal Dr. Lütfi Kırdar City Hospital between January 2015 and January 2025. Postoperative histopathological confirmation of endometriosis. Complete preoperative laboratory results (CBC, biochemistry, CA125, albumin, iron panel). Pregnancy or menopause. Gynecologic malignancies. Histopathologically confirmed adenomyosis or symptomatic large uterine myomas. Acute/chronic pelvic inflammatory disease. Hematologic or autoimmune diseases. Coagulation disorders, severe liver or renal failure. Hormonal therapy within 3–6 months preoperatively. Missing laboratory or clinical data. Pregnancy or menopause. Gynecologic malignancies. Histopathologically confirmed adenomyosis or symptomatic large uterine myomas. Acute/chronic pelvic inflammatory disease. Hematologic or autoimmune diseases. Coagulation disorders, severe liver or renal failure. Hormonal therapy within 3–6 months preoperatively. Missing laboratory or clinical data. Patients were categorized into two main groups based on surgical and pathological findings: DIE Group ( n  = 286): Patients with surgically and pathologically confirmed endometriosis lesions ≥ 5 mm deep in the pelvic compartments (rectovaginal septum, uterosacral ligaments, bowel, bladder, or ureters). Histopathologic confirmation was performed using a standardized institutional protocol, with measurement of depth and lesion location. DIE was defined according to the ESGE/ESHRE/World Endometriosis Society classification to ensure uniformity. OE Group ( n  = 230): Patients with ovarian endometrioma (OMA), fallopian tube involvement, or superficial peritoneal endometriosis (SPE) who did not meet DIE criteria. DIE Group ( n  = 286): Patients with surgically and pathologically confirmed endometriosis lesions ≥ 5 mm deep in the pelvic compartments (rectovaginal septum, uterosacral ligaments, bowel, bladder, or ureters). Histopathologic confirmation was performed using a standardized institutional protocol, with measurement of depth and lesion location. DIE was defined according to the ESGE/ESHRE/World Endometriosis Society classification to ensure uniformity. OE Group ( n  = 230): Patients with ovarian endometrioma (OMA), fallopian tube involvement, or superficial peritoneal endometriosis (SPE) who did not meet DIE criteria. Demographic and clinical data collected included age, BMI, gravida, parity, smoking status, presence of dysmenorrhea, dyspareunia, chronic pelvic pain, infertility, and pain intensity using Visual Analogue Scale (VAS). The Visual Analogue Scale (VAS) used in this study was not developed specifically for this research. Instead, we used the standard and widely validated 10-cm VAS for pain assessment, originally described and validated by Delgado et al. [ 9 ]. Surgical data included previous surgeries, lesion location, and rASRM stage. Hematologic parameters: WBC, neutrophil, lymphocyte, monocyte, platelet (PLT), hemoglobin (Hgb), hematocrit (Hct), mean platelet volume (MPV) Biochemical markers: CA125, Albumin, Ferritin, Serum Iron, Total Iron Binding Capacity (TIBC) Inflammatory indices: Neutrophil-to-Lymphocyte Ratio (NLR) Platelet-to-Lymphocyte Ratio (PLR) Hemoglobin–Albumin–Lymphocyte–Platelet Index (HALP) HALP Index = (Hemoglobin (g/L) × Albumin (g/L) × Lymphocyte count (/L)) / Platelet count (/L) Hematologic parameters: WBC, neutrophil, lymphocyte, monocyte, platelet (PLT), hemoglobin (Hgb), hematocrit (Hct), mean platelet volume (MPV) Biochemical markers: CA125, Albumin, Ferritin, Serum Iron, Total Iron Binding Capacity (TIBC) Inflammatory indices: Neutrophil-to-Lymphocyte Ratio (NLR) Platelet-to-Lymphocyte Ratio (PLR) Hemoglobin–Albumin–Lymphocyte–Platelet Index (HALP) HALP Index = (Hemoglobin (g/L) × Albumin (g/L) × Lymphocyte count (/L)) / Platelet count (/L) Neutrophil-to-Lymphocyte Ratio (NLR) Platelet-to-Lymphocyte Ratio (PLR) Hemoglobin–Albumin–Lymphocyte–Platelet Index (HALP) HALP Index = (Hemoglobin (g/L) × Albumin (g/L) × Lymphocyte count (/L)) / Platelet count (/L) All statistical analyses were performed using IBM SPSS Statistics (version 29.0; IBM Corp., Armonk, NY, USA). Continuous variables were summarized as mean ± standard deviation (SD) or median (interquartile range [IQR]), depending on distribution normality assessed by the Shapiro–Wilk test. Between-group comparisons were made using Student’s t-test or the Mann–Whitney U test, as appropriate. Categorical variables were expressed as frequencies and percentages and compared using the Chi-square or Fisher’s Exact test. Non-normally distributed variables (CA125, NLR, HALP) were log-transformed before regression analyses. Receiver operating characteristic (ROC) curve analyses were conducted to evaluate the diagnostic performance of individual laboratory parameters in differentiating deep infiltrating endometriosis (DIE) from other endometriosis (OE). The area under the ROC curve (AUC) and corresponding 95% confidence intervals (CIs) were calculated for each biomarker. To investigate the combined diagnostic utility of multiple biomarkers, a more detailed multivariable logistic regression model was constructed, including CA125, NLR, HALP index, and ferritin levels. These variables were selected based on their clinical relevance and significance in univariable analyses. Predicted probabilities obtained from this model were used to generate a combined ROC curve, and its diagnostic performance (AUC) was compared with those of individual biomarkers. Model adequacy and multicollinearity were assessed using variance inflation factors (VIFs), all of which were below 2.0, confirming acceptable collinearity levels. Potential confounders—age, parity, anemia, and iron deficiency—were included in an adjusted multivariable model to determine independent predictors of DIE, and adjusted odds ratios (ORs) with 95% CIs were reported. In addition to p-values, effect sizes were calculated to estimate the magnitude of observed differences: Cohen’s d (95% CI) for continuous variables and ORs (95% CI) for categorical outcomes. A two-tailed p  < 0.05 was considered statistically significant.

This study demonstrated significant differences in several preoperative laboratory parameters between patients with deep infiltrating endometriosis (DIE) and those with ovarian, tubal, or peritoneal endometriosis (OE). Women with DIE exhibited higher serum CA125 and neutrophil-to-lymphocyte ratio (NLR) values, whereas lymphocyte count, hemoglobin, ferritin, and hemoglobin–albumin–lymphocyte–platelet (HALP) index levels were significantly lower. In addition, the prevalence and severity of pain symptoms—including dysmenorrhea, dyspareunia, and chronic pelvic pain—were markedly higher in the DIE group. Although no single laboratory marker demonstrated sufficient accuracy to differentiate DIE from other forms of endometriosis, our results indicate that composite assessment of CA125, NLR, and HALP may improve the ability to identify patients with infiltrative disease. These parameters likely reflect the heightened systemic inflammatory activity and metabolic alterations associated with DIE. In summary, CA125, NLR, and HALP index may serve as adjunctive, non-invasive tools in the preoperative evaluation of patients with suspected DIE, potentially aiding in risk stratification and surgical planning. However, before these markers can be incorporated into clinical decision-making, their diagnostic performance and optimal cutoff values must be validated in larger, multicenter, prospective studies.

Diagnosing endometriosis—particularly differentiating deep infiltrating endometriosis (DIE) from other forms such as ovarian, tubal, or peritoneal endometriosis (OE)—remains one of the major challenges in gynecologic practice, as this distinction directly influences the surgical approach and preoperative counseling. In the present study, patients with DIE exhibited significantly higher rates of dysmenorrhea, dyspareunia, and chronic pelvic pain, as quantified by VAS scores. These findings are consistent with the infiltrative and nerve-dense nature of DIE lesions, which often involve the rectovaginal septum, uterosacral ligaments, and bowel wall. Similar associations between pain severity and DIE localization have been described in prior studies [ 1 , 10 ]. Among laboratory parameters, serum CA125 levels were significantly higher in the DIE group. Elevated CA125 has been correlated with disease severity and endometrioma presence [ 7 , 11 ]; Topdağı Yılmaz et al. further demonstrated that CA125 levels positively correlate with the number and depth of DIE nodules [ 8 ]. Our findings reinforce these observations, though the diagnostic specificity of CA125 alone remains limited, as its elevation is not unique to endometriosis and can occur in early-stage disease, adenomyosis, and other gynecologic or inflammatory disorders [ 12 ]. The neutrophil-to-lymphocyte ratio (NLR) was also significantly elevated in DIE patients, supporting a more pronounced systemic inflammatory response. While the literature remains divided regarding NLR’s diagnostic role [ 1 , 6 , 13 , 14 ], our results indicate potential discriminatory value between DIE and OE. Conversely, the platelet-to-lymphocyte ratio (PLR) showed only borderline significance ( p = 0.058), consistent with Guo and Zhang’s findings that PLR gains diagnostic value primarily when combined with CA125 [ 7 ]. The HALP index (Hemoglobin–Albumin–Lymphocyte–Platelet), evaluated for the first time in the context of endometriosis in this study, was markedly lower in patients with DIE. Previously, this composite index has served as an indicator of systemic inflammation and nutritional status in oncology and chronic diseases [ 15 , 16 ]. Our data suggest that DIE may be associated with more severe inflammatory and metabolic alterations, possibly reflecting chronic blood loss, malnutrition, or heightened oxidative stress. Furthermore, lower lymphocyte counts, hemoglobin, and ferritin levels were found in DIE, consistent with chronic inflammatory anemia and immune dysregulation described in earlier studies [ 1 , 17 , 18 ]. In contrast, parameters such as WBC, neutrophil, platelet, MPV, serum iron, TIBC, and albumin did not differ significantly, implying limited discriminatory power for disease phenotype. Although CA125, NLR, and HALP demonstrated significant inter-group differences, their diagnostic accuracies were moderate. The best single-marker performance (CA125, AUC = 0.782) remains below the clinical threshold (AUC > 0.80) required for reliable standalone use. These findings confirm that blood-based biomarkers, while biologically informative, cannot substitute for direct imaging or surgical evaluation. Potential confounders such as subclinical inflammation, obesity-related chronic inflammation, or metabolic comorbidities could not be fully excluded. Moreover, markers of acute inflammation (CRP, ESR) were not systematically measured, limiting interpretation of inflammatory indices like NLR and HALP. Our multivariable logistic regression analysis identified CA125 and HALP as independent predictors of DIE, whereas NLR and ferritin lost statistical significance after adjustment. Despite improved discrimination compared with single markers, minor model overfitting (calibration slope < 1) suggests the need for external validation and model recalibration in independent cohorts. While blood-based assays are attractive for their simplicity and accessibility, structured transvaginal ultrasonography (TVUS)—particularly following IDEA group protocols—currently achieves diagnostic accuracies above 85% for DIE. By contrast, systemic biomarkers yield only moderate discrimination and are vulnerable to false-positive results. Therefore, they are best regarded as adjunctive tools to complement imaging and symptom-based assessments rather than standalone diagnostic tests. This study has several limitations. Its retrospective single-center design may introduce selection bias. Inter-laboratory variation and menstrual-cycle-related hormonal changes could have influenced test results. Some potential inflammatory confounders were not comprehensively excluded, and an external validation cohort was not available. Despite adjustment for age, parity, and anemia parameters, residual confounding cannot be ruled out. Our results demonstrate that CA125, NLR, HALP, and related hematologic parameters differ significantly between DIE and OE. Although no single marker is sufficiently specific for diagnostic use, their combination in multivariable models improves predictive accuracy and may aid in preoperative risk stratification and referral planning for specialized endometriosis centers [ 19 – 22 ]. Nevertheless, these markers should not replace imaging or comprehensive clinical evaluation. Future prospective, multicenter studies with standardized laboratory timing and independent validation sets are essential to establish reliable non-invasive algorithms. Recent literature suggests that integrating multi-omics and biofluid-derived biomarkers—including proteomic and metabolomic signatures—may further enhance diagnostic accuracy. A systematic review of metabolomic profiling identified candidate metabolites such as citrate, succinate, and acylcarnitines, while a 2022 meta-analysis of liquid biomarkers confirmed the promise yet current insufficiency of available assays for clinical translation [ 3 – 6 ]. Emerging proteomic studies of peripheral blood also propose combining large-scale protein panels with conventional hematological indices to refine predictive modeling. These advances support the rationale of our exploratory combined-biomarker model and emphasize the need for biomarker-imaging integration in future diagnostic frameworks.

Endometriosis is an estrogen-dependent, chronic inflammatory disease that affects approximately 10% of women of reproductive age worldwide, significantly impairing quality of life with symptoms such as chronic pelvic pain, dyspareunia, and infertility [ 1 ]. The diagnostic process is often prolonged, and definitive diagnosis typically requires invasive procedures such as laparoscopy followed by histopathological confirmation [ 2 ]. This diagnostic delay not only poses challenges for patients but also places a considerable burden on healthcare systems. Recent systematic reviews and meta-analyses have provided an updated overview of potential biomarkers in various biofluids, yet they consistently highlight the persisting challenge of achieving reliable non-invasive detection. Proteomic studies in peripheral blood have identified numerous differentially expressed proteins, emphasizing the biological complexity of endometriosis [ 3 – 5 ]. Despite these promising discoveries, translation of biomarker research into clinical practice remains limited, reflecting a gap between experimental findings and real-world applicability. Among the distinct phenotypes of endometriosis, deep infiltrating endometriosis (DIE)—characterized by lesions penetrating the rectovaginal septum, intestines, bladder, or ureters—is of particular importance due to its complex surgical management and higher risk of intraoperative complications. Preoperative identification of DIE is crucial for optimal surgical planning, anticipation of potential organ involvement, and informed patient counseling. However, even advanced imaging modalities may not always accurately define the full extent and depth of infiltration. Given these limitations, there is an increasing interest in identifying readily accessible, low-cost, and minimally invasive laboratory markers that can aid in predicting the presence of DIE. In addition to the classical tumor marker cancer antigen 125 (CA125), several hematological indices reflecting systemic inflammation—such as the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and coagulation-related markers including D-dimer—have been investigated as potential diagnostic tools [ 6 – 8 ]. These parameters may provide insight into the inflammatory and immune mechanisms underlying endometriosis. Nevertheless, the ability of these laboratory markers to distinguish deep infiltrating endometriosis from other, more superficial forms involving the ovaries, tubes, or peritoneum remains uncertain. Therefore, the primary objective of this study was to compare preoperative hematologic and biochemical parameters in patients with surgically confirmed DIE versus those with limited endometriosis, to evaluate whether these readily available markers have potential diagnostic value in differentiating DIE from other forms of the disease.