Blood biomarkers for the non-invasive diagnosis of endometriosis.
Level 1 - systematic review of randomized trials
Systematic review and meta-analysis of diagnostic accuracy studies
PubMed 27132058 · doi:10.1002/14651858.CD012179
What was done
This Cochrane systematic review evaluated blood biomarkers as replacement or triage tests for surgical diagnosis of pelvic endometriosis (peritoneal, ovarian, or deep infiltrating). Multiple databases were searched to mid-2015 for cross-sectional or randomized studies of reproductive-aged women undergoing surgical visualization. Predetermined diagnostic performance criteria were set for replacement tests (sensitivity 0.94, specificity 0.79), rule-out triage tests (sensitivity ≥ 0.95, specificity ≥ 0.50), and rule-in triage tests (sensitivity ≥ 0.50, specificity ≥ 0.95). Bivariate models pooled sensitivities and specificities when sufficient datasets existed.
What was found
The review included 141 studies (15,141 participants) evaluating 122 blood biomarkers. Meta-analyses were feasible for only four markers: - Anti-endometrial antibodies (4 studies, 759 women): sensitivity 0.81 (95% CI 0.76 to 0.87), specificity 0.75 (95% CI 0.46 to 1.00). - IL-6 (> 1.90 to 2.00 pg/ml; 3 studies, 309 women): sensitivity 0.63 (95% CI 0.52 to 0.75), specificity 0.69 (95% CI 0.57 to 0.82). - CA-19.9 (> 37.0 IU/ml; 3 studies, 330 women): sensitivity 0.36 (95% CI 0.26 to 0.45), specificity 0.87 (95% CI 0.75 to 0.99). - CA-125: sensitivities ranged across thresholds from 0.40 (95% CI 0.32 to 0.49 at > 35.0-36.0 U/ml) to 0.73 (95% CI 0.67 to 0.79 at > 25.0-26.0 U/ml); specificities ranged from 0.64 (95% CI 0.47 to 0.82 at > 10.0-14.7 U/ml) to 0.91 (95% CI 0.88 to 0.94 at > 35.0-36.0 U/ml). No biomarker met the predefined replacement or triage performance criteria. Eighty-two studies evaluated 97 biomarkers that did not differentiate endometriosis from controls, including 22 with conflicting results across studies.
Why it matters
No blood biomarker currently demonstrates adequate diagnostic accuracy to replace surgical laparoscopy or serve as a reliable triage test in clinical practice.
Limits
All included studies were judged to have poor methodological quality. Substantial heterogeneity, variable threshold cut-offs, and small sample sizes precluded meta-analysis for 118 biomarkers. Included populations were restricted to high-prevalence surgical cohorts (10% to 85% endometriosis frequency), limiting generalizability to primary care or unselected populations. Data were insufficient to statistically pool biomarkers differentiating endometriomas from other benign ovarian masses.
Cited by
- supports The diagnostic gold standard for endometriosis is surgical confirmation because no laboratory test exists.
- supports There is currently no blood test available to diagnose endometriosis.