Diagnostic accuracy of cancer antigen 125 for endometriosis: a systematic review and meta‐analysis

Background The development of a non‐invasive and accurate diagnostic biomarker for endometriosis is urgently needed. Objective Evaluate the diagnostic accuracy of serum cancer antigen 125 (CA 125) for endometriosis. Search strategy We searched EMBASE, MEDLINE, and Web of Science from inception to January 2016. Selection criteria Diagnostic accuracy studies of serum CA 125 (index test) for histologically confirmed endometriosis (reference standard) were included. Data collection and analysis Two authors independently selected trials, extracted study characteristics and data. Methodological quality was assessed using Quality Assessment of Comparative Diagnostic Accuracy Studies (QUADAS‐2) checklist. Main results Twenty‐two studies (16 cohort, six case‐control), 3626 participants, were identified. Bivariate hierarchical models were used to pool accuracy data of 14 studies (2920 participants) using CA 125 ≥ 30 units/ml. Pooled specificity was 93% (95% CI 89–95%) and sensitivity 52% (95% CI 38–66%). CA 125 was significantly more sensitive for the diagnosis of moderate or severe endometriosis compared with minimal disease (63%, 95% CI 47–77% versus 24%, 95%CI 19–32%, P‐value = 0.001). Conclusions CA 125 performs well as a rule‐in test facilitating expedited diagnosis and ensuring investigation and treatment can be confidently tailored for the management of endometriosis. Unfortunately, a negative test, CA 125 < 30 units/ml, is unable to rule out endometriosis.