Diagnostic yield of exome sequencing in nonobstructive azoospermia (NOA): A systematic review and meta-analysis

Azoospermia is considered as the most severe form of male infertility. The application value of exome sequencing (ES) in males with non-obstructive azoospermia (NOA) remains unclear. This study aims to review the known genetic causes of NOA and evaluate the diagnostic yield of ES in males diagnosed with idiopathic NOA. We performed a systematic database search in Ovid MEDLINE, EMBASE, CINAHL, Scopus, Cochrane Central Register of Controlled Trials, and Web of Science from database inception to March 2025. Two independent reviewers assessed the literature and included those studies investigating the utility of ES testing in men diagnosed with NOA and fulfilling the eligibility criteria. The pooled diagnostic yield was calculated using single-proportion analysis with random–effects modeling, and confidence intervals (CI) were estimated using the Clopper–Pearson exact method. A total of nine studies were included, and the qualities were assessed to be moderate to high via the modified STARD. Among the cohorts analyzed (nine studies comprising 1,728 individuals with NOA), the overall diagnostic yield of ES testing was 15% (95% CI: 10%–20%; low-certainty evidence). Of the 270 positive cases identified through ES testing, mutations in 262 genes were detected, with AR, TEX11, FANCM, TDRD9, PNLDC1, M1AP, FBXO15, and DMRT1 being the most frequently observed. Among these cases, only 11.11% (5/45) reported successful testicular sperm extraction. The considerable heterogeneity indicates that the pooled prevalence estimates of the diagnostic yield of ES testing in NOA—approximately 15%—may overestimate the true diagnostic rate in the general NOA population. This estimate should thus be interpreted as an average across diverse clinical and methodological contexts, rather than a precise point estimate reflecting a uniform underlying effect. Future research, particularly large-scale studies using standardized protocols, is crucial to generate more accurate, reliable, and generalizable estimates of the diagnostic yield of ES testing in NOA.