Development of a nomogram for sperm retrieval at microTESE for idiopathic non-obstructive azoospermia in a multi-center cohort study.

backgroundNon-obstructive azoospermia (NOA) affects approximately 10% of infertile men and represents a major challenge in assisted reproductive technology (ART). A model that includes histological variants could be helpful in predicting sperm retrieval rate (SRR) after microdissection testicular sperm extraction (mTESE) in patients affected by NOA and without genetic abnormalities

objectivesTo develop and validate a predictive nomogram integrating clinical and histopathological variables to estimate SRR in NOA patients undergoing microTESE. MATERIAL AND

methodsA multi-center retrospective/prospective cohort study was conducted between 2022 and 2024, enrolling 333 men diagnosed with NOA across six academic centers. Preoperative data, including age, body mass index (BMI), hormonal profile (follicle-stimulating hormone [FSH], luteinizing hormone [LH], testosterone), genetic analysis, and ultrasound-measured testicular volume, were collected. MicroTESE was performed under microscopic guidance with histopathological evaluation. Predictive factors for successful sperm retrieval were analyzed through logistic regression, and a predictive nomogram was constructed and internally validated using bootstrapping techniques. Multi-variate logistic regression was performed to identify independent variables associated with SRR in patients with available final pathology. Therefore, a predictive nomogram was developed and validated using 1000 bootstrap samples.

resultsThe overall SRR was 52.55%. Multi-variate analysis identified FSH levels (odds ratio [OR]: 0.97; p = 0.049), maturation arrest (OR: 0.04; p < 0.01), and Sertoli cell-only syndrome (SCOS; OR: 0.03; p < 0.01) as independent predictors of SRR. The predictive nomogram demonstrated good accuracy, with a C-index of 0.75, sensitivity of 73%, specificity of 82%, and overall accuracy of 77% at a cutoff of 0.33. DISCUSSION: Using preoperative and histology data, we developed a nomogram to predict SR outcomes in patients with NOA undergoing mTESE. Our findings demonstrate that integrating hormonal and histopathological data enhances predictive accuracy of the model, thus providing a valuable tool for preoperative counseling and clinical decision-making in couples presenting with infertility.

conclusionThis validated nomogram effectively predicts sperm retrieval outcomes in NOA patients undergoing microTESE, facilitating improved patient counseling, informed clinical decisions, and optimized patient selection. Prospective external validation and further model refinement are recommended to enhance generalizability and clinical applicability.