Search for a command to run...
Abstract Introduction Microsurgical testicular sperm extraction (mTESE) is the gold-standard method to surgically retrieve sperm in men with non-obstructive azoospermia (NOA). Although sperm retrieval rates range from 30-70% in this population, patient factors reliably associated with successful gamete procurement remain elusive. Objective We evaluated the predictive value of anthropomorphic variables hormonal parameters on sperm retrieval in men with idiopathic NOA naive to prior medical or surgical treatment. Methods We performed a retrospective cohort study on men with idiopathic NOA who underwent mTESE at a large academic institution. Anthropomorphic, laboratory, surgical, and pathology data were captured. Men with prior genitourinary surgery, cryptozoospermia, prior use of testosterone or empirical medical therapy, missing testosterone or gonadotropin levels, or confirmed genetic disease were excluded. Univariate and multivariable logistic regression models were used to assess the predictive utility of anthropomorphic variables and hormonal parameters with regard to successful sperm retrieval (+SR) and failure (-SR). The performance of a clinically-informed partition model was assessed. Association of testicular pathology with sperm retrieval outcome was also evaluated. Results We identified 40 men with idiopathic NOA who underwent mTESE between August 2008 and April 2025 and met our strict inclusion criteria. Half of these men (n=21/40, 53%) had +SR. Median age (35.4 vs 32.8 years, p=0.15) and BMI (28.9 vs 27.0 kg/m^2, p=0.25) were similar between those with +SR and -SR. Follicle stimulating hormone (FSH, 5.9 vs 19.5 IU/l, p=0.0006) and luteinizing hormone (LH, 6.6 vs 10.2 IU/l, p=0.006) were lower among men with +SR, while testosterone (T, 408 vs 433 ng/dl, p=0.76) was similar. Most men exhibited Sertoli Cell Only (SCO) testicular pathology (61%, n=20/33), with approximately one-third exhibiting active spermatogenesis or hypospermatogenesis (36%, n=12/33). Given these endocrinological differences, hormone ratios were interrogated. LH/FSH (0.92 vs. 0.47, p=0.004), T/LH (69 vs. 48, p=0.03), and T/FSH (61 vs. 23, p=0.0003) were found to be higher in men with +SR. T/FSH (OR: 1.04, 95% CI: 1.02-1.08, p=0.0003, AUC: 0.84) was the strongest predictor of +SR on univariate logistic regression, followed by FSH alone (OR: 0.88, 95% CI: 0.80-0.95, p=0.0004, AUC: 0.82). The univariate logistic regression model including T/FSH predicted +SR better than multivariable models including T and FSH (AUC: 0.84 vs 0.83) or T/FSH and LH (AUC: 0.84 vs 0.83). However, a curated partition model (Figure), including indicators FSH <10.8 IU/l and T ≥600 ng/l, was 91% sensitive and 84% specific for +SR and performed better than the best logistic regression model (AUC: 0.88 vs 0.84). Pathological assessment generally correlated with sperm retrieval outcomes. Men with +SR were more likely to exhibit evidence of spermatogenesis on pathological assessment (63 vs 12%, p=0.004), while those with -SR were exhibited a higher prevalence of SCO pathology (88 vs 32%, p=0.001). Conclusions Men with idiopathic NOA with +SR are endocrinologically distinct from those with -SR, exhibiting lower pre-operative levels of LH and FSH. The T/FSH ratio was the single best predictor of +SR and likely reflects a higher degree of efficient testicular function. Men with baseline FSH <10.8 IU/l or T ≥600 ng/dl exhibit a high probability of +SR. Disclosure No
Published in: The Journal of Sexual Medicine
Volume 22, Issue Supplement_4