Intrasurgical parameters associated with successful sperm retrieval in patients with non-obstructive azoospermia undergoing salvage microdissection testicular sperm extraction

Background Patients with non-obstructive azoospermia with a previously failed conventional testicular sperm extraction may undergo a salvage microdissection testicular sperm extraction with the probability of successful sperm retrieval being almost dependent upon the number of previous surgical attempts and to different histopathologic categories. Objectives To determine whether the seminiferous tubules pattern and the histological categories could affect the sperm retrieval rate in patients with non-obstructive azoospermia undergoing salvage microdissection testicular sperm extraction after failed conventional testicular sperm extraction. Materials and methods Seventy-nine patients undergoing unilateral or bilateral salvage microdissection testicular sperm extraction were evaluated. During microdissection testicular sperm extraction, if present, dilated tubules were retrieved, otherwise, tubules with slightly larger caliber than that of the surroundings were removed. When no dilated tubule or tubule with slightly larger caliber was found, not dilated tubules were excised. A prediction model was built with seminiferous tubules pattern and testis histology as covariates. Results Sperm retrieval was successful in 30 out of 79 patients. The prediction model correctly classified 88.3% of cases, explained the 29.7% variability of the outcome, and significantly predicted the microdissection testicular sperm extraction outcome with a sensitivity of 67.7% and a specificity of 90.2%, Both tubules with slightly larger caliber and not dilated tubules were negatively associated with the chance of retrieving spermatozoa. Among the histological categories, only early maturation arrest was significant to the model (log(SSR) = 0.57 - 1.9SDT - 3.3NDT - 1.76EMA) (where SSR is sperm retrieval rate, SDT is tubule with slightly larger caliber, NDT is not dilated tubule, and EMA is early maturation arrest). The model had a clearly useful discrimination (area under the curve = 0.814), the estimated performance was 0.8105, and internal calibration was acceptable (p > 0.05). Discussion Seminiferous tubules pattern and testis histology may reliably explain the salvage microdissection testicular sperm extraction outcome in all patients with non-obstructive azoospermia apart from those with early maturation arrest, where the homogeneous apparent seminiferous tubules pattern may be misleading. Conclusion The outcome of salvage microdissection testicular sperm extraction can be predicted by the same intrasurgical parameters that have been demonstrated to predict the outcome of microdissection testicular sperm extraction in naive patients with non-obstructive azoospermia.