Background: microTESE proved to be the gold standard surgical approach for patients with non-obstructive azoospermia (NOA), but sperm retrieval rates (SRRs) vary considerably among centers. Some authors compared their SRRs with the pattern of seminiferous tubule caliber found at high magnification, but none provided diagnostic accuracy measures. Objective: The present retrospective study sought to verify the diagnostic accuracy of the pattern of seminiferous tubule caliber in predicting the sperm retrieval in NOA patients. Materials and Methods: Data from 143 infertile NOA men undergoing unilateral (64) or bilateral (79) microTESE (222 testes) were retrospectively evaluated. During microTESE, if present, dilated tubules (DTs) were retrieved, otherwise tubules with slightly larger caliber (SDT) (924) than that of the surroundings were removed. When no DT or SDT were found, not dilated tubules (NDTs) were excised. Results: Spermatozoa were retrieved in 95 of 222 testes (42.8%); sperm retrieval was successful in 90% of testes with DTs, in 47% of those with SDTs, and only in 7% of those with NDTs (p < 0.0001). Stepwise binary logistic regression revealed that the combination of seminiferous tubule pattern and testis histology was significantly predictive of SSR, being able to classify 86.8% of testes, with an excellent diagnostic accuracy (AUC 0.93). The median number of spermatozoa retrieved was significantly higher in DTs compared with SDTs and NDTs. Discussion: The results of the present study provide reliable accuracy measures in support of the relationship between seminiferous tubule caliber pattern and SSR in patients with non-obstructive azoospermia. We are proposing for the first time that spermatozoa may be retrieved even from slightly dilated tubules in about half of cases. The pattern of tubules retrieved, together with histology, may represent an additional outcome measure of microTESE. Conclusion: The pattern of seminiferous tubules together with testis histology predicts sperm retrieval with an excellent diagnostic accuracy.
SUPPORTING INFORMATIONAdditional supporting information may be found online in the Supporting Information section at the end of the article.Table S1 Clinical parameters as stratified per tubules caliber pattern.