ABSTRACT: Previous studies, including our own, have reported that spermatozoa isolated from the testis have remarkably higher occurrence of aneuploidy once isolated from azoospermic men. This notion, however, did not translate into a lower pregnancy rate nor a greater proportion of miscarriages. Indeed, ICSI offspring generated from surgically retrieved gametes did not suffer from increased karyotypic aneuploidy than children generated from ejaculated specimens. In recent years, aneuploidy assessments on a larger number of cells and utilizing more chromosome probes have reported a progressive decrease in chromosomal aberrations in spermatozoa directly retrieved from the seminiferous tubules. In light of the availability of more accurate molecular genetic techniques, we have decided to challenge the notion that sampling epididymal and testicular tissues yields spermatozoa with higher incidence of aneuploidy than those retrieved in the ejaculate. In a retrospective manner, we have carried out an analysis by FISH with 9 chromosome probes on at least 1000 cells from the ejaculates of 87 consenting men and the specimens of 6 azoospermic men, while spermatozoa of fertile donors were used as control. Aneuploidy by FISH yielded 0.9% for the donor control but rose in the study group to 3.6% in the ejaculated, 1.2% for the epididymal, and 1.1% for testicular spermatozoa. There were no differences in autosomal or gonosomal disomies, nor nullisomies. In this group, once the specimens of these men were used for ICSI, ejaculated spermatozoa yielded a 22% clinical pregnancy rate that resulted in 62.5% pregnancy loss. The surgically retrieved specimens yielded a 50% clinical pregnancy rate that progressed to term. To confirm our findings, in a prospective analysis, DNA sequencing was carried out on the ejaculates and surgical samples of 22 men with various spermatogenic characteristics. In this comparison, the findings were similar with actually a higher incidence of aneuploidy in the ejaculated spermatozoa (n = 16) compared to those surgically retrieved (n = 6) (P<0.0001). For this group, the clinical pregnancy rate for the ejaculated specimens was 47.2% with 29.4% pregnancy loss, while the surgically retrieved yielded a 50% clinical pregnancy rate, all progressing to term. A subsequent prospective combined assessment on ejaculated and surgically retrieved spermatozoa by FISH and NGS was performed on non-azoospermic men with high DNA fragmentation in their ejaculate. The assessment by FISH evidenced 2.8% chromosomal defects in the ejaculated and 1.2% in testicular biopsies while by NGS became 8.4% and 1.3% (P = 0.02), respectively. Interestingly, we evidenced a pregnancy rate of 0% with ejaculated while 100% with the testicular spermatozoa in this latter group. This indicates that improved techniques for assessing sperm aneuploidy on a wider number of cells disproves earlier reports and corroborates the safe utilization of testicular spermatozoa with a positive impact on chances of pregnancy.
