In patients with oligospermia in the ejaculate or previous ICSI failures if it concurs with high DNA fragmentation, it has been hypothesized that the use of sperm obtained from the testicle would improve the clinical results, since a source of damage to the spermatic DNA is post-testicular in its storage in the epididymis and thus could be avoided. The clinical information available so far is low, of low quality and all the studies present certain limitations susceptible to improvements in further investigations before giving a definitive answer to patients in these circumstances, about whether they should opt for testicular biopsy or for the use of semen in the ejaculate.The intention proposed in our project, is to demonstrate whether using testicular sperm, compared to those available in an ejaculate in these cases, offers a clinically and statistically significant increase in chromosomally normal embryos available that may lead to better reproductive performance of the cycles, in a design never before done, where half of a patient's oocytes are inseminated from ejaculated sperm and the other half from sperm obtained in the testicular biopsy.

Since its first description in 1992, intracytoplasmic sperm injection (ICSI) has been widely used to overcome all forms of severe male factor infertility. Despite achieving acceptable success rates with the use of abnormal sperm, studies suggest that poor sperm quality can be detrimental to ICSI outcomes. Sperm DNA plays a critical role in normal embryonic development because the genetic information passed on to the next generation depends on its integrity (6-8). Deterioration of sperm DNA content has been associated with several conditions, including lifestyle and exposure to gonadotoxins, varicocele, infections in the male accessory glands and advanced paternal age.

Sperm DNA fragmentation (SDF) tests measure the proportion of sperm with damaged chromatin, using probes or stains to identify DNA breaks with the aid of fluorescence microscopy, light microscopy or flow cytometry depending on the type of method. Regardless of the analytical method, DNA fragmentation is more common in the sperm of infertile men than fertile men. Among couples undergoing ICSI, high SDF in the ejaculate is found in 30% of men and is generally associated with abnormal conventional semen parameters. But this high SDF is also a relatively common finding in infertile men with normal semen parameters, up to 20% - 40% according to some authors. Aside from the questionable need to routinely use SDF test during the evaluation of male infertility, recent evidence indicates that knowing the FAE can be clinically informative in estimating the results to be obtained with assisted reproductive techniques (ART). Although sperm with fragmented DNA can fertilize an oocyte with apparently similar efficacy to sperm without DNA fragmentation, the negative impact of damaged paternal chromatin is generally manifested during the development of the embryos produced, leading to developmental blockage, implantation failure or early embryo loss, thus decreasing the success of ART.

Among the many strategies proposed to overcome SDF in couples undergoing ART, the use of testicular sperm instead of ejaculate in males without azoospermia has gained increased attention, due to recent results published by various groups. The biological plausibility of using testicular sperm for ICSI in men with high SDF in the ejaculate is based on observations of less DNA fragmentation in sperm from the testis compared with sperm from the ejaculate. The alteration of sperm chromatin integrity in ejaculated sperm from infertile men may be explained, first, because chromatin compaction is still ongoing during epididymal transit. Secondly, because excessive ROS can be generated in epididymal epithelial cells under stress factors such as high temperatures and environmental conditions. Finally, certain endonucleases can cleave the DNA of mature live sperm and, as a result, sperm DNA damage can occur by different pathways, including hydroxyl radical, nitric oxide, and activation of sperm caspases and endonucleases, thus explaining the positivity for SDF in ejaculated sperm or sperm from infertile men.

Thus, the concept that oxidation-induced sperm chromatin damage can occur in the post-testicular environment or in epididymal transit and storage, and that sperm recovery with improved chromatin integrity can be achieved if the epididymis is bypassed, has led researchers to explore the use of testicular sperm for assisted reproductive treatments. Because of the relevance of the clinical decision to resort to the use of testicular sperm for ICSI, especially because of the inherent risks and clinical implications of such an intervention, the potential benefit of testicular sperm for ICSI needs to be clarified in these cases and, thus, to determine the potential improvement that it could allow.

Our most recent work led us to examine the available evidence regarding ICSI outcomes with testicular sperm (Testi-ICSI) and ejaculated sperm (Ejac-ICSI) among non-azoospermic infertile men with confirmed post-testicular damage through a systematic review and meta-analysis of the available information, where we compared: 1) ICSI outcomes of Testi-ICSI and Ejac-ICSI among infertile men with high SDF (as defined in each study) in the ejaculate; and 2) SDF rates among testicular sperm and ejaculates of infertile men regardless of whether or not ICSI data were provided. Taking into account: 1) study population (history of ICSI failure versus no prior history of ICSI failure); 2) semen analysis profile of participants (oligozoospermia vs. normozoospermia); 3) method of SDF testing; and 4) method of sperm retrieval, and finding conflicting evidence regarding the use of Testi-ICSI in nonazoospermic men and the inherent risks of complications after sperm retrieval, it is necessary to clarify whether any particular male infertility population may derive benefit from this intervention.

Five studies compared SDF levels between ejaculated and testicular sperm for 143 patients who served as their own controls. The mean SDF rates in testicular and ejaculated sperm were 8.9 ± 5.1% and 33.4 ± 12.8% (P <.0001), respectively. Overall, the results indicated that SDF rates were lower in testicular sperm than in ejaculate with the mean difference being 24.58% (95% CI 32.53 to 16.64); I2 = 92%; P <.001. To assess heterogeneity, we performed a subgroup analysis according to the method used to assess SDF. Four studies used TUNEL assay and one study, the SCD test. With TUNEL, pooled results indicated lower rates of SDF in testis sperm than in ejaculates, with a mean difference of 19.78% (95% CI 22.35% to 17.21%; I2 = 15%; P <.001). In the only study using the SCD method, the mean difference was 32.4% (95% CI 34.85% to 29.95%; P <.001) in favor of testicular sperm.Regarding clinical pregnancy rates (CPR), four studies provided CPRs for 379 Testi-ICSI and Ejac-ICSI embryo transfers in patients with high AEF. The results indicated that CPRs were higher with Testi-ICSI than with Ejac-ICSI, being 50.0% and 29.4% (P <.001), respectively, with an OR of 2.42 (95% CI, 1.57 to 3.73; I2 = 34%; P <.001). Subgroup analysis by study population and semen profile revealed that the observed pooled effect size was larger for failed ICSI in studies with normozoospermic men, with greater improvement in Testi-ICSI observed in couples with no history of ICSI failure and oligozoospermic men, although differences between subgroups were not statistically significant. The consistency in the direction of effect across all studies increased confidence in the results of a positive outcome effect of Testi-ICSI in CPR.When considering miscarriage rates in four studies between 151 Testi-ICSI and Ejac-ICSI clinical pregnancies in patients with high AEF. The results indicated that for miscarriage the rates were reduced with Testi-ICSI compared with Ejac-ICSI, being 9.4% and 29.1% (P<.002), respectively, with an OR of 0.28 (95% CI 0.11 to 0.68; P = 0.005; I2 = 11%). The observed pooled effect size was larger for studies involving couples with a history of ICSI failure and normozoospermia than without a history of ICSI failure and oligozoospermia, although the difference between subgroups was not statistically significant. Regarding live newborn rates (LBR), two studies reported LBRs for 272 embryo transfers of Testi-ICSI and Ejac-ICSI in patients with high SDF. The results indicated that LBRs were higher with Testi-ICSI than with Ejac-ICSI, being 46.9% and 25.6% (P <.001), respectively, with an OR of 2.58 (95% CI 1.54-4.35; I2 = 0, P <.001). These results were consistent in suggesting a positive effect of Testi-ICSI on TRNV among men with high SDF.

Most of the studies included in our meta-analysis used TUNEL assay, which has been considered the most accurate method for predicting pregnancy in ART, with an area under the receiver operating characteristic curve of 0.71 (95% CI 0.66-0.76). Four of the studies applied the cutoff value of 30% sperm with fragmented DNA in pure semen to discriminate between normal and high SDF samples (34, 37, 40, 41), and the remaining studies used thresholds of 29%, 15%, and 7%. Notably, three of the four studies providing Testi-ICSI and Ejac-ICSI data agree on the cutoff value used to classify patients with high SDF. Therefore, despite the controversy as to which sperm DNA test and threshold values should be adopted clinically, our results suggest that Testi-ICSI is beneficial when SDF values exceed 29%. Many conditions associated with SDF may be correctable, including varicocele, lifestyle factors, and genital infections, and potentially enable or improve the possibility of natural conception or allow the use of ejaculated sperm for ICSI. In other cases, testicular biopsy may be an option to consider.Finally, despite having low overall DNA damage, testicular sperm may have higher rates of aneuploidy. In a recent paper, SDF rates were almost three times lower in testicular sperm than in ejaculated counterparts (14.9± 5.0% vs. 40.6± 14.8%; P<.05), but aneuploidy rates for chromosomes 13, 18, 21, X, and Y were higher in the former (12.4%± 3.7% vs. 5.7%± 1.2%; P<.05). Nevertheless, the data do not show an increased risk of malformations in offspring generated from ICSI cycles using testicular sperm, but continued monitoring is required until the safety of this strategy is confirmed.

Given that the quality and number of available studies on Test-ICSI in the context of high SDF is still limited, further studies are required, particularly randomized controlled trials that evaluate different aspects still unresolved in previous work. For example, none of the studies genetically analyzed the available embryos, nor their quality, let alone the cumulative rates obtained with frozen embryos, and subsequently in subsequent thawed cycles. The calculation of the results, estimating embryo transfer as the denominator, is introducing a negative bias, that is, unintentionally hiding a greater effect of the intervention, by selecting among the available embryos those with better characteristics, and taking into account only the contribution of these (thus making the differences between the groups to be compared smaller). Any intervention carried out in relation to spermatozoa must take this limitation into account, since it affects the entire embryo cohort, not just the embryos transferred, and thus the magnitude of the effect of the intervention is not adequately assessed. This would improve the adequacy of the effect size estimate and allow a better judgment of the clinical significance of our findings.

In addition, the study of semen samples after freezing and thawing/capacitation allows control of iatrogenic damage caused by laboratory manipulation, which had not been previously considered, and the control of the female contributing factor to reproductive success that is missing in previous studies would also be considered in our work, by statistically controlling for potentially influential variables. For the time being, our meta-analysis indicates that FAE rates are lower in testis sperm than in ejaculated sperm and clinical outcomes are higher for men with confirmed post-testicular SDF when Testi-ICSI rather than Ejac-ICSI is used. Therefore, the existing literature supports the recommendation to use Testi-ICSI in select groups of men with confirmed post-testicular SDF in ejaculate. Given the generally low to moderate quality of the studies reviewed and the potential risks associated with sperm retrieval, the need for further confirmatory evidence using an alternative prospective approach is warranted. At present, Testi-ICSI should be reserved for men with substantial AEF undergoing ART, particularly those who have experienced repeated ICSI failures and when measures to correct other underlying factors causing SDF have failed.

Inclusion Criteria:

• Infertile males with severe oligospermia (<5 mill/ml of spermatozoa in the ejaculate) or infertile males without severe oligospermia (>5 mill/ml of spermatozoa in the ejaculate) but with a previous complete ICSI failure. In addition, all of them have to have a sperm DNA fragmentation test level higher than 30% (SDF>30%), the threshold value for considering the result as abnormal.
• Women with adequate ovarian reserve, understood as those with AMH values >10pM, and Antral Follicular Count (AFC) >10.

Exclusion Criteria:

• Abnormal karyotype (previously known).
• Microdeletions in the Y chromosome (previously known).
• Carriers of known cystic fibrosis gene mutations.
• Presence of varicocele.
• Female age >38 years.
• Presence of uterine pathology that may condition reproductive outcomes (fibroids, uterine malformations).

The oocytes will be randomly assigned to one of the two study groups, and only the person who microinjects them will know which group each oocyte belongs to.