Assessment of reproductive outcomes of fresh versus cryopreserved ejaculated sperm samples-a retrospective analysis of 44423 oocyte donation ICSI cycles.

Abstract

Does the use of frozen sperm affect live birth rate (LBR) and cumulative LBR (CLBR) compared to fresh sperm samples in oocyte donation ICSI cycles? Although there were slight decreases in pregnancy rates (PRs) and LBR, as well as CLBR per embryo replaced and per embryo transfer (ET), when frozen sperm samples were used compared to fresh ejaculates, their clinical impact was limited. Sperm cryopreservation is part of the daily routine in reproduction clinics worldwide because of its many advantages in cycle planning. Nonetheless, there is a lack of agreement in terms of its impact on the outcomes of ICSI cycles. Previous studies showed conflicting conclusions and focused on different populations, which makes reaching consensus on the impact of sperm freezing-thawing complicated. Moreover, classical parameters are used to assess cycle success: pregnancy, live birth and miscarriage rates per ET. This study reports those measurements plus CLBR, which more accurately reflects the impact of the technique on the likelihood of achieving a newborn. A retrospective multicenter observational cohort study, including data from 37041 couples and 44423 ICSI procedures from January 2008 to June 2022, was carried out. The group using frozen sperm included 23852 transferred embryos and 108661 inseminated oocytes, whereas the fresh sample group comprised 73953 embryos replaced and 381509 injected oocytes. Outcomes measured per first ET and per ET were compared between groups using Fisher's exact test and Chi-squared test, as appropriate. Binary-logistics regression models were used to adjust the analyses according to clinically relevant co-variables. Kaplan-Meier curves plotted the CLBR per oocyte inseminated, per embryo replaced and per ET, and compared between groups using the Mantel-Cox test. Cox regressions were employed for the multivariate analyses of CLBR. The frozen sperm group showed a slightly lower biochemical (3.55% and 2.56%), clinical (3.68% and 3.54%) and ongoing (3.63% and 3.15%) PR compared to the cycles using fresh sperm, respectively, both per first ET and per ET. LBR was 4.57% lower per first ET and 3.95% lower per ET in the frozen sperm group than the fresh sperm group. There was also a subtle increase of 2.66% in biochemical miscarriage rate per ET when using frozen versus fresh sperm. All these differences remained statistically significant after the multivariate analysis (adjusted P ≤ 0.001). There were statistically significant differences in CLBR per embryo replaced and per ET but not per oocyte used (adjusted P = 0.071). Despite the statistical significance of the differences between the groups, those using frozen sperm required only 0.54 more oocytes injected, 0.45 more embryos transferred and 0.41 more ET procedures, on average, to achieve a live birth compared to the fresh samples. The retrospective nature of the study subjects the data to biases or potential errors during annotation on the source clinical and cycle records. This study uses multivariate analyses to control biases as much as possible. Using the oocyte donation model also contributes to reducing heterogeneity in the oocyte quality factor. The large sample sizes included in this study allowed for the detection of small changes in cycle success rates between groups. Although statistically significant, the decrease in PRs, LBR, and CLBR when using frozen sperm can be clinically overlooked in favor of the many benefits of sperm cryopreservation. None declared. Not applicable.