Sperm DNA damage and its role in IVF and ICSI

The role of sperm DNA fragmentation in IVF and ICSI

Numerous studies have attempted to assess the association between elevated sperm DNA fragmentation and ART outcomes. Unfortunately, variations between sperm DNA fragmentation assays, protocols, and thresholds and differences in study populations have resulted in systematic reviews and meta-analyses fraught with heterogeneity and unable to come to robust conclusions. To some degree, the controversy surrounding sperm DNA fragmentation is expected. If a male factor that adversely affects reproductive outcomes is present alongside a female factor known to have a strong deleterious effect on reproductive outcomes (such as increased female age), then the stronger female factor may obviate or even reverse the measurable adverse impact of sperm DNA fragmentation on reproductive outcomes. Furthermore, many of the studies have grouped patients undergoing IVF and ICSI together despite differences between the two techniques and have examined different clinical endpoints such as clinical pregnancy, miscarriage, and live birth.

The most recent systematic review and meta-analysis investigated 56 studies broken up into IVF (16 studies), ICSI (24 studies), and mixed IVF and ICSI (16 studies) and measuring sperm DNA damage either by SCSA, TUNEL, SCD, or Comet [31]. Overall, they found that sperm DNA damage predicts poor clinical pregnancy rates after IVF and/or ICSI (OR 1.68, 95 % CI 1.49–1.89, p??0.0001). When stratified by type of ART, the impact of sperm DNA damage on clinical pregnancy persisted (OR 1.65, 95 % CI 1.34–2.04, p??0.0001 and OR 1.31, 95 % CI 1.08–1.59, p??0.0068 for IVF and ICSI, respectively) [31]. In keeping with other previously published meta-analyses, this meta-analysis was limited by poorly controlled female factors and a high study heterogeneity (61 %, p??0.001), making it hard to rely on the odds ratios obtained. While acknowledging the limited trustworthiness of their odds ratios, the authors do attempt to explore the clinical relevance of sperm DNA damage in ART and note a greater clinical relevance for sperm DNA in IVF than in ICSI. With a median positive predictive value (PPV) of 79 %, a median negative predictive value (NPV) of 35 %, and a median clinical pregnancy rate of 32 % in IVF, information gleaned from sperm DNA damage can discriminate between expected IVF clinical pregnancy rates of 21 % (positive result) and 35 % (negative result) [31]. On the other hand, from a median PPV of 64 %, a median NPV of 40 %, and a median clinical pregnancy rate of 36 % in ICSI, sperm DNA damage results can only discriminate between expected ICSI clinical pregnancy rates of 36 % (positive result) and 40 % (negative result) [31].

Another recent systematic review and meta-analysis also sought to assess the effect of sperm DNA damage on live birth rates in IVF and ICSI. The authors identified six prospective cohort studies that investigated the impact of sperm DNA damage on live birth rates, with three using SCSA, two using TUNEL, and one using Comet to measure sperm DNA damage. Their meta-analysis for patients undergoing IVF, which comprised four studies with 553 patients, found a significantly higher live birth rate for men with low amount of sperm DNA damage than for those with high levels of sperm DNA damage (RR 1.27, 95 % CI 1.05–1.52, p?=?0.01) [32]. For patients undergoing ICSI, the authors pooled results from five studies comprising 445 patients and again found a significant increase in live birth rate for men with low levels of sperm DNA damage (RR 1.11, 95 % CI 1.00–1.23, p?=?0.04) [32]. In an attempt to control for female factors, the authors subsequently conducted a subgroup analysis that included only studies accounting for female factors (age and ovarian reserve). They identified only two studies with a very limited number of treated subjects, but found the impact of sperm DNA damage on IVF live birth rates was amplified significantly (RR 2.76, 95 % CI 1.59–4.80, p?=?0.0003) whereas the impact of sperm DNA damage on ICSI live birth rates became nonsignificant (RR 1.08, 95 % CI 0.39–2.96) [32]. This systematic review and meta-analysis, which assessed a highly clinically relevant outcome (live birth rate), suggests that higher levels of sperm DNA damage have a significant impact on live birth rates in IVF, but not in ICSI and advocates for ICSI as a potential therapeutic option for men with elevated levels of sperm DNA damage. Unfortunately, one cannot derive a robust conclusion from this systematic review as it is fraught with heterogeneity, with the six studies using both different methods to assess for sperm DNA damage and also different thresholds to define high versus low sperm DNA damage.

Despite the limitations of the aforementioned systematic review, its conclusions agree with some previously published studies. In particular, Zhao et al. conducted a systematic review and meta-analysis to investigate the impact of sperm DNA damage on clinical pregnancy and miscarriage rates in IVF and ICSI. The authors included sixteen cohort studies with 3106 couples undergoing IVF or ICSI and found a significant decrease in pregnancy rates for men with high DNA damage undergoing IVF (OR 0.66, 95 % CI 0.48–0.90, p?=?0.008) but not for those undergoing ICSI (OR 0.94, 95 % CI 0.70–1.25) [33]. On the other hand, they found a significant increase in miscarriage rates for men with high DNA damage undergoing ICSI (OR 2.68, 95 % CI 1.40–5.14, p?=?0.003), but not for those undergoing IVF (OR 1.84, 95 % CI 0.98–3.46) [33]. Of note, the authors also stratified their meta-analysis by sperm DNA damage assessment method and found that the different tests for sperm DNA damage behaved differently, with only sperm DNA damage as detected by TUNEL being significantly associated with decreased clinical pregnancy rates while only TUNEL and SCSA were significantly associated with increased miscarriage rates [33]. While this systematic review focused on the surrogate outcomes of clinical pregnancy and miscarriage rates rather than live birth rates, it echoes the review published by Osman et al. in suggesting an impact of sperm DNA damage on clinical pregnancy rates in IVF, but not in ICSI. On the other hand, despite the meta-analysis capturing only 47 miscarriages in the ICSI group and 70 miscarriages in the IVF group, it presents contradictory results showing an association between sperm DNA damage with higher miscarriage rates in ICSI, but not IVF. A further important insight was the indication that the various methods of measuring sperm DNA damage are not equivalent. Again, the results of this systematic review must be interpreted with caution due to heterogeneity, with differing threshold values for high versus low sperm DNA damage being used amongst the included studies and with the review unable to account for female factors.

Older systematic reviews and meta-analyses have looked separately at the impact of sperm DNA damage on miscarriage rates [34, 35] or on clinical pregnancy rates [36]. Unlike Zhao’s meta-analysis, which suggests a different impact for sperm DNA damage on IVF versus ICSI, Robinson [34] and Zini [35] both found that the significant associations between high sperm DNA damage and miscarriage rates did not depend on the method of fertilization used. Collins [36] similarly found a statistically and clinically significant association between sperm DNA damage and clinical pregnancy rates when they pooled data from IVF and ICSI studies. However, when the authors did subgroup analyses of IVF and ICSI studies separately, they were unable to demonstrate any statistically significant impact of sperm DNA damage on clinical pregnancy results [36]. Furthermore, while the effect of sperm DNA damage on reproductive outcomes was statistically significant, but not enough to change the decision to pursue assisted reproduction, they concluded that sperm DNA damage had limited clinical value in the initial evaluation of the infertile male. All three of these older systematic reviews face similar limitations as the more recently published meta-analyses, with a great deal of heterogeneity between the tests of sperm DNA damage used, threshold values for high versus low sperm DNA damage, control of female factors, and the use of surrogate outcomes (like clinical pregnancy) rather than live birth rate. Ultimately, as was noted by the American Society for Reproductive Medicine in 2013, despite the multitude of systematic reviews trying to tease out the impact of sperm DNA damage on outcomes in IVF and ICSI, heterogeneity has prevented any robust conclusions from being drawn on the clinical utility of sperm DNA damage prior to treatment with IVF and ICSI [37].

A potential explanation for the disparate findings seen in contemporary meta-analyses on the impact of sperm DNA damage when assessing clinical pregnancy and miscarriage rates in IVF and ICSI could lie in the difference in sperm selection process between the two methods of fertilization. Importantly, none of the methods of assessing sperm DNA damage permits the evaluation of the individual spermatozoon that goes on to fertilize the oocyte via either method of fertilization. In ICSI, sperm with normal morphology and progressive motility are selected for injection, which may improve fertilization and clinical pregnancy rates when compared to IVF, when fertilization depends on the natural selection of a healthy spermatozoon from a group with a high proportion of abnormal sperm. Studies have shown that in infertile men, morphologically normal motile sperm have higher rates of sperm DNA fragmentation (20–60 %) when compared to those in fertile men [38]. While sperm DNA damage appears to play a lesser role in fertilization and early development than abnormalities in centrosome function or oocyte-activation factor [39], the higher miscarriage rates in ICSI may be suggestive of a late paternal effect of abnormal sperm DNA fragmentation on embryo development [40]. Amongst couples who become pregnant, the natural selection of a spermatozoon in IVF may allow the selection of a near normal spermatozoon with lesser degrees of sperm DNA damage when compared to the artificial selection process used in ICSI, in which the selection of a morphologically normal motile spermotozoon may still mask significant sperm DNA damage that exert their influence not at fertilization, but later during embryonic development [40]. Further evidence of sperm DNA damage representing unmeasurable sperm defects in the entire semen sample rather than simply in the few severely damaged sperm detected in the test result comes from the fact that despite processed semen samples having lower levels of sperm DNA damage than neat (or unprocessed) semen samples, only sperm DNA damage as measured by SCSA in the neat (or unprocessed) semen sample is predictive of reproductive outcomes [41]. On the other hand, when using Comet, sperm DNA damage in both the neat and processed semen samples are predictive of reproductive outcomes [42].

While meta-analyses appear to uniformly suggest a deleterious impact for sperm DNA damage on clinical pregnancy and live birth rates in IVF, they are less clear on the effect of sperm DNA damage on ICSI outcomes. Unfortunately, the number of studies assessing ICSI alone is very limited, which makes it difficult to analyze the true effects of sperm DNA damage on ICSI results. Contemporary studies looking at ICSI results alone have corroborated the adverse effect of impaired sperm DNA integrity on ICSI outcomes, with Esteves et al. showing higher miscarriage and lower live birth rates in couples using ejaculated sperm with high sperm DNA damage (mean DFI 40.9 %) compared to those using testicular sperm with low sperm DNA damage (mean DFI 8.3 %) in a prospective cohort trial of 172 couples [43]. Similarly, Mehta et al. found that 50 % of couples who had previously failed one or more IVF-ICSI cycles using ejaculated sperm with a high sperm DNA damage (mean TUNEL 24.5 %) were able to achieve both pregnancy and live birth with ICSI using testicular sperm with low sperm DNA damage (mean TUNEL 4.6 %) [44]. Taken together, this contemporary data suggests a negative impact for high levels of sperm DNA damage on ICSI results.