The effect of sperm DNA fragmentation on intracytoplasmic sperm injection outcome

Our study objective was to assess the effect of various sperm DNA fragmentation levels on clinical intracytoplasmic sperm injection outcome. This retrospective study included 392 patients who underwent ICSI and performed sperm DNA fragmentation testing before the procedure. Based on sperm DNA fragmentation cut‐off values, the patients were differentiated into 3 groups as <20%, 20%–30% and >30%. According to the female status, patients were differentiated into favourable group (n = 259) with female age <35 years and anti‐Mullerian hormone level ≥7.1 pmol/L; and unfavourable group (n = 133) with female age ≥35 years and anti‐Mullerian hormone level ≤7.1 pmol/L. The patient's medical records were reviewed, and patient's demographic, laboratory data including semen analysis, sperm DNA fragmentation determined by means of sperm chromatin dispersion, hormonal profile and data regarding intracytoplasmic sperm injection cycle were collected. This cohort reported that the clinical reproductive outcomes of intracytoplasmic sperm injection showed no statistical significance with increase sperm DNA fragmentation levels. In sperm DNA fragmentation above 30%, favourable females had significantly higher clinical pregnancy rate and live birth rate than unfavourable females, while fertilisation rate and miscarriage rate showed no significance between the subgroups. High sperm DNA fragmentation is linked to poor semen parameters.

total and progressive motility, vitality and morphology (World Health Organization, 2010). Yet, these parameters do not always predict male fertility, as infertile men might have normal semen parameters in up to 15% of cases (World Health Organization, 2010).
Alternatively, sperm DNA fragmentation (SDF) testing provides a different parameter for the analysis of the male factors affecting fertility. In the recent years, SDF has been identified as a predictor of fertility with good reliability because sperm DNA integrity can impact fertilisation, embryogenesis, implantation and pregnancy outcomes (Majzoub, Agarwal, Cho et al., 2017;. SDF may include breakage of single-strand DNA (ssDNA) or double-strand DNA (dsDNA), base deletion or modification, and inter-or intra-cross linkage (Sergerie et al., 2005). SDF occurs during late spermatogenesis due to defects in the repair system of DNA (Bui et al., 2018) which can be caused by different pathological mechanisms including apoptosis, elevated oxidative stress due to an increase in reactive oxygen species (ROS), and dysregulation of the chromatin protamine and histone components. These changes may occur as a result of various factors like drug use, tobacco smoking, environmental pollution, high testicular temperature, and advanced age (Sergerie et al., 2005). Furthermore, the impact of DNA damage on fertility outcome is not only believed to be influenced by sperm chromatin integrity, but also by the oocyte repair capacity. Oocytes play an important role in repairing SDF, depending on the oocyte's cytoplasmic and genomic quality (Fernández-Díez et al., 2016, García-Rodríguez et al., 2018, Setti et al., 2021. The exact mechanism by which the oocyte can repair the SDF remains unknown although human oocytes occupy DNA repair genes and can be linked to maternal mRNA repair (Osman et al., 2015).

Various tests for SDF have been identified with Sperm Chromatin
Structure Assay (SCSA), Sperm Chromatin Dispersion (SCD), Terminal deoxynucleotidyl transferase dUTP nick-end labelling (TUNEL) and Single Cell Gel Electrophoresis (COMET) being most commonly applied (Majzoub, Agarwal, Cho et al., 2017;. Recent efforts have identified implications of SDF on fertility and hence clinical guidelines were published highlighting the importance of SDF testing in couples with unexplained/idiopathic infertility, recurrent natural pregnancy loss, clinical varicocele, lifestyle exposures, recurrent intrauterine insemination (IUI) and in vitro fertilisation (IVF) failure and recurrent miscarriages following ICSI (Agarwal et al., 2019). There is a contradiction in literature with respect to the effect of SDF on fertilisation and pregnancy rates with ICSI. Evidence extracted from three meta-analyses has indicated that higher SDF is not associated with a negative impact on ICSI outcomes (Zini, 2011;Li et al., 2006;Zhao et al., 2014). On the contrary, another meta-analysis by Simon et al. including 24 ICSI studies revealed that SDF can have a significant impact on the pregnancy rate with an OR of 1.31 (95% CI 1.08-1.59, p = .006) (Simon et al., 2017).
To bridge the above-mentioned gap regarding the role of SDF on ICSI, our aim in the current study was to investigate the effect of different levels of SDF on ICSI in our centre with regard to fertilisation, clinical pregnancy, live birth and abortion.

| ME THODOLOGY
This is a retrospective cohort study which was carried out in the assisted conception unit at Hamad Medical Corporation, Doha, Qatar.
The study duration was over a 5-year period from 1 August 2014 to 1 August 2019. Ethical Approval from the Institutional Review Board was obtained (IRB MRC-01-19-349) for the protocols and procedures of this study.
The charts of 1922 patients who underwent ICSI were screened for inclusion in the study. Inclusion criteria were patients that underwent ICSI using ejaculated spermatozoa and had a recorded sperm DNA fragmentation (SDF) test done within a week before ICSI.
While couples with (a) severe male factor (severe oligozoospermia, azoospermia), (b) male genetic abnormalities, (c) female factor infertility (tubal factor, uterine abnormality, PCO), (d) history of mumps orchitis, (e) history of receiving chemotherapy or radiotherapy, and (f) history of testicular tumour were excluded.
The electronic medical records of male patients and their female partners were reviewed by two investigators for extraction of data including patients' age, related past medical history, any surgical procedure of relevance to fertility, consanguinity or history of infertility in the family. Clinical examination data involving general and local genital examination as well as laboratory data including semen analysis, SDF and hormonal profile (FSH, LH, Testosterone, PRL and E2) were extracted. Data retrieved for the female partners included basal FSH, anti-Mullerian hormone (AMH) and oestradiol. All information regarding the ICSI cycle was extracted from the medical records of the spouse including the protocol for ovarian stimulation and different ICSI outcomes.

| Semen analysis
Following 3-5 days of sexual abstinence, patients were asked to produce semen samples by masturbation. After liquefaction, semen analysis was carried out according to World Health Organization, 2010 standard protocol (WHO 5th edition guidelines).

| Sperm DNA fragmentation test
The Halosperm G2 test kit (Halotech, Madrid, Spain) was used to determine SDF. This test can be performed upon fresh and frozen semen samples. The SCD process is an indirect method of observing sperm DNA damage. During the test, controlled acid denaturation of the DNA and removal of nuclear proteins of the semen sample takes place. As a result, a large halo can be seen in sperm with intact DNA by using fluorescence microscope, while spermatozoon with dispersed chromatin materials or fragmented DNA will not produce a halo. SDF cut-off value was taken as ≥30% to differentiate between infertile and fertile men using a standard protocol (Fernández et al., 2005).

| Study protocol for ICSI
Patients performed an ICSI trial using ejaculated spermatozoa.
Ovarian stimulation in a controlled manner was started using recombinant FSH after desensitisation of pituitary with the dose guided by patient age, previous ovarian stimulation, oestradiol level and follicles scanned on ovarian transvaginal ultrasound scans (Yim et al., 2001). Continued monitored stimulation was done until at least two follicles matured and reached a mean diameter of 18 mm. Ovulation induction was then performed using 10,000 IU subcutaneous of hCG 36 hr before oocyte retrieval.
Oocytes were aspirated guided by transvaginal ultrasound.
Following ICSI, fertilised oocyte culturing was done, and the embryo quality was evaluated using the Veeck score. Embryo transfer was carried out on day 2 or 3 and patients were then daily given vaginal progesterone (Crinone 8%) for luteal phase support. (Veeck, 1988).

| Statistical analyses
The Shapiro-Wilk test for normality was performed to identify the distribution of the study variables. Frequencies were used to report categorical data, while median [95% CI] were used to present continuous values. Spearman's correlations were performed to assess the relationship between various study variables. The chi-square test was used to compare ICSI outcomes in various SDF levels. Kruskal-Wallis test was used to compare continuous variables between the three SDF levels. The Jonckheere-Terpstra test was used to test for an ordered difference in medians of the three SDF levels stating the direction of this order (trend). A p-value below .05 was considered statistically significant. All statistical analyses of collected data were performed using MedCalc ® Statistical Software version 19.8 (MedCalc Software Ltd).

| RE SULTS
The records of 1,922 patients who underwent ICSI from 2014 to 2019 were screened in our study. A total of 392 patients had SDF data within 7 days before the ICSI procedure were included after applying the inclusion and exclusion criteria. There were 63.5% of the patients complaining of primary infertility (n = 249) and 36.5% of secondary infertility (n = 143). The mean male age was 37.32 ± 6.7 years while the mean female age was 33.8 ± 6.1 years. Table 1 shows the clinical characteristics of the whole study population (n = 392) including semen parameters, hormonal analysis, and testicular size. As for the whole patients' samples, clinical pregnancy occurred in 44.1% (173/392) of the cases, live birth rate was 34.4% (121/352), and 9% (12/134) of patients had miscarriage. Table 2 shows correlations between SDF and different clinical parameters. There was a highly significant negative correlation between SDF and sperm count and motility, while there was a highly significant positive correlation between SDF and abnormal morphology. No correlation could be found between SDF and male age and progressive motility.
The patients were grouped according to SDF value into three groups, SDF <20% (n = 156), SDF 20%-30% (n = 149), and SDF >30% (n = 87). Table 3 shows the comparison between the three SDF groups with regard to semen and hormone parameters. Semen volume, sperm total motility and normal morphology were significantly different between the 3 groups with higher values found in lower SDF groups. No statistical significance was observed for age, testis site, sperm concentration, sperm progressive motility and hormones between SDF groups. However, there was a trend of better sperm concentration and lower FSH levels in patients with low SDF.
Fertilisation (p = .441), clinical pregnancy (p = .265) and live birth rates (p = .861) were greater in patients with lower SDF than in those with higher SDF. However, the difference was insignificant. Similarly,  our results, failed to report any significant correlation between male age and SDF (Sun et al., 1997;Winkle et al., 2009;Colin et al., 2010;Brahem et al., 2011;Nijs et al., 2011).

| D ISCUSS I ON
Our data show that there is no significance between SDF groups as regards clinical pregnancy, fertilisation rate, miscarriage rate and live birth rate. This result is in agreement with other studies.
Antonouli et al. used the SCD test to measure SDF within 150 patients and found no significant correlation between SDF and fertilisation rate in ICSI patients (Antonouli, 2019). Sun et al. also used SCD in ICSI patients and found no significant difference between <30% SDF and ≥30% groups as regards ICSI outcome (p = .458) (Sun et al., 2018). The impact of SDF on ICSI outcome is believed to be undermined by the sperm selection processes that usually choose the best quality sperm, i.e. with lower SDF, for the procedure. Indeed, Liffner et al. measured SDF levels in selected sperm using a colloid discontinuous gradient for patients with high SDF and reported levels that were even lower than those of normozoospermic donors (Liffner et al., 2019).
In the current study, SDF was found to have no significant effect on clinical pregnancy. This finding was in agreement with a study by Simon and coworkers who found that there was no significant correlation between SDF and clinical pregnancy following ICSI (Simon, 2013). Similarly, another systematic review by Agarwal and coworkers found that there was no correlation between SDF and clinical pregnancy rate (OR 0.94, 95% CI) (Cho and Agarwal, 2018).
Moreover, in a study by Antonouli et al. (2019) in a cohort of 150 couples with donated oocytes, SCD was utilised at a cut-off value of 25% to test its correlation with ICSI outcome. The study found that there was no statistically significant difference in pregnancy outcome between low SDF (< 25%) and high SDF (> 25%) groups (pregnancies 68.2% and 65.1% respectively). Additionally, Sun and coworkers concluded that using a 30% cut-off value, that there was no significant effect of SDF on IVF and ICSI outcomes using SCD (Tie-Cheng, 2018). On the contrary, in a recent meta-analysis on 24 ICSI studies, the authors concluded that there was a negative significant correlation between SDF and clinical pregnancy rate (OR 1.31, 95% CI, p < .007) (Simon, 2013 (Osman et al., 2015).
The main limitation of our study was the retrospective nature of the study with the possibility of incomplete data. However, we extensively searched in all medical records of both spouses in addition to assisted conception unit records to overcome this limitation. The data were also retrieved from one ART centre in Qatar; therefore, our data lacked diversity within methodologies for IVF and SDF testing, as well as lacked geographical distribution. However, the geographical distribution should not be a problem in our patients' cohort because we are serving patients from more than 117 countries from different geographical and ethnic backgrounds. A prospective, multicentre study on the effect of SDF on ICSI outcome is highly recommended to consolidate the data on this issue.

| CON CLUS ION
Sperm DNA fragmentation was found to be significantly correlated with conventional semen parameters highlighting its significance as a robust diagnostic test during male fertility evaluation.
In this study, while patients with higher SDF values had worse reproductive outcomes with ICSI, the results did not reach statistical significance. In case of older female age with lower ovarian reserve, if the husband has high sperm DNA fragmentation this will significantly affect ICSI outcome with regard to clinical pregnancy and live birth rate. Therefore, in these cases, more intervention is needed through sperm selection methods either intracytoplasmic morphologically selected sperm injection (IMSI), physiological intracytoplasmic sperm injection (PICSI) or using testicular sperm for ICSI.

ACK N OWLED G EM ENT
Open Access funding provided by the Qatar National Library.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request. Abbreviation: SDF, Sperm DNA fragmentation.