There is a need for new markers that might better

Transcription

1 Journal of Andrology, Vol. 30, No. 3, May/June 2009 Copyright E American Society of Andrology Are Tests of Sperm DNA Damage Clinically Useful? Pros and Cons Review ARMAND ZINI*{ AND MARK SIGMAN{{ From the *Division of Urology, Department of Surgery, McGill University, Montreal, Quebec, Canada; and the ÀDivision of Urology, Department of Surgery, Brown University, Providence, Rhode Island. ABSTRACT: The advent of assisted reproductive technologies, particularly intracytoplasmic sperm injection (ICSI), has revolutionized the treatment of male-factor infertility. However, there are many unanswered questions regarding the safety of these techniques. These safety concerns are relevant because 1) these technologies often bypass the barriers to natural selection; 2) infertile men, particularly those with severe male-factor infertility, possess substantially more sperm DNA damage than do fertile men; and 3) experimentally, sperm DNA damage has been shown to adversely affect the developing embryo. This review discusses the etiology of sperm DNA damage, describes the individual tests of sperm DNA damage, and explores the relationship between sperm DNA damage and pregnancy outcomes. Based on a systematic review of the literature, sperm DNA damage is associated with lower natural, intrauterine insemination (IUI), and in vitro fertilization (IVF) pregnancy rates, but not with ICSI pregnancy rates. The literature also suggests that that sperm DNA damage is associated with an increased risk of pregnancy loss in those couples undergoing IVF or ICSI. Nonetheless, the true clinical utility of sperm DNA damage tests remains to be established, because the available studies are small and few in number and the study characteristics are heterogeneous. Although current data suggest that impaired sperm DNA integrity may have the greatest effect on IUI pregnancy rates and pregnancy loss by IVF and ICSI, further prospective studies are needed before testing should become a routine part of patient management. Key words: Fertility, ICSI, IVF, DNA integrity, meta-analysis. J Androl 2009;30: There is a need for new markers that might better discriminate infertile from fertile men, and that may predict pregnancy outcome and the risk of adverse reproductive events. Conventional semen parameters are poor predictors of reproductive outcomes. Although fertile men as a group have higher mean sperm parameters (concentration, motility, and morphology) than infertile men, there is significant overlap between these groups (Guzick et al, 2001). There is now some evidence to suggest that markers of sperm DNA integrity may help differentiate fertile from infertile men, but the clinical value of sperm DNA integrity testing remains to be defined (Kodama et al, 1997; Evenson et al, 1999; Spano et al, 2000; Zini et al, 2001). The study of sperm DNA damage is particularly relevant in an era in which advanced forms of assisted reproductive technologies (ARTs) are frequently used, because iatrogenic transmission of genetic defects is possible when natural selection barriers to fertilization Correspondence to: Dr Armand Zini, St Mary s Hospital, 3830 Lacombe Ave, Room 2304, Montreal, Quebec, Canada H3T 1M5 ( ziniarmand@yahoo.com). Received for publication September 23, 2008; accepted for publication December 1, { Dr Zini presents the pro viewpoint and Dr Sigman the con viewpoint. DOI: /jandrol are bypassed. Moreover, animal studies indicate that DNA damage is associated with both poor embryo development and poor pregnancy outcomes at in vitro fertilization (IVF; Fernández-Gonzalez et al, 2008). To date, the short- and long-term ramifications of successful fertilization and development with DNAdamaged spermatozoa are unknown. Clearly, the understanding that sperm DNA damage is common in infertile men, together with the concerning preliminary reports on genetic and epigenetic abnormalities in the offspring associated with intracytoplasmic sperm injection (ICSI), urges us to explore the subject of sperm DNA damage further. DNA that possesses measurable damage (specifically, DNA oxidation) may cause misreading errors to occur during DNA replication, and this may result in the generation of de novo mutations (Kuchino et al, 1987). Although the concept has not been tested in the context of mammalian reproduction, we cannot dismiss the possibility that successful fertilization with DNA-damaged sperm may cause de novo mutations in the offspring (despite the ability of the oocyte and embryo to repair this DNA damage; Ahmadi and Ng, 1999). Studies have found that children of fathers who smoked cigarettes preconceptually have a higher risk of developing childhood cancers (Ji et al, 1997). These studies suggest that there may be a link between sperm DNA damage and the subsequent development of childhood diseases. 219

2 220 Journal of Andrology N May ÙJune 2009 Human Sperm DNA and Chromatin Structure Unlike the relatively loose structure of chromatin in somatic cells, sperm chromatin is very tightly compacted by virtue of the unique associations between the DNA and sperm nuclear proteins (predominantly protamines; Ward and Coffey, 1991; Brewer et al, 1999). During the later stages of spermatogenesis, the spermatid nucleus is remodeled and condensed, and this is associated with the displacement of histones by transition proteins and then by protamines (Steger et al, 2000). The DNA strands are tightly wrapped around the protamine molecules, forming tight and highly organized loops (Brewer et al, 1999). Intermolecular and intramolecular disulfide cross-links between the cysteine-rich protamines are responsible for the compaction and stabilization of the sperm nucleus, and it is thought that this nuclear compaction is important to protect the sperm genome from external stresses such as oxidation or temperature elevation (Kosower et al, 1992). In humans, sperm chromatin is tightly packaged by protamines, but up to 15% of the DNA remains packaged by histones at specific DNA sequences (ie, there is a nonrandom association between histones and DNA sequences; Gatewood et al, 1987). The histonebound DNA sequences are less tightly compacted, and it is thought that these DNA sequences and/or genes may be involved in fertilization and early embryo development (Gatewood et al, 1987; Gineitis et al, 2000). Infertile men have a higher sperm histone to protamine ratio when compared with fertile controls (Oliva, 2006; Zhang et al, 2006). An excess of nuclear histones may result in poorer chromatin compaction and an increased susceptibility to DNA damage (Cho et al, 2001, 2003; Aoki et al, 2005, 2006). Etiology of Sperm DNA Damage The etiology of sperm DNA damage is multifactorial: it may be due to intrinsic (eg, protamine deficiency, excess reactive oxygen species [ROS] levels, apoptosis), or extrinsic factors (eg, testicular hyperthermia, environmental toxins). Sperm DNA damage is clearly associated with male infertility (and abnormal spermatogenesis) but a small percentage of spermatozoa from fertile men also possess detectable levels of DNA damage (Kodama et al, 1997; Evenson et al, 1999; Spano et al, 2000; Zini et al, 2001). Intrinsic Factors Sperm DNA damage has been associated with protamine deficiency (Cho et al, 2001; Aoki et al, 2005, 2006). Infertile men have an increased sperm histone to protamine ratio when compared with fertile controls, and an important subset of infertile men (,5% 15%), but not of fertile men, possesses a complete protamine deficiency (Zhang et al, 2006). Protamine deficiency (absolute or relative) can potentially result in defective chromatin compaction (Aravindan et al, 1997b) and in an increased susceptibility to DNA damage (Aoki et al, 2005). Studies on transgenic animal models with targeted protamine deficiency demonstrate a link between protamine deficiency, sperm DNA damage, and poor fertilizing capacity at IVF (Cho et al, 2001, 2003). The association between sperm DNA damage and protamine deficiency (Aoki et al, 2005, 2006) suggests that DNA damage may in part be due to a defect in spermiogenesis (the period during which sperm protamines are deposited; Steger et al, 2000). Sperm DNA damage has been associated with high levels of ROS (approximately 25% of infertile men have high levels of semen ROS; Zini et al, 1993; Irvine et al, 2000). Although the generation of low levels of ROS is necessary for normal sperm function, high levels of ROS (generated by defective spermatozoa and by semen leukocytes) can cause sperm dysfunction. The association between sperm DNA damage and sperm-derived ROS suggests that DNA damage may be due to a defect in spermiogenesis (Gomez et al, 1996), whereas the association between sperm DNA damage and leukocytederived ROS suggests that the DNA damage may be due to a posttesticular defect (eg, epididymitis; Ochsendorf, 1999). Sperm DNA damage may be the result of an aborted apoptosis (programmed cell death), although this theory has been challenged (Muratori et al, 2000; Sakkas et al, 2003). During normal spermatogenesis, germ cell apoptosis prevents overproliferation and selectively aborts abnormal sperm forms (Sinha Hikin and Swerdloff, 1999). Sakkas et al (2003) have proposed that some of the spermatozoa with DNA damage have initiated and then subsequently escaped apoptosis ( abortive apoptosis ). In line with this theory, Marcon and Boissoneault (2004) have suggested that DNA damage may be the result of incorrect repair of transient DNA nicks that are introduced during spermiogenesis (the nicks are introduced to relieve the torsional strain on the DNA helix during the exchange of histones to protamines). Advancing age and gonadotoxins have been associated with reduced levels of germ cell apoptosis in the testicle, and an increase in the percentage of ejaculated spermatozoa with DNA damage, suggesting that in these men, both spermatogenesis and apoptosis have been disrupted (Brinkworth and Nieschlag, 2000; Singh et al, 2003). Extrinsic Factors Extrinsic factors that may cause sperm DNA damage include drugs (eg, chemotherapy), cigarette smoking, genital tract inflammation, testicular hyperthermia, and varicoceles.

3 Zini and Sigman N Are Tests of Sperm DNA Damage Clinically Useful? 221 Young men with cancer (eg, Hodgkin lymphoma and testicular cancer) typically have poor semen quality and sperm DNA damage even prior to cancer-specific therapy (O Flaherty et al, 2008). They then experience cumulative dose damage during therapy (chemotherapy, radiation), often rendering them completely sterile (Fossa et al, 1997; Morris, 2002). The recovery of spermatogenesis may occur months to years after therapy, but evidence of sperm DNA damage may often persist beyond that period (Fossa et al, 1997). Patients who are scheduled to undergo definitive cancer therapy (surgery, chemotherapy, and/or radiation) are strongly encouraged to cryopreserve sperm for future use (Lee et al, 2006). Studies have shown that cigarette smoking is associated with lower sperm counts and motility and with an increase in abnormal sperm forms and sperm DNA damage (Potts et al, 1999). It is postulated that smoking causes increased leukocyte-derived ROS production with subsequent adverse effects on mature sperm (Potts et al, 1999). Exposure to pesticides (organophosphates) and air pollution have also been associated with increased levels of sperm DNA damage (Sanchez-Pena et al, 2004; Rubes et al, 2005). Posttesticular genital tract infection and inflammation, as seen clinically in epididymo-orchitis or prostatitis, result in leukocytospermia, and have been associated with increased levels of ROS and subsequent sperm DNA damage (Ereinpress et al, 2002). Testicular hyperthermia has been shown experimentally to cause sperm DNA damage, as well, as an increase in the histone to protamine ratio (Sailer et al, 1997; Banks et al, 2005). Clinically, certain behaviors (eg, hot baths, saunas) and occupations (eg, welding, baking, occupations that involve prolonged driving) have been associated with increased scrotal temperatures (Jung et al, 2002). Moreover, an association between hyperthermia and reduced male fertility potential has also been demonstrated (Thonneau et al, 1998). Corroborative clinical studies on sperm DNA damage and hyperthermia are lacking. Varicoceles have been associated with sperm DNA damage (Saleh et al, 2003). The level of sperm DNA damage is related to the high levels of oxidative stress found in the semen of these infertile men (Saleh et al, 2003). We have shown that varicoceles are associated with the abnormal retention of sperm cytoplasmic droplets (a morphologic feature associated with high levels of semen ROS) and that these retained droplets are correlated with sperm DNA damage in these infertile men (Zini et al, 2000; Fischer et al, 2003). Furthermore, sperm DNA integrity has been shown to improve after varicocele repair (Zini et al, 2005a). There is experimental evidence to show that hormonal deficiency can cause sperm chromatin defects. FSH receptor knockout mice have reduced levels of sperm nuclear protamines, lower testosterone, impaired fertility, and increased levels of DNA damage as compared with wild-type mice (Xing et al, 2003). Recently, investigators have shown that serum testosterone levels are inversely related to sperm DNA damage in a cohort of infertile men (Meeker et al, 2008). Tests of Sperm DNA Damage A variety of assays have been developed to measure sperm DNA damage (Aravindan et al, 1997a; Evenson et al, 1999; Chohan et al, 2006). The use of these tests has been driven largely by the growing use of ARTs and the concern that the integrity of the sperm genome is of importance in this context. It is important to consider several issues when evaluating studies of sperm DNA integrity. Although many assays profess to determine DNA integrity, it is important to understand what each assay actually measures (Table 1). Some procedures, such as the COMET and terminal deoxynucleotidyl transferase-mediated dutp nick end-labeling (TUNEL) assays, detect actual DNA strand breaks. Other approaches measure the susceptibility of DNA to denaturation that is, the formation of single-stranded DNA from native double-stranded DNA. These approaches depend on the fact that nicked DNA denatures more easily than double-stranded DNA. Finally, some assays rely on the differential binding of dyes to single-stranded and double-stranded DNA. Despite differences in assay methodology, the results of most assays correlate highly with each other, one exception being the manual acridine orange test (Chohan et al, 2006). Current assays are limited because they do not selectively differentiate clinically important DNA fragmentation from clinically insignificant fragmentation. A limited amount of singlestranded DNA breaks may be repaired by the oocyte, whereas double-stranded DNA breaks are irreversible (Ahmadi and Ng, 1999). In addition, some DNA nicking occurs as a normal process during winding and unwinding of DNA; current assays do not differentiate physiologic from pathologic nicking. Finally, the assays generally do not evaluate the genes that may be affected by the fragmentation; it is possible that fragmentation in areas containing certain genes may be more detrimental than fragmentation in relatively inactive regions of the genome. What the assays do attempt to do is simply to determine the amount of DNA fragmentation. This approach is based on the belief that more DNA nicking or fragmentation is pathologic. Most studies define an upper normal level of the

4 222 Journal of Andrology N May ÙJune 2009 Table 1. Basics of common sperm DNA integrity assays Basis of Assay Measured Parameter Direct assays TUNEL Adds labeled nucleotides to free DNA ends % Cells with labeled DNA Template independent Labels SS and DS breaks COMET Electrophoresis of single sperm cells % Sperm with long tails (tail length, DNA fragments form tail Intact DNA stays in head Alkaline COMET Alkaline conditions, denatures all DNA Identifies both DS and SS breaks Neutral COMET Does not denature DNA Identifies DS breaks, maybe some SS breaks % of DNA in tail) In situ nick translation Incorporates biotinylated dutp at SS DNA breaks with DNA polymerase I % Cells with incorporated dutp (fluorescent cells). Template-dependent Labels SS breaks, not DS breaks Indirect assays DNA break detection FISH Denatures nicked DNA Amount of fluorescence Whole genome probes bind to SS DNA proportional to number of DNA breaks SCD Individual cells immersed in agarose % Sperm with small or absent halos Denatured with acid then lysed Normal sperm produce halo Acridine orange flow cytometric assays (examples: SCSA, SDFA) Mild acid treatment denatures DNA with SS or DS breaks Acridine orange binds to DNA DS DNA (nondenatured) fluoresces green SS DNA (denatured) fluoresces red Flow cytometry counts thousands of cells DFI the percentage of sperm with a ratio of red to (red + green) fluorescence greater than the main cell population Acridine orange test Same as above, hand-counting of green and red cells % Cells with red fluorescence Abbreviations: DFI, DNA fragmentation index; DS, double-stranded; FISH, fluorescence in situ hybridization; SCD, sperm chromatin dispersion test; SCSA, sperm chromatin structure assay; SDFA, sperm DNA fragmentation assay; SS, single-stranded; TUNEL, terminal deoxynucleotidyl transferase-mediated dutp nick end-labeling. percentage of cells with DNA fragmentation. Samples with assay results above this threshold percentage are considered to have high DNA fragmentation. Each assay has advantages and disadvantages that impact its availability and utility (Table 2). Influence of Sperm DNA Damage on Natural and Assisted Pregnancy Several studies suggest that sperm DNA damage is associated with lower rates of natural insemination and intrauterine insemination (IUI) pregnancies. Indeed, couples in whom the husband has a high percentage of spermatozoa with DNA damage have very low potential for natural fertility and a prolonged time to pregnancy (Evenson et al, 1999; Spano et al, 2000; Loft et al, 2003). A recent meta-analysis indicates a strong association between sperm DNA damage and failure to achieve a natural pregnancy (Evenson and Wixon, 2008). High levels of sperm DNA damage have generally been associated with lower IUI pregnancy rates (Duran et al 2002; Muriel et al, 2006; Bungum et al, 2007). An estimation of the diagnostic odds ratio (OR) can only be obtained from the Bungum et al (2007) study (a table cannot be constructed from the data in the Duran et al, 2002, and Muriel et al, 2006, studies). The OR derived from the Bungum et al (2007) study (OR, 9.9; 95% confidence interval [CI], ; P,.001) indicates that sperm DNA damage is associated with a significantly lower pregnancy rate at IUI. Numerous studies have examined the influence of sperm DNA integrity on pregnancy rates after standard IVF. A systematic review and meta-analysis of IVF studies indicates that sperm DNA damage is associated (albeit weakly) with lower IVF pregnancy rates with a combined OR of 1.57 (95% CI, ; P,.05). The characteristics of the 9 evaluable IVF studies included in the meta-analysis are shown in Table 3. Nine otherwise valid studies were excluded from the meta-analysis because 1) they reported on a mixed population (IVF and IVF/ICSI) (Larson-Cook et al, 2003; Seli et al, 2004;

5 Zini and Sigman N Are Tests of Sperm DNA Damage Clinically Useful? 223 Table 2. Advantages and disadvantages of various DNA integrity assays Pros Cons Direct assays TUNEL Can perform on few sperm Thresholds not standardized Expensive equipment not required Variable assay protocols COMET Sensitive Labor intensive Can perform on few sperm Requires imaging software Alkaline: identifies all breaks Variable assay protocols Neutral: may identify more clinically relevant breaks Alkaline: may identify clinically unimportant fragmentation May induce breaks at alkaline-labile sites Neutral: less sensitive In situ nick translation Simple Unclear thresholds Less sensitive Indirect assays DNA break detection FISH Can perform on few sperm Limited clinical data SCD Easy, can use bright-field microscopy Limited clinical data Acridine orange flow Many cells rapidly examined Expensive equipment required cytometric assays Most published studies reproducible Small variations in lab conditions affect results Calculations involve qualitative decisions Manual acridine orange test Simple Difficulty with indistinct colors, rapid fading, heterogeneous staining Abbreviations: FISH, fluorescence in situ hybridization; SCD, sperm chromatin dispersion test; TUNEL, terminal deoxynucleotidyl transferasemediated dutp nick end-labeling. Virro et al, 2004; Payne et al, 2005; Velez de la Calle et al, 2008) or 2) a table could not be constructed (often because a cutoff or threshold DNA damage level was not reported; Tomlinson et al, 2001; Tomsu et al, 2002; Bakos et al, 2008; Meseguer et al, 2008). These observations are in keeping with a recent meta-analysis (Collins et al, 2008) and suggest that sperm DNA damage has a modest impact on pregnancy rates at IVF. A systematic review and meta-analysis of ICSI studies indicates that sperm DNA damage is not associated with ICSI pregnancy rates (combined OR, 1.14; 95% CI, , P 5.65). The characteristics of the 11 evaluable ICSI studies are shown in Table 4. Seven otherwise valid studies were excluded from the meta-analysis because 1) they reported on a mixed population (IVF and IVF/ICSI; Larson-Cook et al, 2003; Seli et al, 2004; Virro et al, 2004; Payne et al, 2005; Velez de la Calle et al, 2008), 2) a table could not be constructed (Bakos et al, 2008) or 3) the assay type used in the study is not widely recognized (Virant-Klun et al, 2002). These observations are in keeping with a recent meta-analysis (Collins et al, 2008) and suggest that sperm DNA damage has no measurable impact on pregnancy rates at ICSI. It is possible that the stringent process of sperm and embryo selection at ICSI (in humans) mitigates the potential adverse effect(s) of sperm DNA damage on reproductive outcomes (Gandini et al, 2004). Several studies have reported a relationship between sperm DNA damage and pregnancy loss after both standard IVF and IVF/ICSI (see Table 4). A systematic review and meta-analysis of IVF and ICSI studies shows that sperm DNA damage is associated with a significant increase in the rate of pregnancy loss after IVF and ICSI Table 3. Selected diagnostic properties of studies on sperm DNA damage and pregnancy after IVF Study n Assay %hdd Sens Spec PPV NPV OR 95% CI Host et al, TUNEL , 4.04 Henkel et al, TUNEL , 4.58 Huang et al, TUNEL , 2.56 Boe-Hansen et al, SCSA , Borini et al, TUNEL , 8.28 Lin et al, SCSA , 2.19 Benchaib et al, TUNEL , 2.00 Bungum et al, SCSA , 2.26 Frydman et al, TUNEL , 6.32 Abbreviations: %hdd, proportion of samples with high sperm DNA damage; CI, confidence interval; NPV, negative predictive value; OR, odds ratio; PPV, positive predictive value; SCSA, sperm chromatin structure assay; Sens, sensitivity; Spec, specificity; TUNEL, terminal deoxynucleotidyl transferase-mediated dutp nick end-labeling.

6 224 Journal of Andrology N May ÙJune 2009 Table 4. Selected diagnostic properties of studies on sperm DNA damage and pregnancy after ICSI Study n Assay %hdd Sens Spec PPV NPV OR 95% CI Host et al, TUNEL , 2.25 Henkel et al, TUNEL , 12.0 Gandini et al, SCSA , 2.08 Huang et al, TUNEL , 4.27 Zini et al, 2005b 60 SCSA , 3.22 Check et al, SCSA , 3.43 Boe-Hansen et al, SCSA , 2.72 Borini et al, TUNEL , 32.4 Benchaib et al, TUNEL , 3.41 Bungum et al, SCSA , 1.14 Lin et al, SCSA , 3.23 Abbreviations: %hdd, proportion of samples with high sperm DNA damage; CI, confidence interval; ICSI, intracytoplasmic sperm injection; NPV, negative predictive value; OR, odds ratio; PPV, positive predictive value; SCSA, sperm chromatin structure assay; Sens, sensitivity; Spec, specificity; TUNEL, terminal deoxynucleotidyl transferase-mediated dutp nick end-labeling. with a combined OR of 2.48 (95% CI, ; P,.0001). The characteristics of the 11 evaluable studies (from 7 reports) are shown in Table 5. There was no significant difference in the OR according to treatment type (IVF or ICSI). These data suggest that the adverse effect of sperm DNA damage on live birth rates is probably more significant than is reflected by the clinical pregnancy rates alone. The value of tests of sperm DNA integrity must be evaluated by the same criteria as any diagnostic test using a combination of statistical parameters and clinical judgment. As a test to diagnose infertility, the assay must identify patients with infertility by having a high sensitivity. In addition, fertile patients should test negative; that is, the assays should have a high specificity. These 2 properties infertile patients should test positive and fertile patients should test negative mean that pregnancy should be less common in those with a positive test (high DNA fragmentation) than in those with a negative test. This relationship is commonly evaluated by examining ORs the odds of pregnancy in those with a positive test divided by the odds of pregnancy in those with a negative test. It is important to understand that ORs tend to overestimate this relationship as compared with the likelihood ratio, which utilizes proportions instead of odds. Because ORs are not affected by the prevalence of the disease (infertility), a relatively high OR may have limited value in situations of low disease prevalence. In addition to having appropriate statistical characteristics, a diagnostic test of infertility must have clinical utility to have value beyond a research tool. That means the results should be able to be used to affect clinical management. To be useful as a clinical tool, the results should be repeatable and consistent over time. Equally important, the test must be able to be performed far enough in advance to make management decisions. Lastly, the difference in prognosis must be clinically significant, not just statistically significant. With these concepts in mind, the value of DNA integrity assays should be examined as Table 5. Selected diagnostic properties of studies on sperm DNA damage and PL after IVF and IVF/ICSI Study n ART Assay PL, % Abn Test, % Sens Spec PPV NPV OR (95% CI) Check et al, ICSI SCSA ( ) Zini et al, 2005b 60 ICSI SCSA ( ) Borini et al, IVF TUNEL ( ) Borini et al, ICSI TUNEL ( ) Benchaib et al, IVF TUNEL ( ) Benchaib et al, ICSI TUNEL ( ) Bungum et al, IVF SCSA ( ) Bungum et al, ICSI SCSA ( ) Frydman et al, IVF TUNEL ( ) Lin et al, IVF SCSA ( ) Lin et al, ICSI SCSA ( ) Abbreviations: Abn Test, proportion of abnormal sperm DNA test amongst documented pregnancies; ART, assisted reproductive technology; CI, confidence interval; ICSI, intracytoplasmic sperm injection; IVF, in vitro fertilization; NPV, negative predictive value; OR, odds ratio; PL, pregnancy loss; PPV, positive predictive value; SCSA, sperm chromatin structure assay; Sens, sensitivity; Spec, specificity; TUNEL, terminal deoxynucleotidyl transferase-mediated dutp nick end-labeling.

7 Zini and Sigman N Are Tests of Sperm DNA Damage Clinically Useful? 225 tests to predict failure of pregnancy by intercourse, IUI, IVF, or ICSI. As will become clear, these assays have much less value than they are often promulgated to have. Clinical Value of Tests of Sperm DNA Damage Three clinical scenarios are described below to illustrate the clinical value of sperm DNA testing. The pro and con statements (for sperm DNA testing) are based on 1) a systematic review of the literature, 2) the sperm DNA test characteristics (eg, reproducibility, sensitivity, positivity rate), and 3) disease prevalence (eg, pregnancy, pregnancy loss). First-pregnancy planners PRO argument: A.Z. These couples may benefit from testing because sperm DNA damage is associated with a significantly lower natural pregnancy rate (combined OR, 7.15; 95% CI, , P,.0001). An analysis of the 2 studies (with a median pregnancy rate of 64%), revealed a median positive predictive value (PPV) of 73% and median negative predictive value (NPV) of 68% (Evenson et al, 1999; Spano et al, 2000). This means that in populations with an overall pregnancy rate of 64%, the pregnancy rate is estimated at 27% when there is an abnormal test result and at 68% when the test result is normal. Thus, in this analysis, sperm DNA damage assessment provides clinically valuable information, because it can discriminate between pregnancy rates of 27% and 68%. CON argument: M.S. The value of tests of sperm DNA integrity is often stated be greatest in predicting pregnancy by intercourse and IUI. In one of the most commonly quoted studies (Evenson et al, 1999) an OR of 7.6 was reported (Evenson and Wixon, 2008); although the OR is statistically significant, it ignores the prevalence of infertility in the study population. Although 25.6% of the patients failed to achieve pregnancy, the test was abnormal in only 6.2% of patients. The test failed to identify infertility in 4 out of 5 infertile patients because the sensitivity of the assay was only 19%. In addition, in an age of concern about cost-effectiveness, it is worrisome that out of the 132 patients tested, only 6 (3.7%) patients were found to have high DNA fragmentation and did not achieve pregnancy; the remaining 96.3% of patients derived no benefit from the test. Out of all patients with an abnormal test result, pregnancy still occurred in 40% of the spouses. Finally, in both studies of pregnancy by intercourse, the predictive ability was best if the test was done close to the time of intercourse; the test performed more poorly when it was performed farther ahead of time, further limiting the clinical value in this population. IUI Candidates (Couples With Mild Male-Factor Infertility) PRO argument: A.Z. These couples may benefit from testing because sperm DNA damage is associated with a significantly lower IUI pregnancy rate (OR, 9.9; 95% CI, , P,.0001). The study reported by Bungum et al (2007), with a pregnancy rate of 20%, revealed a PPV of 97% and NPV of 24%. This means that in populations with an overall IUI pregnancy rate of 20%, the pregnancy rate is estimated at 3% when there is an abnormal test result and at 24% when the test result is normal. Thus, in this analysis, sperm DNA damage assessment provides clinically valuable information, because it can discriminate between pregnancy rates of 3% and 24%. Couples with high levels of sperm DNA damage should consider IVF and/ or ICSI rather than IUI. CON argument: M.S. Although an OR of 9.9 may be impressive, it would be wrong to conclude that the Bungum study demonstrates that couples planning IUI should first be tested for sperm DNA integrity (Bungum et al, 2007). Out of all samples tested, abnormal results were obtained in only 17%, whereas overall 80% of patients did not conceive. This poor test performance is because the sensitivity of the assay was only 20.7%: it failed to diagnose failure to conceive in 4 out of 5 cases. Although it performed better with IUI than with intercourse, only 16.6% of patients would have benefited from the test. In addition, this is the only study of the relationship between pregnancy by IUI and sperm DNA integrity that allows for proper statistical evaluation. More studies are needed before a change in practice patterns should be instituted. More concerning is the fact that the DNA fragmentation assay was performed on the same sample that was used for IUI hardly enough time to use the results to affect management. There has been no study demonstrating that sperm DNA integrity testing performed well in advance of IUI is predictive of pregnancy. Most worrisome, sperm DNA integrity is not always stable over time. A recent study examined a group of couples undergoing IUI, IVF, or ICSI for more than 1 cycle. In those couples with an initial normal DNA fragmentation, tests in subsequent cycles showed high DNA fragmentation in 15%. Even more worrisome was the finding that of those with initially high DNA fragmentation scores, 37% were found to have normal scores on subsequent testing (Erenpreiss et al, 2006). IVF or ICSI Candidates (Couples With Severe Male- Factor Infertility) PRO Argument: A.Z. These couples may benefit from testing because sperm DNA damage is associated with a significantly lower IVF pregnancy rate (combined OR, 1.57; 95% CI, ; P,.05). An analysis of the 9 IVF studies (with a median pregnancy rate of 33%) revealed a median PPV of 71% and median NPV of 34% (Table 3). Thus, in populations with an overall IVF pregnancy rate of 33%, sperm DNA

8 226 Journal of Andrology N May ÙJune 2009 damage assessment can discriminate between IVF pregnancy rates of 29% (positive test) and 34% (negative test), a small difference with modest clinical value. Couples with high levels of DNA damage may be better served by proceeding to ICSI, in which pregnancy rates are not related to sperm DNA damage test results (combined OR, 1.14; 95% CI, ; P 5.65, see Table 4). More importantly, these couples can benefit from testing because sperm DNA damage is associated with a significantly higher rate of pregnancy loss after IVF or ICSI (combined OR, 2.48; 95% CI, ; P,.0001). An analysis of the pregnancy loss studies (with a median pregnancy loss rate of 18%), revealed a median PPV of 37% and median NPV of 90%. This means that in populations with an overall pregnancy loss of 18%, the rate of pregnancy loss is estimated at 37% when there is an abnormal test result and at 10% when the test result is normal. Thus, sperm DNA damage assessment provides clinically valuable information, because it can discriminate between pregnancy loss rates of 37% and 10%. The effect of DNA damage on pregnancy loss should be discussed with patients prior to undergoing ART. CON Argument: M.S. The use of sperm DNA integrity testing in IVF candidates demonstrates the importance of understanding the difference between statistical significance and clinical significance. In the current meta-analysis, a statistically significant OR is obtained. However, the pregnancy rate in the normal (negative test) group was 34%, whereas in the abnormal result (positive test) group it was still 29%. It is unlikely that a 5% difference in pregnancy rates warrants the added expense (and increased potential risk) of ICSI. In addition, a recently published metaanalysis found no statistical significance to the OR in IVF cycles (Collins et al, 2008). This points out that the difference is so slight that different calculations with slightly different studies may give different statistical results. It is clear from the present analysis, as well as others, that in ICSI cycles, the assays have an OR no different from 1 and therefore have no predictive value for pregnancy. The most interesting data involves the value of sperm DNA integrity testing in predicting pregnancy loss. The current analysis demonstrates a statistically significant OR of 2.48 corresponding to a pregnancy loss rate of 10% with a normal (negative) test and 37% with an abnormal (positive) test. It certainly may be argued that this difference is also clinically significant. Despite these impressive numbers, it should be understood that the assay failed to identify 60% of the pregnancy loss cases because of a sensitivity of only about 40%. On the flip side, because of a low PPV, 67% of patients with an abnormal (positive) test will have a successful pregnancy and delivery. One can hardly tell a couple not to proceed with IVF/ICSI because they have high DNA fragmentation when they have a two-thirds chance that any pregnancy will go to term. In addition, because there are no proven therapies to improve DNA integrity in most patients, the result of the test will not change management. Summary Successful human reproduction depends on the inherent integrity of the sperm DNA. Indeed, there appears to be a threshold of sperm DNA damage beyond which embryo development and subsequent pregnancy outcome are impaired. There is now clinical evidence to show that, sperm DNA damage is detrimental to some reproductive outcomes and that the spermatozoa of infertile men possess substantially more DNA damage than that of fertile men. However, our understanding of the etiology/etiologies of sperm DNA damage, and the full impact of this sperm defect on reproductive outcomes in humans, remains rudimentary. Additional studies are needed to fully define the clinical value of sperm DNA damage testing and the optimal test to be used in this setting. Finally, therapies for impaired DNA integrity need to be developed and subsequent improvements in sperm DNA integrity from such therapies need to be correlated with improved reproductive outcomes. References Ahmadi A, Ng SC. Fertilizing ability of DNA-damaged spermatozoa. J Exp Zool. 1999;284: Aoki VW, Emery BR, Liu L, Carrell DT. Protamine levels vary between individual sperm cells of infertile human males and correlate with viability and DNA integrity. JAndrol. 2006;27: Aoki VW, Moskovtsev SI, Willis J, Liu L, Mullen JB, Carrell DT. DNA integrity is compromised in protamine-deficient human sperm. J Androl. 2005;26: Aravindan GR, Bjordahl J, Jost LK, Evenson DP. Susceptibility of human sperm to in situ DNA denaturation is strongly correlated with DNA strand breaks identified by single-cell electrophoresis. Exp Cell Res. 1997a;236: Aravindan GR, Krishnamurthy H, Mougdal NR. Enhanced susceptibility of follicle-stimulating-hormone-deprived infertile bonnet monkey (Macaca radiata) spermatozoa to dithiothreitolinduced DNA decondensation in situ. J Androl. 1997b;18: Bakos HW, Thompson JG, Feil D, Lane M. Sperm DNA damage is associated with assisted reproductive technology pregnancy. Int J Androl. 2008;31: Banks S, King SA, Irvine DS, Saunders PT. Impact of a mild scrotal heat stress on DNA integrity in murine spermatozoa. Reproduction. 2005;129: Benchaib M, Lornage J, Mazoyer C, Lejeune H, Salle B, Francois Guerin J. Sperm deoxyribonucleic acid fragmentation as a

9 Zini and Sigman N Are Tests of Sperm DNA Damage Clinically Useful? 227 prognostic indicator of assisted reproductive technology outcome. Fertil Steril. 2007;87: Boe-Hansen GB, Fedder J, Ersboll AK, Christensen P. The sperm chromatin structure assay as a diagnostic tool in the human fertility clinic. Hum Reprod. 2006;21: Borini A, Tarozzi N, Bizzaro D, Bonu MA, Fava L, Flamigni C, Coticchio G. Sperm DNA fragmentation: paternal effect on early post-implantation embryo development in ART. Hum Reprod. 2006;21: Brewer LR, Corzett M, Balhorn R. Protamine induced condensation and decondensation of the same DNA molecule. Science. 1999; 286: Brinkworth MH, Nieschlag E. Association of cyclophosphamideinduced male-mediated, foetal abnormalities with reduced paternal germ-cell apoptosis. Mutat Res. 2000;447: Bungum M, Humaidan P, Axmon A, Spano M, Bungum L, Erenpreiss J, Giwercman A. Sperm DNA integrity assessment in prediction of assisted reproduction technology outcome. Hum Reprod. 2007; 22: Check JH, Graziano V, Cohen R, Krotec J, Check ML. Effect of an abnormal sperm chromatin structural assay (SCSA) on pregnancy outcome following (IVF) with ICSI in previous IVF failures. Arch Androl. 2005;51: Cho C, Jung-Ha H, Willis WD, Goulding EH, Stein P, Xu Z, Schultz RM, Hecht NB, Eddy EM. Protamine 2 deficiency leads to sperm DNA damage and embryo death in mice. Biol Reprod. 2003;69: Cho C, Willis WD, Goulding EH, Jung-Ha H, Choi YC, Hecht NB, Eddy EM. Haploinsufficiency of protamine-1 or -2 causes infertility in mice. Nat Genet. 2001;28: Chohan KR, Griffin JT, Lafromboise M, DeJonge CJ, Carrell DT. Comparison of chromatin assays for DNA fragmentation evaluation in human sperm. J Androl. 2006;27: Collins JA, Barnhart KT, Schlegel PN. Do sperm DNA integrity tests predict pregnancy with in vitro fertilization? Fertil Steril. 2008;89: Duran EH, Morshedi M, Taylor S, Oehninger S. Sperm DNA quality predicts intrauterine insemination outcome: a prospective cohort study. Hum Reprod. 2002;17: Erenpreiss J, Bungum M, Spano M, Elzanaty S, Orbidans J, Giwercman A. Intra-individual variation in sperm chromatin structure assay parameters in men from infertile couples: clinical implications. Hum Reprod. 2006;21: Erenpreiss J, Hlevicka S, Zalkalns J, Erenpreisa J. Effect of leukocytospermia on sperm DNA integrity: a negative effect in abnormal semen samples. J Androl. 2002;23: Evenson DP, Jost LK, Marshall D, Zinaman MJ, Clegg E, Purvis K. de Angelis P, Claussen OP. Utility of the sperm chromatin assay as a diagnostic and prognostic tool in the human fertility clinic. Hum Reprod. 1999;14: Evenson DP, Wixon R. Data analysis of two in vivo fertility studies using Sperm Chromatin Structure Assay-derived DNA fragmentation index vs. pregnancy outcome. Fertil Steril. 2008;90: Fernández-Gonzalez R, Moreira PN, Pérez-Crespo M, Sánchez- Martín M, Ramirez MA, Pericuesta E, Bilbao A, Bermejo-Alvarez P, de Dios Hourcade J, de Fonseca FR, Gutiérrez-Adán A. Longterm effects of mouse intracytoplasmic sperm injection with DNAfragmented sperm on health and behavior of adult offspring. Biol Reprod. 2008;78: Fischer MA, Willis J, Zini A. Human sperm DNA integrity: correlation with sperm cytoplasmic droplets. Urology. 2003;61: Fossa SD, De Angelis P, Kraggerud SM, Evenson D, Theodorsen L, Clausen OP. Prediction of post-treatment spermatogenesis in patients with testicular cancer by flow cytometric sperm chromatin structure assay. Cytometry. 1997;30: Frydman N, Prisant N, Hesters L, Frydman R, Tachdjian G, Cohen- Bacrie P, Fanchin R. Adequate ovarian follicular status does not prevent the decrease in pregnancy rates associated with high sperm DNA fragmentation. Fertil Steril. 2008;89: Gandini L, Lombardo F, Paoli D, Caruso F, Eleuteri P, Leter G, Ciriminna R, Culasso F, Dondero F, Lenzi A, Spano A. Full-term pregnancies achieved with ICSI despite high levels of sperm chromatin damage. Hum Reprod. 2004;19: Gatewood JM, Cook GR, Balhorn R, Bradbury EM, Schmid CW. Sequence-specific packaging of DNA in human sperm chromatin. Science. 1987;236: Gineitis AA, Zalenskaya IA, Yau PM, Bradbury EM, Zalensky AO. Human sperm telomere-binding complex involves histone H2B and secures telomere membrane attachment. J Cell Biol. 2000;151: Gomez E, Buckingham DW, Brindle J, Lanzafame F, Irvine DS, Aitken RJ. Development of an image analysis system to monitor the retention of residual cytoplasm by human spermatozoa: correlation with biochemical markers of the cytoplasmic space, oxidative stress, and sperm function. J Androl. 1996;17: Guzick DS, Overstreet JW, Factor-Litvak P, Brazil CK, Nakajima ST, Coutifaris C, Carson SA, Cisneros P, Steinkampf MP, Hill JA, Xu D, Vogel DL, National Cooperative Reproductive Medicine Network. Sperm morphology, motility, and concentration in infertile and fertile men. N Engl J Med. 2001;345: Henkel R, Kierspel E, Hajimohammad M, Stalf T, Hoogendijk C, Mehnert C, Menkveld R, Schill WB, Kruger TF. DNA fragmentation of spermatozoa and assisted reproduction technology. Reprod Biomed Online. 2003;7: Host E, Lindenberg S, Smidt-Jensen S. The role of DNA strand breaks in human spermatozoa used for IVF and ICSI. Acta Obstet Gynecol Scand. 2000;79: Huang CC, Lin DP, Tsao HM, Cheng TC, Liu CH, Lee MS. Sperm DNA fragmentation negatively correlates with velocity and fertilization rates but might not affect pregnancy rates. Fertil Steril. 2005;84: Irvine DS, Twigg JP, Gordon EL, Fulton N, Milne PA, Aitken RJ. DNA integrity in human spermatozoa: relationships with semen quality. J Androl. 2000;21: Ji BT, Shu XO, Linet MS, Zheng W, Wacholder S, Gao YT, Ying DM, Jin F. Paternal cigarette smoking and the risk of childhood cancer among offspring of nonsmoking mothers. J Natl Cancer Inst. 1997;89: Jung A, Schill WB, Schuppe HC. Genital heat stress in men of barren couples: a prospective evaluation by means of a questionnaire. Andrologia. 2002;34: Kodama H, Yamaguchi R, Fukuda J, Kasai H, Tanaka T. Increased oxidative deoxyribonucleic acid damage in the spermatozoa of infertile male patients. Fertil Steril. 1997;68: Kosower NS, Katayose H, Yanagimachi R. Thiol-disulfide status and acridine orange fluorescence of mammalian sperm nuclei. J Androl. 1992;13: Kuchino Y, Mori F, Kasai H, Inoue H, Iwai S, Miura K, Ohtsuka E, Nishimura S. Misreading of DNA templates containing 8-hydroxydeoxyguanosine at the modified base and at adjacent residues. Nature. 1987;327: Larson-Cook KL, Brannian JD, Hansen KA, Kasperson KM, Aamold ET, Evenson DP. Relationship between the outcomes of assisted reproductive techniques and sperm DNA fragmentation as measured by the sperm chromatin structure assay. Fertil Steril. 2003;80:

10 228 Journal of Andrology N May ÙJune 2009 Lee SJ, Schover LR, Partridge AH, Patrizio P, Wallace WH, Hagerty K, Beck LN, Brennan LV, Oktay K. American Society of Clinical Oncology recommendations on fertility preservation in cancer patients. J Clin Oncol. 2006;24: Lin MH, Kuo-Kuang Lee R, Li SH, Lu CH, Sun FJ, Hwu YM. Sperm chromatin structure assay parameters are not related to fertilization rates, embryo quality, and pregnancy rates in in vitro fertilization and intracytoplasmic sperm injection, but might be related to spontaneous abortion rates. Fertil Steril. 2008;90: Loft S, Kold-Jensen T, Hjollund NH, Giwercman A, Gyllemborg J, Ernst E, Olsen J, Scheike T, Poulsen HE, Bonde JP. Oxidative DNA damage in human sperm influences time to pregnancy. Hum Reprod. 2003;18: Marcon L, Boissonneault G. Transient DNA strand breaks during mouse and human spermiogenesis new insights in stage specificity and link to chromatin remodeling. Biol Reprod. 2004;70: Meeker JD, Singh NP, Hauser R. Serum concentrations of estradiol and free T4 are inversely correlated with sperm DNA damage in men from an infertility clinic. J Androl. 2008;29: Meseguer M, Martinez-Conejero JA, O Connor JE, Pellicer A, Remohi J, Garrido N. The significance of sperm DNA oxidation in embryo development and reproductive outcome in an oocyte donation program: a new model to study a male infertility prognostic factor. Fertil Steril. 2008;89(5): Morris ID. Sperm DNA damage and cancer treatment. Int J Androl. 2002;25: Muratori M, Piomboni P, Baldi E, Filimberti E, Pecchioli P, Moretti E, Gambera L, Baccetti B, Biagiotti R, Forti G, Maggi M. Functional and ultrastructural features of DNA-fragmented human sperm. J Androl. 2000;21: Muriel L, Meseguer M, Fernandez JL, Alvarez J, Remohi J, Pellicer A, Garrido N. Value of the sperm chromatin dispersion test in predicting pregnancy outcome in intrauterine insemination: a blind prospective study. Hum Reprod. 2006;21: Ochsendorf FR. Infections in the male genital tract and reactive oxygen species. Hum Reprod Update. 1999;5: O Flaherty C, Vaisheva F, Hales BF, Chan P, Robaire B. Characterization of sperm chromatin quality in testicular cancer and Hodgkin s lymphoma patients prior to chemotherapy. Hum Reprod. 2008;23: Oliva R. Protamines and male infertility. Hum Reprod Update. 2006;12: Payne JF, Raburn DJ, Couchman GM, Price TM, Jamison MG, Walmer DK. Redefining the relationship between sperm deoxyribonucleic acid fragmentation as measured by the sperm chromatin structure assay and outcomes of assisted reproductive techniques. Fertil Steril. 2005;84: Potts RJ, Newbury CJ, Smith G, Notarianni LJ, Jefferies TM. Sperm chromatin damage associated with male smoking. Mutat Res. 1999;423: Rubes J, Selevan SG, Evenson DP, Zudova D, Vozdova M, Zudova Z, Robbins WA, Perreault SD. Episodic air pollution is associated with increased DNA fragmentation in human sperm without other changes in semen quality. Hum Reprod. 2005;20: Sailer BL, Sarkar LJ, Bjordahl JA, Jost LK, Evenson DP. Effects of heat stress on mouse testicular cells and sperm chromatin structure. J Androl. 1997;18: Sakkas D, Seli E, Bizzaro D, Tarozzi N, Manicardi GC. Abnormal spermatozoa in the ejaculate: abortive apoptosis and faulty nuclear remodeling during spermatogenesis. Reprod Biomed Online. 2003;7: Saleh RA, Agarwal A, Sharma RK, Said TM, Sikka SC, Thomas AJ Jr. Evaluation of nuclear DNA damage in spermatozoa from infertile men with varicocele. Fertil Steril. 2003;80: Sanchez-Pena LC, Reyes BE, Lopez-Carrillo L, Recio R, Moran- Martinez J, Cebrian ME, Quintanilla-Vega B. Organophosphorous pesticide exposure alters sperm chromatin structure in Mexican agricultural workers. Toxicol Appl Pharmacol. 2004;196: Seli E, Gardner DK, Schoolcraft WB, Moffatt O, Sakkas D. Extent of nuclear DNA damage in ejaculated spermatozoa impacts on blastocyst development after in vitro fertilization. Fertil Steril. 2004;82: Singh NP, Muller CH, Berger RE. Effects of age on DNA doublestrand breaks and apoptosis in human sperm. Fertil Steril. 2003;80: Sinha Hikin AP, Swerdloff RS. Hormonal and genetic control of germ cell apoptosis in the testis. Rev Reprod. 1999;4: Spano M, Bonde JP, Hjollund HI, Kolstad HA, Cordelli E, Leter G. Sperm chromatin damage impairs human fertility. Fertil Steril. 2000;73: Steger K, Pauls K, Klonisch T, Franke FE, Bergmann M. Expression of protamine-1 and -2 mrna during human spermiogenesis. Mol Hum Reprod. 2000;6: Thonneau P, Bujan L, Multigner L, Mieusset R. Occupational heat exposure and male fertility: a review. Hum Reprod. 1998;13: Tomlinson MJ, Moffatt O, Manicardi GC, Bizzaro D, Afnan M, Sakkas D. Interrelationships between seminal parameters and sperm nuclear DNA damage before and after density gradient centrifugation: implications for assisted conception. Hum Reprod. 2001;16: Tomsu M, Sharma V, Miller D. Embryo quality and IVF treatment outcomes may correlate with different sperm comet assay parameters. Hum Reprod. 2002;17: Velez de la Calle JF, Muller A, Walschaerts M, Clavere JL, Jimenez C, Wittemer C, Thonneau P. Sperm deoxyribonucleic acid fragmentation as assessed by the sperm chromatin dispersion test in assisted reproductive technology programs: results of a large prospective multicenter study. Fertil Steril. 2008;90: Virant-Klun I, Tomazevic T, Meden-Vrtovec H. Sperm singlestranded DNA, detected by acridine orange staining, reduces fertilization and quality of ICSI-derived embryos. J Assist Reprod Genet. 2002;19: Virro MR, Larson-Cook KL, Evenson DP. Sperm chromatin structure assay (SCSA) parameters are related to fertilization, blastocyst development, and ongoing pregnancy in in vitro fertilization and intracytoplasmic sperm injection cycles. Fertil Steril. 2004;81: Ward WS, Coffey DS. DNA packaging and organization in mammalian spermatozoa: comparison with somatic cells. Biol Reprod. 1991;44: Xing W, Krishnamurthy H, Sairam MR. Role of follitropin receptor signaling in nuclear protein transitions and chromatin condensation during spermatogenesis. Biochem Biophys Res Commun. 2003;312: Zhang X, San Gabriel M, A Zini A. Sperm nuclear protamine to histone ratio in fertile and infertile men: evidence of heterogeneous sub-populations of spermatozoa in the ejaculate. J Androl. 2006;27: Zini A, Bielecki R, Phang D, Zenzes MT. Correlations between two markers of sperm DNA integrity, DNA denaturation and DNA fragmentation, in fertile and infertile men. Fertil Steril. 2001;75: Zini A, Blumenfeld A, Libman J, Willis J. Beneficial effect of microsurgical varicocelectomy on human sperm DNA integrity. Hum Reprod. 2005a;20: