DNA Integrity Tests in Clinical Practice: It Is Not a Simple Matter of Black and White (or Red and Green)

Abstract

Note: Postings to Androlog have been lightly edited before publication. In vitro fertilization (IVF) presents a significant physical, emotional, and financial commitment on the part of the prospective parents. Couples often cling to hopes of successful IVF even when presented with adverse odds. Thus, any information that increases the accuracy of the estimate of success is of paramount importance. This is illustrated by a question from Mark Jutras, titled “Extremely Abnormal SCSA”: Have a guy seen for oligoasthenoteratozoospermia. He was found to be hypothalamic and was treated with hCG [human choriogonadotropin] 1500 SQ 3 times a week with normalization of T. Several months later there was some improvement in count but no normal sperm on any sample and all samples showing 50%-60% with 2 tails. This was also true before treatment with hCG. Specimen sent for SCSA [sperm chromatin structure assay] with call from [Donald] Evenson. Almost no signal in the normal range. Cannot even assign numbers to the value. His chromosomes were normal. They want to do IVF but in the current situation we have to tell them that we don't expect a pregnancy. Best plan I can come up with is since a repeat SCSA is indicated anyway, continue hCG and start Fertile One and recheck SCSA after 3 months of Fertile One. Will prep as for IVF/ICSI [intracytoplasmic sperm injection] and do the SCSA on that sample. My tech is skeptical that she can get anything through a column with current samples. If no go at that time thought I would send him to Cornell Urology for second opinion. Anyone else want him? Ideas and thoughts welcomed. Peter Schlegel was not surprised that the SCSA was abnormal given the gross morphological problems: The 2-tailed sperm are usually diploid. Sperm chromosome analysis (FISH [fluorescence in situ hybridization]) may be worthwhile to test this possibility before IVF/ICSI. Sounds like a failure of spermatid separation—no surprise that DNA integrity is so poor. Jutras agreed with Schlegel's assessment and brought up an interesting anecdote on another patient with an abnormal SCSA: Brings up another interesting case we had. Couple with recurrent spontaneous abortion. Standard evaluation including genetics on both were normal. SCSA showed 2 distinct populations. About 50% of sperm had normal content of DNA and 50% were diploid. Best solution would have been IVF with PGD [preimplantation genetic diagnosis]. Since couple had one normal daughter they decided best solution was vasectomy. The reliance on SCSA results to predict clinical outcomes was questioned by this author (Craig Niederberger), who brought up the recent study by Payne et al (2005) on the predictive value of SCSA in assisted reproductive technology (ART): Regarding the patient with an abnormal SCSA: it's important to keep in mind that the original reported outcomes of this test have not yet been clinically verified by independent investigators. Payne and coauthors recently reported in an independent study that SCSA outcomes failed to identify a DFI [DNA fragmentation index] threshold above which pregnancy outcomes were negative after IVF or IVF/ICSI. In fact, almost half of the 19 couples with a DFI greater than 27% had clinical pregnancies. Interestingly, patients with low DFI (defined as less than 9%), what one would expect to have the highest quality sperm by this assay, were the least likely to become pregnant. Peter Schlegel concurred that more studies are required to further define the clinical relevance of DNA integrity testing: I agree with Dr Niederberger that additional information on SCSA and its relationship to pregnancy outcome is needed. Many studies have been done on this subject with highly variable findings, but essentially none that mirror the predictive value of SCSA for natural or IVF-ICSI cycles suggested by Evenson et al and Larson et al [Larson et al, 2000; Larson-Cook et al, 2003] in their original publications. Dr Schlegel then went on to give a nuanced view of the value of DNA integrity tests. While citing studies showing an association of paternal DNA damage with poor embryo development and recurrent miscarriage, he cautioned against using absolute SCSA cutoff points to predict pregnancy outcomes: Unfortunately, the study design of some of the additional reports has been poor. Payne et al [2005] essentially tested the hypothesis that DFI was the only variable affecting IVF outcome. No one would expect this to be true and, ignoring other known factors that affect IVF outcome (eg, female age), may well have explained the apparently contradictory results of this study. So, we are left with a field that has conflicting data. I do believe SCSA (and other tests of DNA integrity) have some value in identifying male factors involved with fertility. For example, patients with recurrent miscarriage have a substantive increase in DNA breaks than couples undergoing IVF for other indications or fertile donors (Carrell et al, 2003). In addition, the observation of increased DNA breaks in sperm from couples with poor embryo development suggests a possible role of this test in identifying a male factor (Tesarik et al, 2004). Interestingly, DNA breaks (abnormal SCSA results) do not appear to predict fertilization failure. At present, I see tests like DNA integrity evaluations (TUNEL [terminal uridine nick end labeling], SCSA, Comet) as well as sperm aneuploidy testing as allowing us new insight into male factor contributions to infertility that we've never considered before. However, as Dr Niederberger appropriately points out, there is no absolute cut-point for SCSA (as originally suggested) that will preclude success for natural pregnancy nor IVF/ICSI. On the other hand, I expect we will find that the man with 70% DFI (DNA fragmentation index) has a lower chance of contributing to a pregnancy than other couples in the same situation. Certainly we (and many other centers) have reported pregnancies with sperm obtained from samples with high DFI levels. However, having fragmented and unstable DNA cannot be good for sperm. Confounding factors that complicate the interpretation of DNA integrity testing were neatly summarized and discussed by Juan Alvarez. He first pointed out that different tests measure distinct aspects of DNA: I recently published a letter to the editor in Human Reproduction in which I address the issue of the predictive value of the SCSA test in ART (Alvarez, 2005). First, I point out that we have to differentiate between tests that measure “real” DNA fragmentation vs tests that measure “susceptibility” to DNA denaturation. In the former group we have tests like TUNEL, in situ nick translation and COMET under neutral pH conditions. In the latter group we have tests like the SCSA, COMET under alkaline or acidic conditions, SCD [sperm chromatin dispersion], chromomycin, etc. Dr Alvarez also discussed the confounder of heterogeneity in the sperm population. He questioned whether the abnormal sperm as detected by SCSA are a distinct subpopulation of otherwise healthy sperm or whether they are the worst representatives of a globally damaged sperm population (ie, the “tip of the iceberg”; Evenson et al, 1999, 2002): Another factor that should be considered is whether DNA damage affects the whole sperm population in a given semen sample or only a fraction of it. For example, DNA fragmentation induced by apoptosis usually affects a fraction of the spermatozoa whereas DNA fragmentation induced by oxidative stress and radiation therapy leads to double-stranded DNA fragmentation and nucleotide damage of the 8-OH-2-deoxyguanosine (oxo8dG) type in most spermatozoa. DNA fragmentation induced by the hydroxyl radical results in the formation of oxo8dG in a first stage followed by double-stranded DNA fragmentation thereafter (Cui et al, 2000). While DNA damage of the first type could be repaired to some extent by the oocyte, double-stranded DNA damage is virtually irreversible and incompatible with the development of a viable pregnancy. Since DNA fragmentation values in ejaculated spermatozoa above 10%, as assessed by TUNEL (Benchaib et al, 2003), or above 30%, as assessed by the SCSA test (Evenson et al, 1999), are associated with low pregnancy rates, one would think that the remaining 90% or 70% of the spermatozoa, respectively, could fertilize the egg and result in a viable pregnancy. However, in addition to double-stranded breaks, a significant proportion of these spermatozoa could have DNA nucleotide modifications of the oxo8dG type. Therefore, the probability that a spermatozoon with normal DNA would fertilize the egg would be much lower than that expected from a DNA fragmentation value of 10% or 30%, respectively. That is, in addition to the measurable 10% and 30% of spermatozoa with DNA fragmentation, the remaining 90% and 70% of spermatozoa would have some type of DNA damage that is not compatible with the development of a viable pregnancy. This concept has been designated as the “iceberg effect” (Evenson et al 1999). Another confounder is that not all DNA damage is lethal. Dr Alvarez reminded us that the majority of DNA is noncoding and that the oocyte is capable of DNA repair. Another factor is whether DNA damage affects exons vs introns. Since more than 90% of DNA is comprised of non protein-coding regions or introns, DNA damage is most likely to affect these noncoding regions of DNA and, therefore, may not affect embryo development. This could explain, at least in part, why relatively high levels of sperm DNA fragmentation can result in viable pregnancies. Another important factor is the capability of the oocyte to repair sperm DNA damage. DNA damage in the fertilizing spermatozoon may be repaired by the oocyte. This is most likely to occur in those cases where female age is <35 years. Dr Alvarez then concluded that these confounding factors place inherent limitations on the predictive value of DNA integrity tests. He proposed combining DNA integrity tests with measures of reactive oxygen species damage to get a clearer picture of sperm DNA quality: Therefore, based on the above, it can be concluded that the predictive value of sperm DNA fragmentation tests will always have DNA region and oocyte-derived uncertainty factors and, cannot have a negative predictive as was originally suggested by the “iceberg effect” proposed by Evenson et al The of DNA fragmentation and nucleotide modifications of the oxo8dG type in sperm DNA may the predictive value of DNA fragmentation tests in when using tests like TUNEL, in situ nick or COMET under neutral pH that measure “real” DNA damage. tests should have a predictive value than those that measure “susceptibility” to DNA like the This is by a recent by et al, in which they showed that of spermatozoa with TUNEL test values in a pregnancy of of spermatozoa with TUNEL values or in a second in these same couples in a pregnancy of et al, 2005). Peter Schlegel further on the issue of sperm population He that since the predictive value of SCSA is when to and not the fraction in then SCSA be some that affects the sperm Dr Alvarez has summarized some of the on DNA integrity and IVF Several points and of the were in these In the original studies reported by Evenson et al et al, Larson et al, 2000; Larson-Cook et al, DNA integrity was by SCSA in the semen This result predict IVF whereas the sperm SCSA test was less So, the sperm for IVF/ICSI not have many DNA The of the sperm sample as a whole to DNA breaks with treatment a DNA oxygen other that may affect IVF including on et al and et al embryo development and results that we should the concept of sperm with DNA in a semen sample with abnormal DNA integrity and that abnormal DNA integrity results may have an on and assisted a the proposed by may not be the best when the on patients who have SCSA, TUNEL, or it is interesting that these evaluations other in most whereas of DNA damage (ie, may have different results with the tests for DNA this a of the clinical value of SCSA by Evenson. He by a of the study by Payne et The of the by Payne et al (2005) the relationship between sperm fragmentation as by the sperm chromatin structure and outcomes of assisted reproductive is The the integrity of sperm the the pregnancy outcome. This in to the couple of published on and studies that the integrity of the sperm DNA is highly with good embryo development and normal pregnancy. Payne et al to the relationship between SCSA and ART with which is an to a that the SCSA sperm samples in be most of these couples had female infertility The couples in the Payne et al study female infertility 2 and male If these female patients were the of sperm in the less than DFI this factor could why pregnancy were lower in this In the Payne et al (2005) a of were to of 3 whether any patient had more than 3 and and this to the of the in on the of obtained and If DNA fragmentation affects some of the these will to to or after to abortion. However, the of the may be normal. For example, the of obtained in a couple is are of poor quality to sperm DNA and 2 are of high of these 2 may result in pregnancy and one may that DNA fragmentation is not with poor outcome in This would be most in a case where are obtained and all In this DNA fragmentation will a negative with pregnancy association between sperm DNA fragmentation, poor embryo quality (Tesarik et al and failed pregnancy in et al, et al, has been Dr Evenson then on the brought up by Dr Alvarez the of DNA damage with and the negative predictive value of the SCSA I the by Dr Juan Alvarez on the relevance of the of DNA If this were by apoptosis breaks the negative predictive value would be much lower than it were by oxidative stress where most of the DNA is damage damage DNA damage nucleotide damage as damage in cases of apoptosis and DNA is in cases of oxidative stress there is damage. Therefore, the DFI value is but DNA fragmentation was by there is a chance that a healthy spermatozoon would fertilize the oocyte. to be a that most of the ejaculated sperm DNA damage is by may also In that the SCSA 30% DFI threshold may the of and damage. The published by et al (2005) that The is that most of those patients had sperm DNA fragmentation as by the that of the had DNA fragmentation as by TUNEL, of in ejaculated spermatozoa and in spermatozoa only male had values in spermatozoa. suggested by Dr had the levels of oxo8dG been they would have been and “iceberg effect” (Evenson and would have in this That is likely why the pregnancy with ejaculated spermatozoa in the first was and with spermatozoa. Dr Evenson then that the DNA integrity tests the SCSA is for in the Payne et al (2005) it was pointed out by Dr Schlegel (2005) that all of the DNA fragmentation tests are in their of a threshold that a man at an infertility is not at this time what between TUNEL and SCSA data. However, significant between SCSA and TUNEL have been et al with have also been between TUNEL and SCSA for and et al, While the TUNEL is thought to breaks, it may also less DNA breaks et al, a in the highly sperm DNA a DNA may with He then on the of the SCSA to predict He that the SCSA not an absolute cutoff for and that its with pregnancy outcomes on the of or of the SCSA was also the important point we do not the of SCSA and TUNEL damaged sperm the are that ejaculated semen samples that a high proportion of sperm with damaged DNA SCSA have a for a successful pregnancy. The are that are a part of most SCSA clinical reports have the is important to that a DFI value above 30% not preclude a pregnancy. that the male is into a group of that a time to a more IVF/ICSI increased of spontaneous or no SCSA have been to be independent from semen analysis to important and clinically information not obtained by any other current SCSA conditions, 30% of the sperm have DNA damage to the 30% the other 70% may have DNA damage to and DNA (Evenson et al, and 2005). This is in with miscarriage the highest for ART in patients DFI was et al, et al, 2005). Dr Schlegel (2005) pointed out that of couples with spontaneous have DFI values (Carrell et al 2003). is also but was a point of for a of that the of fertilization have a on the predictive value of the SCSA test SCSA results from a study of couples using cannot be to predict the outcomes of In a study at of male factor infertility with Dr the SCSA for couples by natural the was greater for a successful pregnancy the DFI was (Evenson et al, et al, 2000). No couple when the DFI in that was However, some who had a 3 of DFI in months a in months occur with a DFI of but a time to Thus, there was a of a success but an absolute at of the were from the SCSA data. this was published it is from recent by (Evenson and the a of studies the relationship between sperm DNA fragmentation and pregnancy outcome using in and indicated that the patients were and times more likely to a the DFI was IVF was couples were and times more likely to become their DFI was of studies using IVF showed a where patients were and times more likely to a the was Payne et al (2005) that that of have The was some time since sperm DNA integrity is only one factor in the of a pregnancy. The has been and to poor DNA The SCSA and fertility. to this fact, for SCSA are only for the threshold of DFI and high DNA followed by a of the of patients above that threshold that or do become using This is in to the of the in which of and for testing where there are absolute outcomes (ie, a is present, and it is In to when the SCSA the for a pregnancy with a DFI are but there are for pregnancy to some may have a that could be as the test having predictive However, when the is to a threshold for which is the for the SCSA the of a pregnancy for a threshold of DFI are up to greater than when DFI is and The value of this same was in a recent by et al using the TUNEL of analysis with or as and of values for further significant between and sperm were for the pregnancy vs Thus, the probability for the female of by IVF is almost as or 50% at the same the of sperm in the is low Payne et al as well as other have 2 studies showing no pregnancies when the SCSA DFI only in the that the et al in Dr SCSA measures in Evenson showed that pregnancies by IVF/ICSI were obtained with DFI values In the case of et al Payne et al (2005) that no pregnancies Thus, I to further on these The first clinical study of SCSA in ART was at the of IVF of Dr on couples et al, 2000). The SCSA were done in Dr and the sent to Dr The showed that no pregnancies of the sperm showed DNA (DNA Dr current IVF at the of success for pregnancy with DFI even when has the SCSA and in Thus, a question this to the of the couples involved or to the ART involved in the The same question can also be of the Larson-Cook et al study. Payne et al (2005) that the DFI 30%, they et al, low of development and no This is not In et al reported that the pregnancy with a DFI was less than the couples having a DFI was a pregnancy with greater than Payne et al (2005) a we that the interpretation of the SCSA is based on the that when it to fragmented DNA and when it to is possible that the between this and DNA is more than this and that results are the first that with not a poor with patients undergoing In there are of on the of and the is that with or a of the that a from to At least of couple infertility is to a male and sperm DNA fragmentation is one important to this is a factor in their However, they be that a high DFI not preclude a normal pregnancy and that they should to the of using sperm their DFI is However, after with a of cases where couples have up to a cycles with no and that the DFI is in the this information should the to a different DNA fragmentation test is and relatively and information that will the the most to pregnancy. This the the of sperm DNA integrity in the field of male factor This is also by the of and on this published recently and Evenson and Schlegel and and et al, et al, 2005). However, it is that of DNA tests into clinical some Schlegel and 2005). The first is the of different tests that measure different aspects of DNA damage. by Alvarez and and Schlegel and DNA integrity tests in what they TUNEL and DNA et al, and in situ nick translation et al, measure breaks in the The SCSA (Evenson et al, Schlegel and and SCD or et al, measure to denaturation. In addition to in SCSA, other as et al, and et al, have also been to sperm chromatin and and have in been to other et al, et al, 2005). However, it would be to expect that they are in all of DNA damage. Thus, we should be in results of one to another we do not yet when and why these tests give different of of the of sperm DNA damage is another test is to a (eg, a to we cannot tests like SCSA in a there is yet no of sperm DNA and cases of DNA damage are et al, The clinical studies have been by the of patients with conditions, from female factors et al, to IVF failure et al, to male infertility et al, 2003). is no on the patient in these tests would be most we do not what DNA Several have been including damage et al, apoptosis et al, abnormal DNA et al, and in paternal DNA repair and et al, 2005). SCSA and other tests like TUNEL are in these of these tests is and not Another of the SCSA is the In (Evenson and SCSA is a test with repeat However, et al (2005) showed that the is to in This of studies from different more This is to in the of with the SCSA in studies et al, 2005). these the SCSA has been as an to predict ART IVF/ICSI. Unfortunately, this clinical has yet to be by by Dr the the hypothesis that high DNA fragmentation low is in studies of natural and (Evenson et al, et al, 2000; et al, Evenson and 2005). However, the is less on well SCSA IVF in cases et al, 2004). The by Dr Evenson in (Evenson and different studies of SCSA and IVF/ICSI. The analysis showed a for of pregnancy with DFI vs DFI but the result of and not some recent studies that found results et al, Payne et al, et al, 2005). all and to that have SCSA to IVF/ICSI cases clinical pregnancy as an and DFI of 27% or 30%, we found that a majority of but not found that high DFI with pregnancy They on whether SCSA fertilization a is that many studies have only been presented at and we to the of of SCSA, as that by et al The that a high DFI should a couple from or IVF et al, to be by further The value of DNA integrity testing is when one that the male is a is the of a genetic to the In to the measures of and that the DNA tests the content of the the role of the DNA integrity testing become the more However, we believe that current of DNA fragmentation yet the clinical of these tests.