Abstract. Introduction. RBMOnline - Vol 7. No Reproductive BioMedicine Online; on web 8 September 2003

Transcription

1 RBMOnline - Vol 7. No Reproductive BioMedicine Online; on web 8 September 2003 Article Effects of cryopreservation on testicular sperm nuclear DNA fragmentation and its relationship with assisted conception outcome following ICSI with testicular spermatozoa Dr Sheena Lewis is Reader in Obstetrics and Gynaecology at Queen s University of Belfast in Northern Ireland. She is also Scientific Director of the Andrology Laboratory at The Royal Maternity Hospital in Belfast. She has published more than 60 research papers in the field of male infertility and has been invited to speak around the world. She reviews for many of the major journals and funding bodies in her field. Her special interests are the influences of sperm nuclear and mitochondrial DNA on assisted reproductive outcome, cryopreservation and antioxidant status, comparisons of semen profiles and assault susceptibility of spermatozoa from fertile and infertile men. Dr Sheena Lewis MEM Thompson-Cree 1, Neil McClure 1, 2, Eilish T Donnelly 1, Kristine E Steele 1, Sheena EM Lewis 1,3 1 School of Medicine, Obstetrics and Gynaecology, Queen s University Belfast, Institute of Clinical Science, Grosvenor Road, Belfast BT12 6BJ, Northern Ireland, UK; 2 Regional Fertility Centre, Royal Jubilee Maternity Service, Grosvenor Road, Belfast BT12 6BA, Northern Ireland, UK 3 Correspondence: Tel: ; Fax: ; s.e.lewis@qub.ac.uk Abstract The objective of the study was to investigate the effects of ing on testicular sperm DNA fragmentation, fertilization rates and pregnancy rates following intracytoplasmic sperm injection with testicular spermatozoa (TESE). This ongoing prospective study included 88 couples attending for infertility treatment where the man presented with obstructive azoospermia at the Regional Fertility Centre, Belfast, UK. Patients were allocated to receive TESE treatment with fresh or freeze-thawed spermatozoa. Sperm aliquots were stored in liquid nitrogen at 196ºC following static phase vapour cooling or cooling at controlled rates using a programmable freezer. Samples were thawed at either room temperature or 37ºC. Sperm nuclear DNA; assessed by the alkaline Comet assay, was significantly damaged by slow freezing followed by fast thawing. Pregnancies were more likely to be achieved with spermatozoa displaying markedly less DNA damage. However, no differences were observed in the fertilization rates, the number of blastomeres or the cumulative embryo score between TESE cycles using either fresh or frozen thawed testicular spermatozoa. The pregnancy rates tended to be higher following fresh TESE cycles (30%) compared with TESE cycles using frozen thawed testicular spermatozoa (26%), although this difference did not reach statistical significance. It is concluded that cryopreservation of testicular spermatozoa may reduce pregnancy rates, although this will only be confirmed by a much larger multi-centre trial. Keywords: cryopreservation, DNA fragmentation, fertilization rates, pregnancy rates, testicular intracytoplasmic sperm injection Introduction Although there is little information about the effects of sperm freezing on genetic integrity, cryopreservation of ejaculated spermatozoa has been routine practice in assisted conception units for many years. It is an essential step in donor insemination programmes and also provides insurance against future fertility problems for cancer patients before initiation of radiotherapy, cytotoxic chemotherapy or surgical treatment. With the extension of intracytoplasmic sperm injection (ICSI) to include the use of testicular sperm (TESE) of varying maturities, even testicular spermatozoa are being frozen, thus, due to the invasive nature of this procedure, eliminating the need to perform multiple biopsies if subsequent treatment cycles are required. Although many apparently healthy babies have been born around the world after TESE, concerns remain about its longterm effects. ICSI completely circumvents natural physical barriers to fertilization, as well as facilitating the use of spermatozoa of varying degrees of immaturity with highly abnormal motility and morphology. Thus, the natural mechanisms that select against structurally defective or metabolically abnormal spermatozoa are overridden. This potential for the random injection of a spermatozoon with 449

2 abnormal DNA must increase the risk of endowing progeny with genetic disorders. Although there is a desire to advance therapeutic options for patients, it is possible that the drive to be the first to introduce new treatments and to push back the clinical boundaries of fertility has led to this high-risk strategy. As with ICSI, this procedure has become clinical practice so quickly that the potential hazards of using immature spermatozoa, together with the further trauma of cryopreservation, have not yet been fully assessed. In animal spermatozoa, where there is already higher quality morphology and motility, cryosurvival is measured by quantitative assessments of acrosomal and plasma membrane integrity and even tests of sperm penetration. Yet, with human spermatozoa, no better tests are performed than a cursory assessment of cryosurvival amounting to prodding a spermatozoon to see if it can still twitch, or testing its response to a simple hypo-osmotic swelling (HOS) test. In addition to the concerns about transmission of genetic disease, there has been considerable debate as to whether cryopreservation of testicular spermatozoa affects assisted conception success. Some groups have shown that although fertilization rates are similar using ed testicular spermatozoa, pregnancy and birth rates using such cryopreserved spermatozoa are substantially lower: although no mechanisms have been identified to explain this (Friedler et al., 1997; De Croo et al., 1998; Madgar et al., 1998). However, others have shown no differences in assisted conception success (Habermann et al., 2000; Kupker et al., 2000). Cryopreservation can be performed in a variety of ways, including plunge freezing at +25 C to 186 C for 10 s, or programmable freezing where the temperature of the spermatozoa is reduced much more gradually (1 25 C/min). Conversely, thawing may be performed rapidly (in a 37 C water bath) or more slowly (at room temperature). These different rates of freezing and thawing, and combinations of each, have major ramifications in terms of cryoinjury caused by intracellular ice formation (Watson, 1995). It has been shown that ejaculated spermatozoa from fertile men are less susceptible to cryoinjury than those from infertile men (Donnelly et al., 2001). It has also been shown that addition of seminal plasma during freezing protects sperm DNA from damage (Donnelly et al., 1999), and this may be particularly beneficial to spermatozoa from infertile men, since these spermatozoa have less antioxidant protection in their seminal plasma (Lewis et al., 1995). Testicular spermatozoa, by their immaturity and the complete absence of this protection, may be even more vulnerable. This study compares the effects of programmable rate with plunge freezing and thawing of spermatozoa at room temperature, 37 C, on testicular sperm DNA and the relationship with fertilization and pregnancy rates following TESE. Materials and methods Subjects Testicular biopsies were performed on 88 men with obstructive azoospermia attending the Regional Fertility Centre, Belfast. Patients with obstructive azoospermia were identified on the basis of two azoospermic semen analyses, normal FSH and LH concentrations, a normal karyotype and bilaterally normal vas deferens on examination Forty-six couples had treatment using fresh testicular spermatozoa. The remaining 42 couples underwent treatment using frozen thawed testicular spermatozoa. Informed written consent for participation was obtained, and the project was approved by the Queen s University Belfast Research and Ethics Committee. Testicular sperm retrieval and preparation (after Steele et al., 2000) Briefly, the spermatic cord was located and 10 ml of 0.5% bupivacaine injected around it. A testicular biopsy was performed by passing a 14G Trucut biopsy needle (Baxter Healthcare Ltd, Thetford, Norfolk, UK) into the testis after the skin over the site had been anaesthetized. The biopsy was transferred into Biggers, Whitten and Whittingham medium (BWW) (Biggers et al., 1971). Testicular spermatozoa were retrieved from the seminiferous tubules by milking the tubular contents with size 5 jeweller s forceps into BWW, under a dissecting microscope. The contents of the seminiferous tubules were then centrifuged at 110 g for 10 min to remove debris, after which the sperm pellet was resuspended in BWW. Cryopreservation of spermatozoa Sperm aliquots were placed into cryovials (Nalge Company, Rochester, NY, USA) and mixed 1:0.7 with Enhance Spermfreeze cryoprotectant (Conception Technologies, San Diego, CA, USA). The major components in this freezing medium were glycerol (15%), HEPES (18 mmol/l) and human serum albumin (HSA, 0.4%). No antibiotics were included. The mixture was left to equilibrate at room temperature for 10 min. Half of the samples from each patient were placed in a horizontal position, parallel to the surface of the liquid nitrogen (10 cm above the surface for 15 min) prior to being plunged into liquid nitrogen ( 196 C). The other half were placed into the chamber of a computerized slow stage freezer (Planer, Kryo-10 MRV, UK) before cooling according to the following programme: (i) cooling rate of 1 C/min from +20 to +5 C; (ii) 10 C/min from +5 to 80 C; (iii) 25 C/min from 80 to 130 C and (iv) plunged into liquid nitrogen where the samples were stored until required. All clinical cycles used spermatozoa that had been plunge frozen and thawed at room temperature. 450

3 Determination of DNA integrity by modified alkaline single cell gel electrophoresis (COMET) assay The following procedure (adapted from Hughes et al., 1997 and Donnelly et al., 1999) was performed under yellow light to prevent further induced DNA damage. Embedding of spermatozoa in agarose gel Fully frosted microscope slides (Richardsons Supply Co. Ltd, London, UK) were gently heated, covered with 100 μl of 0.5% normal melting point agarose in Ca 2+ and Mg 2+ free phosphate buffered saline (PBS; Sigma-Aldrich Company Ltd, Poole, Dorset, UK) at less than 45 C and immediately covered with a large (22 50 mm) coverslip. The slides were placed in a chilled metal tray and left at 4 C for at least 30 min to allow the agarose to solidify. The coverslips were then removed and spermatozoa in 10 μl BWW were mixed with 75 μl of 0.5% low melting point agarose at 37 C. This cell suspension was rapidly pipetted on top of the first agarose layer, covered with a coverslip and allowed to solidify at room temperature. Lysing of cells and decondensation of DNA The coverslips were removed and the slides immersed in a Coplin jar containing freshly prepared cold lysing solution (2.5 mol/l NaCl, 100 mmol/l Na 2 EDTA, 10 mmol/l Tris; ph 10, with 1% Triton X-100 (Sigma, UK) added just before use for 1 h at 4 C. Subsequently, slides were incubated for 30 min at 4 C with 10 mmol/l DTT followed by a 90 min incubation at 20 C with 4 mmol/l lithium diiodosalicyclate (LIS; Sigma, UK). Unwinding of DNA Slides were removed from the lysis solution, DTT and LIS and carefully drained of any remaining liquid. A horizontal gel electrophoresis tank was filled with fresh alkaline electrophoresis solution (300 mmol/l NaOH, 1 mmol/l EDTA, ph 13.0: Sigma, UK) at C. The slides were placed into this tank side by side, with the agarose end facing the anode, and covered with electrophoresis buffer to a level of approximately 0.25 cm above the slide surface. The slides were left in this high ph buffer for 20 min to allow DNA in the cells to unwind. Separation of DNA fragments by electrophoresis Electrophoresis was conducted for 10 min at 25 V (0.714 V/cm) adjusted to 300 ma by raising or lowering the buffer level in the tank. After electrophoresis, the slides were drained, placed on a tray and flooded with three changes of neutralization buffer (0.4 mol/l Tris; ph 7.5; Sigma) for 5 min each. This removed any remaining alkali and detergents that might interfere with ethidium bromide staining. The slides were then drained and stained with 50 μl of 20 μg/ml ethidium bromide (Sigma) and covered with a large coverslip. Image analysis The slides were viewed using a Nikon E600 epifluorescence microscope that was equipped with an excitation filter of nm from a 100 W mercury lamp and a barrier filter of 590 nm. Fifty images were captured and analysed by an image analysis system using the programme Komet 3.1 (Kinetic Imaging Ltd, Liverpool, UK). TESE procedure All testicular ICSI cycles were performed according to routine procedures. Ovulation induction was achieved with Metrodin HP (Serono, Welwyn Garden City, UK) following a long protocol of pituitary desensitization with buserelin acetate, Suprefact (Hoechst, Hounslow, UK) or Synarel (Syntex Pharmaceuticals Ltd, Maidenhead, UK). Human chorionic gonadotrophin (HCG; Pregnyl, Organon) was administered when there were at least four follicles with mean diameter of 17 mm, 36 h before oocyte retrieval. Mature, metaphase II oocytes obtained by vaginal ultrasound-guided aspiration were cultured in universal IVF medium, Medicult (Imperial Laboratories, Andover, UK) at 37 C, 5% CO 2 in air. The injection procedure has been described in detail before (Van Steirteghem et al., 1993). In brief, the injection pipette was filled with a small amount of polyvinylpyrrolidone (PVP) and a free, motile spermatozoon was aspirated head-first from 5 μl of universal IVF medium into the pipette. The single spermatozoon was then placed, without any cell debris, in another droplet of PVP and immobilized by breaking its tail. The spermatozoon was aspirated again into the injection pipette tail-first and injected into an oocyte. Fertilization was recorded h after injection if two pronuclei were detected. The embryos were transferred into the uterine cavity after an additional 24 h. Luteal support was provided by Cyclogest vaginal pessaries (400 mg; Shire Pharmaceuticals Ltd, Basingstoke, UK). Cumulative embryo score (CES) Two days after oocyte recovery, prior to embryo transfer, the embryos were graded (after Steer et al., 1992) as follows: grade 4, equal sized symmetrical blastomeres; grade 3, uneven blastomeres with <10% fragmentation; grade 2, (10 15%) blastomeric fragmentation; and grade 1 (>50%) blastomeric fragmentation or pronucleate single cell embryos. The morphological grade of the embryo was then multiplied by the number of blastomeres to produce a quality score for each embryo. The scores of all embryos transferred per patient were summed to obtain the CES, to be used as an index of the total embryonic contribution to pregnancy, regardless of the number of embryos transferred (Steer et al., 1992). Statistical analysis Data were analysed using SPSS 10 for windows ( The Student s t-test for independent samples was used to analyse differences in clinical outcomes between fresh and ed testicular spermatozoa. The Student s t-test for dependent samples was employed to determine differences in DNA damage in fresh and ed spermatozoa. A general linear model, using a two-way analysis of variance, with subjects as blocking 451

4 variables and method of cryopreservation as the variable of interest was also used. Standard tests were performed to check the validity of this method (normality of residuals, independence and equality of variants). A post-hoc test was also applied using Duncan s tests for homogeneous subgroups. Results Comparison of different ing techniques on sperm DNA fragmentation Sperm DNA was markedly damaged (+28%) by rapid freezing followed by slow thawing (Table 1). The test for homogeneous subsets drew a clear distinction between the group that was plunge frozen and then thawed at room temperature. In contrast, rapid freezing followed by rapid thawing, slow freezing followed by slow thawing or slow freezing followed by rapid thawing, did not induce DNA damage. Table 1. Effects of ing on sperm DNA fragmentation. The test for homogeneous subsets drew a clear distinction between the group that was plunge frozen and then thawed at room temperature. Group n DNA P-value fragmentation (group versus (%) fresh) Fresh ± 0.86 n/a Plunge ± 0.89 <0.001 at RT Plunge ± 0.78 NS at 37 C Programmed ± 0.91 NS at RT Programmed ± 0.73 NS at 37 C Relationship between DNA fragmentation and assisted conception success DNA fragmentation in spermatozoa that did not achieve a pregnancy was substantially higher (63%) than that of spermatozoa that resulted in pregnancies (Table 2). Effects of ing on fertilization rates The fertilization rate in the 46 cycles of TESE using fresh spermatozoa was 75%, compared with 70% in the 42 cycles of TESE using ed spermatozoa (Table 3). Table 2. Relationship between DNA fragmentation and assisted conception success. Successful Unsuccessful P-value cycles cycles n DNA ± 1.30 a ± 1.96 a <0.001 fragmentation (%) a Significantly different. NS = not significant; n/a = not applicable. Table 3. Effects of ing of spermatozoa on assisted conception success. Fresh Freeze thawed P-value Successful Unsuccessful Successful Unsuccessful (fresh versus outcome outcome outcome outcome ed) 452 Fertilization rate 0.75 ± ± ± ± 0.04 NS (n = 14) (n = 32) (n = 11) (n = 31) Blastomere no. at 5.5 ± ± ± ± 0.3 NS embryo transfer Cumulative embryo score 39.1 ± ± ± ± 2.5 NS (n = 11) (n = 27) (n = 11) (n = 23) Clinical pregnancy rate 30% 70% 26% 74% NS (n = 14) (n = 32) (n = 11) (n = 31) Take-home babies 6 boys; 2 girls; 6 boys; 3 girls; NS 3 sets of 1 set of twin boys twin girls

5 Effects of ing on embryo cleavage There was no significant difference in the number of blastomeres or the embryo quality as assessed by CES between embryos created by fresh or ed spermatozoa at embryo transfer (Table 3). Effects of ing on pregnancy rates and numbers of take-home babies Where fresh spermatozoa were used, a clinical pregnancy rate of 30% was achieved compared with a rate of 26% pregnancy when ed spermatozoa were used. There was no significant difference between these two groups (P = 0.42). Similarly, there was no significant difference in the small numbers of babies conceived with ed rather than fresh spermatozoa, or in their genders (Table 3). Discussion Since the practice of using testicular spermatozoa in ICSI often necessitates sperm storage, a number of studies have compared a range of outcome parameters to determine if cryopreservation reduces TESE success. However, there is considerable conflict in the literature, with some studies showing no differences in success rates (Gil Salom et al., 1996; Friedler et al., 1997; Madgar et al., 1998) and some showing distinct reductions after cryopreservation (Verheyen et al., 1997; De Croo et al., 1998). This discrepancy may be due to one of the major difficulties encountered in this study, namely lack of randomization. Some patients chose for personal reasons to opt for a fresh or frozen cycle. Other patients undergoing treatment with a fresh biopsy were doing so because a previous biopsy had been unsuitable for cryopreservation. Similarly, those having treatment cycles using frozen biopsies usually did so because their spermatozoa had been shown previously to withstand the trauma of freezing without obvious damage. This might suggest that, in these cases, the frozen spermatozoa from these individuals were superior to fresh spermatozoa from other men. Thus, any conclusions that success rates with fresh spermatozoa are better than those from ed spermatozoa must take these confounding factors into consideration. To overcome this problem, only first-time fresh or frozen samples that had been allocated to each group were included. However, by doing this, the potential numbers of couples in the 4-year study were halved, reducing them from 175 to 88. There was no relationship between cryopreservation of testicular spermatozoa and fertilization rates in TESE. These results are in conflict with those from De Croo et al. (1998), where fertilization rates were significantly lower with frozen thawed spermatozoa, but in agreement with Friedler et al. (1997) and Madgar et al. (1998), both of whom observed no differences in fertilization rates after freezing. A more important outcome parameter than fertilization is the clinical pregnancy rate. The present study has shown that pregnancy rates were slightly reduced (from 30% to 26%) with cryopreserved spermatozoa. This may be associated with a loss of sperm nuclear DNA integrity, since all the spermatozoa in the clinical arm of the study were plunge frozen and thawed at room temperature; a protocol that causes significant DNA damage (Table 2). Again, data reported here support those of Friedler et al. (1997), who found that although pregnancy rates fell markedly from 21 to 9%, the difference did not reach significance. Similarly, De Croo et al. (1998) showed a reduction in pregnancy rate from 38 to 26%, culminating in a lower live birth rate of only 8% with ed spermatozoa compared with 19% when fresh spermatozoa were used. This is further supported by a study from Macas et al. (1998), in which the pregnancy rates were 10% less with frozen spermatozoa. In those studies where no differences were found (such as Gil Salom et al., 1996 and Ben Yosef et al., 1999), the frozen cycles are not unbiased, since spermatozoa from these men will have been tested for cryosurvival before a treatment cycle and only those men whose spermatozoa survived will be offered this treatment. Of course, the most important outcome parameter of all these studies is the take home baby rate. Unfortunately, due to the low success rates in ART worldwide, this study suffers, like so many others, in having too few data to draw any conclusions on this point. However, in the second part of the study, by comparing the effects of different freezing protocols on DNA fragmentation in testicular spermatozoa, robust data show that undamaged paternal DNA is important for reproductive success. This is especially important in view of the recent study by Ramos et al. (2002), showing that there is a much higher level of DNA breaks in spermatozoa from men with obstructive azoospermia compared with donors. These data add to the belief that a larger study is warranted to assess these clinical implications properly. The study adds to the increasing body of evidence that fertilization failure, poor embryo cleavage and failure to implant are all associated with damaged sperm DNA. Unfortunately, even more serious consequences of DNA damage may only become apparent at a later stage of pregnancy, during fetal or childhood development, or even in adulthood. DNA breakages lead to chromosomal damage and most sperm-derived genetic damage occurs through chromosomal breakage, rather than chromosomal rearrangement, as in the oocyte (Cathcart et al., 1984). In addition, studies of genetic abnormalities arising from germ line mutations show a higher frequency of paternal rather than maternal origin (Vogel and Rathenberg, 1975). As well as lack of repair, this may be because there are more cell divisions in the male germ cell during spermatogenesis than oogenesis, allowing more opportunities for mutations to occur. Unlike somatic cells, where DNA repair and recombination are normal features, sperm DNA strand breakage inevitably results in DNA damage. This is because both testicular and ejaculated spermatozoa are post-spermiogenetic, a process that results in discarding of the spermatic cytoplasm, at which stage transcription ceases and the spermatozoon cannot undertake DNA repair. Hence, any damage that occurs after spermiogenesis will be irreversible until fertilization. However, even then, the oocyte has only the capacity to repair small amounts of sperm DNA damage (<8%) (Ahmadi and Ng, 1999). Thus, damage to sperm DNA has far reaching consequences and is strongly correlated with mutagenic events (Fraga et al., 1991), resulting in birth defects and an increased risk of both genetic diseases and cancer in offspring (Sorahan et al., 1997). 453

6 454 Work from Schatten s laboratory (Hewitson et al., 2002) shows that ICSI can result in abnormal nuclear remodelling during sperm decondensation, due to the presence of the sperm acrosome and perinuclear theca, structures normally removed at the oolemma during IVF; this is turn causes a delay in the onset of DNA synthesis. These unusual modifications raise their concerns that the ICSI procedure itself may result in chromatin damage during DNA decondensation. Like us, this causes them to highlight the need for a move rigorous assessment of methods of assisted reproduction prior to their global application. The use of a variety of cryopreservatives and different combinations of freezing and thawing rates have improved cell survival (reviewed by Watson, 2000). Propanediol and dimethylsulphoxide are commonly used for embryos, but methods perfected for the embryo do not necessarily suit the spermatozoon, as it is unique with its non-spherical shape and small cell volume to large surface area ratio. Further, testicular spermatozoa need protocols optimized for them, since structurally they are rather like hybrids between ejaculated spermatozoa and embryos in that their cytoplasm and surface area are less than those of embryos but still greater than those of ejaculated spermatozoa. Different regimes are required even for spermatozoa from different mammals. For example, some species are more susceptible to cold shock or chilling injury (Watson, 2000). Here, a cryoprotectant composed of glycerol (15%), the most widely used and successful preservatives for mature spermatozoa, was used. One major benefit of this commercially prepared medium over egg yolk based media is that it does not require vigorous washing to remove the yolk nor inhibit sperm motility by its stickiness. A major cause of all cryoinjury is the formation of intracellular ice (Muldrew and McGann, 1990). In this study, those spermatozoa most damaged by cryopreservation were those frozen rapidly and then thawed slowly. Rapid freezing is known to cause more cryoinjury than slow controlled freezing (Verheyen et al., 1997). This may be explained by the slower cooling allowing cytoplasmic water to exit the cell and freeze outside thus minimizing ice crystal formation within the cells. This damage to DNA is reflected in the only clinical data available from this protocol (Friedler et al., 1997) where the pregnancy rate was reduced by 12%. However, it was found that if rapid freezing was followed by rapid thawing, the damage could be minimized. This has been explained by Watson (1995) as the rapidly cooled cells (with the greatest likelihood of intracellular ice) being rescued by rapid rewarming, which limits the time for recrystallization to occur. Other studies suggest that gentle thawing at room temperature rather than 37 C appears to facilitate rehydration without membrane and chromatin damage. However, the results would suggest that this only protects the cell from the limited damage caused by gentle freezing. If the cell has already been substantially damaged by rapid freezing, slow thawing gives additional time for recrystallization to occur. Although testicular sperm cryopreservation is convenient and allows several cycles of treatment from one invasive biopsy, there is now a body of literature suggesting that it reduces the couple s chances of a pregnancy. The present study provides a mechanism by which this occurs showing that it may be due to sperm DNA damage induced by a commonly used protocol, for example in the study of Friedler et al. (1997). The clinical data from this study add to the ever-increasing number of similar small (n < 100 cycles) one-centre studies. The consensus of data available from these studies is that using ed testicular spermatozoa reduces the couple s chance of a pregnancy. However, all of these studies are of limited individual value because of the small numbers available to each group and the variation in protocols from laboratory to laboratory. Another problem is that, again because of small sample numbers, many of the differences do not reach statistical significance and on reading only the abstract, this is the take-home message. However, this can be most misleading, since trends can show marked reductions following freezing. An example of this is seen in the study of De Croo et al. (1998), where it was concluded that a pregnancy rate of 38% with fresh spermatozoa was not statistically different from one of 26% following freezing and therefore unimportant. Such a difference may well be of clinical importance. In order to draw conclusions of clinical value, it would be most useful to readdress this important issue on a multi-centre basis, where much larger numbers could be evaluated using identical protocols. Only this approach will provide data with sufficient statistical power to answer this dilemma. Acknowledgements Financial support from The Wellcome Trust project no /Z/99 is gratefully acknowledged. The authors also wish to acknowledge Mr Michael Stevenson for his assistance with the statistical analysis and the contributions of the embryologists in the Regional Fertility Centre. References Ahmadi A, Ng S-C 1999 Fertilizing ability of DNA-damaged spermatozoa. Journal of Experimental Zoology 284, Ben Yosef D, Yogev L, Hauser R et al Testicular sperm retrieval and cryopreservation prior to initiating ovarian stimulation as the first line approach in patients with nonobstructive azoospermia. Human Reproduction 14, Biggers JD, Whitten WK, Whittingham DG 1971 The culture of mouse embryos in vitro. In: Daniel JC (ed.) Methods in mammalian embryology. Freeman, San Francisco, pp Cathcart R, Schwiers E, Saul RL et al Thymine glycol and thymidine glycol in human and rat urine a possible assay for oxidative DNA damage. Proceedings of the National Academy of Sciences of the USA 81, De Croo I, Van der Elst J, Everaert K et al Fertilization, pregnancy and embryo implantation rates after ICSI with fresh or frozen thawed testicular spermatozoa. Human Reproduction 13, Donnelly ET, McClure N, Lewis SEM 1999 The effect of ascorbate and α-tocopherol supplementation in vitro on DNA integrity and hydrogen peroxide induced DNA damage in human spermatozoa. Mutagenesis 14, Donnelly ET, McClure N, Lewis SEM 2001 Cryopreservation of human semen and prepared sperm: effects on motility parameters and DNA integrity. Fertility and Sterility 76, Fraga CG, Motchnik PA, Shigenaga MK et al Ascorbic acid protects against endogenous oxidative DNA damage in human sperm. Proceedings of the National Academy of Sciences of the USA 88, Friedler S, Raziel A, Soffer Y et al Intracytoplasmic injection of fresh and cryopreserved testicular spermatozoa in patients with nonobstructive azoospermia a comparative study. Fertility and Sterility 68,

7 Gil Salom M, Romero J, Minguez Y et al Pregnancies after intracytoplasmic sperm injection with cryopreserved testicular spermatozoa. Human Reproduction 11, Habermann H, Seo R, Cieslak J et al In vitro fertilization outcomes after intracytoplasmic sperm injection with fresh or frozen thawed testicular spermatozoa. Fertility and Sterility 73, Hewitson L, Simerly CR, Schatten G 2002 Fate of sperm components during assisted reproduction: implications for infertility. Human Fertility 5, Hughes CM, Lewis SEM, McKelvey-Martin V et al Reproducibility of human sperm DNA measurements using a single cell gel electrophoresis assay. Mutation Research 374, Kupker W, Schlegel PN, Al-Hasani S 2000 Use of frozen thawed testicular sperm for intracytoplasmic sperm injection. Fertility and Sterility 73, Lewis SEM, Boyle PM, McKinney KA et al Total antioxidant capacity of seminal plasma is different in fertile and infertile men. Fertility and Sterility 64, Macas E, Imthurn B, Borsos M et al Impairment of the developmental potential of frozen thawed human zygotes obtained after intracytoplasmic sperm injection. Fertility and Sterility 69, Madgar I, Hourvitz A, Levron J et al Outcome of in vitro fertilization and intracytoplasmic injection of the epididymal and testicular sperm extracted from patients with obstructive and nonobstructive azoospermia. Fertility and Sterility 69, Muldrew K, McGann LE 1990 Mechanisms of intracellular ice formation. Biophysical Journal 57, Ramos L, Kleingeld P, Meuleman E et al Assessment of DNA fragmentation of spermatozoa that were surgically retrieved from men with obstructive azoospermia. Fertility and Sterility 77, Sorahan T, Prior P, Lancashire RJ et al Childhood cancer and paternal use of tobacco: deaths from 1971 to British Journal of Cancer 76, Steele EK, Kelly JD, Lewis SEM et al Testicular sperm extraction by Trucut needle and milking of seminiferous tubules: a technique with high yield and patient acceptability. Fertility and Sterility 74, Steer CV, Mills CL, Tan SL et al The cumulative embryo score: a predictive embryo scoring technique to select the optimal number of embryos to transfer in an in-vitro fertilization and embryo transfer programme. Human Reproduction 7, Van Steirteghem AC, Nagy P, Joris H et al High fertilization rates after intracytoplasmic sperm injection. Human Reproduction 8, Verheyen G, Nagy Z, Joris H et al Quality of frozen thawed testicular sperm and its preclinical use for intracytoplasmic sperm injection into vitro-matured germinal-vesicle stage oocytes. Fertility and Sterility 67, Vogel F, Rathenberg R 1975 Spontaneous mutation in man. Advances in Human Genetics 5, Watson PF 1995 Recent developments and concepts in the cryopreservation of spermatozoa and the assessment of their postthawing function. Reproduction Fertility and Development 7, Watson PF 2000 The causes of reduced fertility with cryopreserved semen. Animal Reproduction Science 60 61, Paper based on contribution presented at the First meeting of the Mediterranean Congress on Reproductive Medicine in Taormina, Sicily, November Received 16 April 2003; refereed 13 May 2003; accepted 17 July