Comparison of different hypo-osmotic swelling solutions to select viable immotile spermatozoa for potential use in intracytoplasmic sperm injection

Human Reproduction Update 1997, Vol. 3, No. 3 pp. 195 203 European Society for Human Reproduction and Embryology Comparison of different hypo-osmotic swelling solutions to select viable immotile spermatozoa for potential use in intracytoplasmic sperm injection G.Verheyen 1, H.Joris, K.Crits, Z.Nagy, H.Tournaye and A.Van Steirteghem Centre for Reproductive Medicine, University Hospital, Dutch-speaking Brussels Free University (Vrije Universiteit Brussel), Laarbeeklaan 101, B-1090 Brussel, Belgium TABLE OF CONTENTS Introduction 195 Materials and methods 196 Results 198 Discussion 200 References 202 The hypo-osmotic swelling test, originally developed as a diagnostic sperm test, is used to discriminate viable from non-viable spermatozoa for intracytoplasmic sperm injection (ICSI) in cases of complete asthenozoospermia. In the present study, three hypoosmotic solutions were compared, i.e. (A) Jeyendran solution containing sodium citrate and fructose; (B) a mixture of 50% culture medium and 50% milli-q water; and (C) milli-q water. While both the percentage of swelling and vitality assessed by eosin Y remained unchanged after 5 30 min of sperm exposure to solutions A and B, incubation in water for only 5 min was in itself detrimental. Ten frozen thawed donor samples and 10 asthenozoospermic patient samples were exposed to the three solutions for given periods of time. Percentages of swelling and the results of the eosin Y test were well correlated for solutions B and C, but only weakly correlated for solution A. Percentage viability was further assessed by eosin Y and motility of spermatozoa after 2 h and 24 h exposure to the three solutions was compared with unexposed control spermatozoa. While a significant decrease in both parameters was observed for all three solutions in comparison with the control, sperm quality was significantly higher after exposure to solution B than after exposure to solutions A and C. It may be concluded that solution B (composed of 50% culture medium and 50% water) is to be preferred for the selection of viable immotile spermatozoa for ICSI. Key words: eosin Y/hypo-osmotic swelling/icsi/ immotile spermatozoa/viability Introduction Intracytoplasmic sperm injection (ICSI) is the most successful procedure for assisted fertilization in cases of severe male-factor infertility and unexplained fertilization failure after conventional in-vitro fertilization (IVF) (Palermo et al., 1992, 1993; Van Steirteghem et al., 1993a,b). High fertilization and pregnancy rates are obtained with ejaculated, epididymal and testicular spermatozoa (Tournaye et al., 1994, 1996a; Nagy et al., 1995a; Silber et al., 1995; Kahraman et al., 1996; Tsirigotis et al., 1996). Although the success of ICSI is generally not affected by the quality of the basic sperm parameters (Nagy et al., 1995b; Oehninger et al., 1995), as long as spermatozoa of almost normal shape can be injected (Mansour et al., 1995), complete asthenozoospermia, where only immotile ejaculated spermatozoa are available for injection, lowers the fertilization rate considerably, reflecting the reduced proportion of vital spermatozoa in the final sperm suspension (Nagy et al., 1995b). This observation seems self-evident as long as motility is used as the sole marker for viability and no distinction is made between immotile live and immotile dead spermatozoa. Total immotility may be caused by necrozoospermia, ultrastructural defects or metabolic immaturity of the spermatozoa. In necrozoospermia, all spermatozoa in the ejaculate are non-viable, and testicular sperm extraction (TESE) may be successful in retrieving viable motile spermatozoa (Tournaye et al., 1996b). In case of ultrastructural defects, however, viable but immotile spermatozoa may be present in the ejaculate and, 1 To whom correspondence should be addressed. Tel: +32 2 477 5050; Fax: +32 2 477 5060

196 G.Verheyen et al. since many ultrastructural defects have a genetic origin (Afzelius, 1981), it is unlikely that motile spermatozoa will be found in the testis. Absence of motile spermatozoa is also occasionally observed in epididymal sperm aspirations and in testicular biopsy specimens in cases of obstructive and non-obstructive azoospermia. Viability tests based on dye exclusion (eosin Y, Hoechst 3358) are very reliable for diagnostic purposes, but the possible toxic effects of these stains when they are used to select viable but immotile spermatozoa for ICSI are unknown. The hypo-osmotic swelling test (HOST) was developed by Jeyendran et al. (1984) in order to evaluate the functional integrity of the sperm membrane. Live spermatozoa with normal membrane function show swelling of the tail due to water influx when exposed to hypo-osmotic conditions. The clinical predictive value of this functional test, however, is still being debated (Avery et al., 1990; Liu and Baker, 1992). Nevertheless, the HOST might be a less invasive procedure than supravital staining by which to select viable spermatozoa from an immotile sperm population. Its use in distinguishing viable from non-viable immotile spermatozoa for ICSI was first proposed by Desmet et al. (1994) who obtained a 30% fertilization rate with the injection of selected spermatozoa into 1 day old oocytes which had failed to fertilize after conventional IVF. Casper et al. (1996) applied the procedure with relative success in eight clinical ICSI cycles. Unfortunately, that study was not carried out under controlled conditions. No information is available on the possible detrimental effects of the non-physiological solution, composed of sodium citrate, fructose and water, developed to carry out the HOST diagnosis, so that questions may be raised about the use of this procedure to select spermatozoa for ICSI in cases of complete asthenozoospermia. In the following study, the effects of different exposure periods to three different hypo-osmotic solutions was investigated in order to define the appropriate incubation period for each of the solutions. Longer term effects on vitality and motility of exposed spermatozoa were also studied, with regard to both frozen thawed donor semen samples and asthenozoospermic patient samples, in order to define the least toxic procedure before introducing the test into clinical practice. Materials and methods Semen samples Ten frozen thawed donor semen samples and 10 fresh or frozen thawed asthenozoospermic patient samples were used for the study. Sperm quality was evaluated on both the fresh and the cryopreserved specimens immediately after thawing. Parameters assessed were concentration, percentage motility (A+B+C) and vitality by eosin Y, according to the World Health Organization criteria (WHO, 1992). Before applying the hypo-osmotic swelling tests, sperm preparation was carried out by two-layer Percoll gradient centrifugation (95%, 47.5%) at 300 g for 20 min. The 95% fraction was aspirated and washed twice with Earle s medium followed by centrifugation at 500 g for 5 min. The final pellet was resuspended in 250 µl Earle s medium and the above sperm parameters were reassessed. Hypo-osmotic solutions and eosin Y test Three hypo-osmotic solutions were used in this study: (A) Jeyendran solution (Jeyendran et al., 1984); (B) a solution composed of isotonic Earle s medium and milli-q water (50:50 v/v); and (C) milli-q water alone. Solution A (Jeyendran solution) was prepared by mixing 7.35 g sodium citrate (Na 3 C 6 H 5 O 7.2H 2 O) and 13.51 g fructose in 1 l of milli-q water. The osmolality measured (Osmomat 030 Gonotec; Van Hopplynus, Brussels, Belgium) was 155 mosmol/kg. Solution B was composed of half the volume of milli-q water, half the volume of Earle s medium containing 0.48% (v/v) HEPES and 1% human serum albumin (osmolality 285 mosmol/kg), resulting in a final osmolality of 139 mosmol/kg. The three solutions were sterilized by passing through a 0.22 µm filter before use. Spermatozoa prepared by Percoll fractionation were mixed with each of the hypo-osmotic solutions in a proportion of 1:10 (v/v) and incubated at a temperature of 37 C for the time period defined below. A total of 100 spermatozoa were evaluated as swollen or non-swollen, according to the criteria described by Jeyendran et al. (1984). Only clearly swollen cells of different types were considered live. Spermatozoa with a straight or a blunt tail were both considered to be non-swollen. The eosin Y test was carried out by mixing one drop of sperm suspension with one drop of 0.5% eosin Y (Sigma, Poole, Dorset, UK) solution. In all, 100 spermatozoa were counted as stained (dead) or unstained (viable). Experimental procedure Before comparing efficiency of the three solutions, the appropriate incubation time was defined for each of them; this is the shortest exposure time necessary to induce the maximum proportion of swollen spermatozoa. From one donor, five frozen thawed sperm samples prepared by Percoll fractionation were exposed to each of the three solutions A, B and C (20 µl spermatozoa:200 µl solution) at 37 C for 5, 10, 15, 20, 25 and 30 min. At each time interval,

Comparison of hypo-osmotic swelling solutions to select viable spermatozoa 197 Figure 1. Experimental flow chart. 5 µl of the suspension was removed and mixed with 5 µl of eosin Y for assessment of the percentage of dye exclusion and the percentage swelling at 400 magnification under a phase-contrast microscope. Appropriate incubation times were defined as 10 15 min for solutions A and B and 10 s for solution C. In order to determine the specificity and sensitivity of the swelling and viability tests, 100 individual spermatozoa from each of seven semen samples were incubated with solutions A or B for 5 min. After exposure to eosin Y, swelling and staining were evaluated and compared. Survival of spermatozoa after exposure to the three hypo-osmotic solutions was then assessed in 20 donor and asthenozoospermic patient samples. For each sample, 50 µl of the final Percoll fraction was diluted in 500 µl of each of the hypo-osmotic solutions (A, B, C) and incubated at 37 C, for 10 15 min for solutions A and B, and for 10 s for solution C. The percentage of swelling was evaluated under a phase-contrast microscope at 400 magnification. After incubation, the hypo-osmotic solution was washed out with Earle s medium in order to restore the physiological environment. After centrifugation at 800 g for 5 min, the pellet (250 µl) was resuspended in Earle s medium and incubated for 2 or 24 h; viability (eosin Y) and motility were then assessed. An untreated sperm fraction from each donor or patient was incubated for the assessment of motility and viability by eosin Y at 0, 2 and 24 h. At each time point one droplet of each of the suspensions was used for eosin Y and motility testing. The flow chart of the experiment is presented in Figure 1. Statistical analysis The viability and motility of spermatozoa exposed to eosin Y and to the three different solutions were compared two-by-two by the paired t-test. Spearman rank correlation tests were used to compare the different viability tests. Table I. Percentage of swollen and viable spermatozoa following incubation for different time periods in three hypo-osmotic solutions (n = 5) 5 min 10 min 15 min 20 min 25 min 30 min Solution A HOST 86 86 87 85 86 85 eosin Y 67 66 67 69 67 66 Solution B HOST 87 89 87 87 86 85 eosin Y 67 68 68 66 68 68 Solution C HOST 86 88 84 87 88 87 eosin Y 6 1 0 0 0 0 HOST = % swollen spermatozoa. Eosin Y = % spermatozoa exclusing eosin Y.

198 G.Verheyen et al. Results Defining the appropriate incubation time for each solution Five frozen thawed semen specimens of one donor were analysed before and after Percoll gradient centrifugation. The washed 95% Percoll fraction showed a mean concentration of 8 10 6 /ml, 25% type A, 23% type B, 13% type C motility, and 73% viability (eosin Y). The percentages of tail swelling over time are shown in Table I for the different hypo-osmotic solutions, together with the results of the eosin Y test which was carried out on the swollen spermatozoa. The maximum proportion of swollen/ unswollen spermatozoa appeared to be reached by 5 min incubation and there was no further increase observed in the proportion of swollen spermatozoa between 5 30 min of incubation, indicating that prolonged exposure to the solutions, as described by Jeyendran et al. (1984), is not required. Incubation in water (solution C) was detrimental to the spermatozoa, as shown by the absence of viable cells following the eosin Y test after 10 min of incubation, while viability was stable when the spermatozoa were exposed to solutions A and B for 5 30 min. Based on the results of this test, in further experiments spermatozoa were exposed to water for a maximum of 30 s, while incubation in solutions A and B was performed for 10 15 min. In this preliminary experiment, a remarkable difference was observed between the results of the swelling test and the eosin Y test. For solutions A and B, the percentage of swollen cells (HOST) was ~20% higher than the percentage of viable cells not stained by eosin Y. This implies that ~20% of the swollen, so-called live spermatozoa had taken up the dye after swelling. Assessment of specificity and sensitivity of the swelling and viability tests The results of the evaluation of 700 individual spermatozoa (seven semen samples) incubated for 5 min in solution A or B and afterwards exposed to eosin Y are presented in Table II. Although the specificity of the swelling test was high (>98% for solutions A and B), implying that a low rate of non swollen cells were unstained (low false-positive rate), the sensitivity was poor, as a large number of swollen cells were stained by eosin Y (high false-negative rate). Since interference between the tests, therefore, cannot be excluded, the combined viability test was not applied in the subsequent experiments. Table II. Specificity of the swelling test (solutions A or B) and the eosin Y test on the same group of spermatozoa (n = 700). Values are expressed as no. spermatozoa Eosin Y dead live Solution A HOST dead 75 (a) 3 (b) live 122 (c) 500 (d) Solution B HOST dead 76 (a) 9 (b) live 54 (c) 561 (d) Solution A = Sensitivity (a/a+c) = 38.1%; specificity (d/b+d) = 99.4%. Solution B = Sensitivity = 58.5%; specificity: 98.4%. Comparison of the three swelling tests and the eosin Y test The mean quality of the 10 frozen thawed donor and 10 patient samples before and after Percoll gradient centrifugation is presented in Table III. While the donor samples possessed good motility, total motility (A+B+C) of the patient samples (whether fresh or frozen) was severely impaired. For both types of samples, the quality of the spermatozoa was only slightly improved by Percoll gradient centrifugation. Sperm viability was comparable for patient and donor samples. Table III. Mean quality of donor and asthenozoospermic patient samples before (immediately after thawing) and after Percoll gradient centrifugation to remove cryporotectant Concentration Percentage motility Percentage viability ( 10 6 /ml) type A type B type C (eosin Y) Donor samples (n = 10) before Percoll 78 17 11 9 51 SEM ± 25 ± 3 ± 2 ± 2 ± 3 after Percoll 36 17 12 12 57 SEM ± 10 ± 3 ± 2 ± 1 ± 3 Patient samples (n = 10) before Percoll 55 2 4 5 42 SEM ± 17 ± 1 ± 1 ± 2 ± 7 after Percoll 20 3 6 6 49 SEM ± 8 ± 1 ± 2 ± 2 ± 6

Comparison of hypo-osmotic swelling solutions to select viable spermatozoa 199 Figure 2. Percentage viability as assessed by three hypo-osmotic swelling tests on donor samples (n = 10) and asthenozoospermic patient samples (n = 10), compared with percentage viability of a control sperm suspension from each donor or patient, assessed by eosin Y exclusion. Table IV. Spearman rank correlation tests between the results of the eosin Y test and the results of each of the different hypo-osmotic swelling tests (HOST; n = 20) Eosin Y HOST solution A solution B Solution A 0.40 a (P = 0.08) Solution B 0.58 0.75 (P = 0.007) (P <0.001) Solution C 0.56 0.50 0.71 (P = 0.01) (P = 0.026) (P = 0.001) a Correlation coefficient. Table V. Paired t-test between the results of the eosin Y test and the results of each of the different hypo-osmotic swelling tests (HOST; n = 20) Eosin Y HOST solution A solution B Solution A P = 0.006 Solution B P <0.001 P = 0.102 Solution C P <0.001 P = 0.066 P = 0.288 Each of the prepared sperm specimens (n = 20) was exposed to the three hypo-osmotic solutions for an incubation time as previously defined. The results of these hypo-osmotic swelling tests were compared with each other and with the results of the viability test using eosin Y, which was carried out in parallel on a control sperm fraction, independently of the swelling test (Figure 2). For the donor specimens, mean percentage swelling was 5% (solution A) to 10% (solution C) higher than the percentage viability, depending on which hypo-osmotic solution was used. Analysing the data per donor, however, revealed that the lower viability could be ascribed entirely to three samples showing 10 30% lower eosin Y exclusion than hypo-osmotic swelling. A greater difference (~20% of the mean values) between the results of the swelling and the eosin Y tests was prevalent among the patient samples: the percentage of swelling was consistently higher than the percentage of cells unstained by eosin Y in six out of 10 samples, which resulted in significant differences between the dye exclusion and the swelling test for each of the solutions (paired t-test). The results of the Spearman rank correlation tests and the paired t-tests between the dye exclusion test and each of the swelling tests on the 20 samples together are presented in Tables IV and V respectively. Although a reasonable correlation was observed between the percentages of swelling in solution B or C and the percentage dye exclusion, highly significant differences were found in the paired t-test. The different hypo-osmotic solutions, however, showed a good correlation and no differences in the paired t-test. Survival of spermatozoa after exposure to the three hypo-osmotic solutions Vitality The motility and viability (eosin Y) of the spermatozoa were assessed 2 and 24 h after exposure to, and subsequent removal of, the hypo-osmotic solutions. A sperm fraction not exposed to hypo-osmotic solution was used as control at the different time points. The change in viability of spermatozoa exposed to the different solutions is shown in Figure 3. In the untreated control group, viability decreased significantly by ~10% after 2 h of incubation, both for donor (P <0.001) and for patient samples (P = 0.002). A further slight decrease was observed after 24 h of incubation. For spermatozoa exposed to the hypo-osmotic solutions, a greater loss of viability was seen than in the control group. Exposure for 10 s to water (solution C) caused dramatic cell death (Figure 3). Spermatozoa treated with solutions A and B also showed a significantly lower viability than the control at each time point, both for donor and patient samples (Figure 3). The only point at which no

200 G.Verheyen et al. Figure 3. Percentage exclusion of eosin Y of spermatozoa exposed to hypo-osmotic solutions (A, B, C) in comparison with unexposed spermatozoa after 2 and 24 h of incubation, for donor samples (n = 10) and asthenozoospermic patients samples (n = 10). Figure 4. Percentage motility of spermatozoa exposed to hypo-osmotic solutions (A, B, C) in comparison with unexposed spermatozoa after 2 and 24 h of incubation, for samples from donors (n = 10) and asthenozoospermic patients (n = 10). difference from the control fraction was observed was the 2 h incubation of donor spermatozoa after exposure to solution B. While spermatozoa exposed to solutions A or B both survived significantly better (P <0.001) than those exposed to solution C, exposure to solution B led to a higher survival rate (P <0.05) than exposure to solution A for the asthenozoospermic patient samples, but not for the donor samples. In-vitro incubation of spermatozoa following exposure to different hypo-osmotic solutions therefore adversely affected sperm viability, but this effect was somewhat less pronounced for solution B. Motility Compared with viability, the parameter of motility (A+B+C) was even more adversely affected by exposure to the hypo-osmotic solutions (Figure 4). For donor samples, 2 h (P = 0.002) and 24 h (P <0.001) incubation of control spermatozoa significantly reduced the proportion of motile spermatozoa. This was not observed for patient samples, where a low percentage of motile spermatozoa remained for at least 24 h. Spermatozoa exposed to any of the three solutions showed a sharp significant decrease in motility after 2 or 24 h of further incubation, in comparison with their respective controls. Short-time exposure to solution C resulted in almost complete loss of motility after 24 h of incubation. An important observation is that both donor and patient spermatozoa exposed to solution B maintained motility significantly better (P <0.05) after 2 and 24 h of incubation than spermatozoa exposed to solution A. Discussion In cases of complete asthenozoospermia, the fertilization rate after ICSI of randomly selected immotile spermatozoa is usually very low, especially with ejaculated spermatozoa (Nagy et al., 1995b; Casper et al., 1996; Tournaye et al., 1996b). This reduced success rate might be due to the injection of non-viable spermatozoa, which cannot be distinguished from viable spermatozoa with no motility. Application of the hypo-osmotic swelling test, however, seems a promising method of identifying live spermatozoa for ICSI.

Comparison of hypo-osmotic swelling solutions to select viable spermatozoa 201 Only viable sperm cells, with a chemically and physically intact membrane, undergo tail swelling due to water influx under hypo-osmotic conditions. Casper et al. (1996) obtained a higher fertilization rate after injecting spermatozoa selected by the HOST (43%) than after injecting randomly selected spermatozoa (26%). This fertilization rate exceeded that obtained following viability assessment (30%) by supravital staining, indicating the efficiency of this sperm selection procedure for ICSI. Nevertheless, a fertilization rate of 43% is still poor, in the light of mean values of >60% with ejaculated spermatozoa as reported by most centres. This observation raises questions about the reliability of the procedure itself, or of the application of the test in cases of complete asthenozoospermia. While several groups have investigated the correlation between the HOST and other sperm parameters, and the prognostic value of the HOST for IVF outcome, surprisingly few data are available in the literature on comparisons of the results of the HOST and the different dye-exclusion tests. Inter-assay correlations are frequently observed, with coefficients ranging from r = 0.47 (Liu et al., 1988) to r = 0.76 (Coetzee et al., 1989; Avery et al., 1990), but absolute data in percentages for both tests together are mostly lacking. Although the present study confirms a correlation between the HOST and the eosin Y test for solutions B and C, only a weak, not significant correlation with the Jeyendran solution (solution A) was observed, which agrees with results obtained by Jeyendran et al. (1984) themselves. Correlations between the eosin Y and the three hypoosmotic solutions, however, revealed highly significant differences in the paired t-test (Tables IV and V), which can be explained by the consistency in deviation between the HOST and the eosin Y tests, at least for asthenozoospermic patient samples. The differences in results between the two tests are hard to explain. The higher results obtained by the HOST cannot be ascribed to overestimation of positively reacting cells, as spermatozoa with a blunt tail tip were considered non-viable, and this reduced the risk of including false-positive results. Only clearly swollen cells were considered viable. Neither can the deviation be ascribed to differences underlying the two principal viability tests. While exclusion of eosin Y requires only structural integrity of the cell membrane, reaction to hypo-osmotic conditions is dependent on both structural and functional integrity of the membrane. In the light of this difference, one would expect a higher percentage viability with eosin Y than with the HOST, which runs counter to our observations. It might be important to mention here a phenomenon frequently observed in our hands. It happens repeatedly that capacitated spermatozoa exposed to eosin Y show an increasing, time-dependent cell death during the evaluation of the test. The higher susceptibility of capacitated spermatozoa than fresh spermatozoa to eosin Y might be explained by a change in the composition and probably the permeability of the cell membane after capacitation. The degree of the effect varies between samples. In order to avoid it, assessments of the viability tests were carried out within a minimal time period, but some degree of underestimation of the percentage of live cells with the eosin Y test cannot be excluded. A pronounced, significant difference between the in results of the two types of viability test was already apparent in the preliminary experiment in which the eosin Y test was applied to hypo-osmotically swollen spermatozoa, which yielded complete cell death in conjunction with solution C and 20% higher values by HOST for solutions A and B, indicating that 20% of swollen cells had taken up the dye. Combining the two tests seemed to result in loss of viability by a limited population of swollen spermatozoa, which occurred either immediately after swelling, or simply as a result of exposure of the hypo-osmotically-incubated cells to eosin Y, as mentioned above for capacitated spermatozoa. Swelling might have caused additional changes in membrane permeability resulting in an easier dye passage through the sperm membrane. It is anyway clear from these results that it is not correct to validate the results of the HOST by exposure of swollen cells to eosin Y. The eosin Y test cannot be considered to be the gold standard, since its results might be adversely affected by the previously applied HOST, so that interference between the tests cannot be excluded. While incubation in the three solutions resulted in comparable percentages of sperm swelling (Figure 2), long-term motility and vitality of these spermatozoa after restoration of an isotonic environment showed clear differences, which were most pronounced in the use of water. In 1991, Lomeo and Giambersio described the use of distilled water as hypo-osmotic solution for a diagnostic swelling test as reliable and more simple than the Jeyendran medium, which was later confirmed by Fuse et al. (1993). The present results, however, indicate that water, although resulting in a swelling rate comparable to that of other solutions, is not appropriate in discriminating viable from non-viable spermatozoa for later use in ICSI. The strong hypotonicity after combining water and spermatozoa at a ratio of 1:10 causes severe osmotic stress.

202 G.Verheyen et al. The large volume of water entering the cell results in lysis of the cell membrane and cell death, even after short-time incubation of only 10 s, as is clear from the results of Table I and Figures 3 and 4. This excludes further use of water as a hypo-osmotic solution by which to identify viable cells for ICSI. Exposure for 10 15 min either to the Jeyendran solution or to 50% medium/50% water had a significantly negative effect on motility and vitality after 2 h and 24 h of further incubation in an isotonic environment, when compared with the control group at the respective time points. The maintenance of motility was significantly better with solution B than with solution A, both after 2 h and 24 h of incubation, and for both donor and patient spermatozoa. Survival, however, was higher only for patient spermatozoa after exposure to solution B rather than to solution A at both time points. That the quality of spermatozoa incubated in solution B is consistently better than that incubated in solution A might be explained by the more appropriate composition of solution B. Solution B is made up of all the components required for optimal sperm incubation, while the Jeyendran medium contains only fructose and sodium citrate. Although it cannot be ignored that sperm survival and motility are affected by exposure to hypo-osmotic solutions, this does not necessarily imply a decreased functionality of such exposed spermatozoa after injection into the oocyte. The use of HOST-selected spermatozoa for ICSI is in any case probably not completely safe and the efficiency is apparently below that of untreated motile spermatozoa, as shown by the relatively low fertilization rate (43%) obtained by Casper et al. (1996). On the other hand, it must be said that the exposure time to the hypo-osmotic solution can be considerably reduced. Incubation for 1 h in the hypo-osmotic solution, as advised for the diagnostic test by Jeyendran et al. (1984), and even for 10 15 min as performed in the present experiment, seemed far too long. Sperm swelling remained unchanged after >5 min of exposure to any of the hypo-osmotic solutions. Furthermore, aspiration of single spermatozoa with the injection pipette and expulsion into a small droplet containing the hypo-osmotic solution resulted in swelling after only 10 s, after which the sperm cell could be removed and replaced in a fresh droplet of isotonic medium to restore the physiological condition (unpublished data). The percentage swelling obtained with the manipulation of individual spermatozoa (three samples; 50 spermatozoa per sample) was similar to the percentage swelling observed with the conventional incubation procedure. The negative effects on long-term survival and motility observed in the present experiments might be reduced by this short-time exposure (10 s), but this needs to be further investigated. In a preclinical study, in-vitro-matured germinal-vesicle-stage oocytes could be injected with spermatozoa selected by either the Jeyendran solution or the solution composed of 50% medium and 50% water, in order to check if there is any benefit either in terms of fertilization and cleavage rate. Nevertheless, it may be concluded from these data that the solution composed of 50% isotonic medium and 50% distilled water is preferable for the selection of viable spermatozoa for ICSI in cases of complete asthenozoospermia. It is, however, not clear whether all problems of complete asthenozoospermia might be overcome by this method of selecting viable spermatozoa. When, on the day of ICSI, the semen sample of an occasional patient shows complete immotility, HOST selection of viable spermatozoa may be successful and TESE may be avoided. In cases of complete asthenozoospermia due to ultrastructural defects, however, it is far from certain that the use of the HOST will result in normal fertilization and cleavage. Centriolar defects are more frequently observed in immotile than in motile spermatozoa (Sathananthan, 1994). If fertilization occurs after injection of an immotile spermatozoon, a higher incidence of mitotic spindle defects may result in impaired, retarded or arrested embryo development (Sathananthan et al., 1996; Van Blerkom, 1996). This hypothesis needs further investigation. Acknowledgements We wish to thank Mr Frank Winter of the Language Education Centre for correcting the manuscript. This work was supported by a grant from the Belgian Fund for Medical Research. References Afzelius, B.A. (1981) Genetical and ultrastructural aspects of the immotile-cilia syndrome. Am. J. Hum. Genet., 33, 852 864. Avery, S., Bolton, V.N. and Mason B.A. (1990) An evaluation of the hypo-osmotic sperm swelling test as a predictor of fertilizing capacity in vitro. Int. J. Androl., 13, 93 99. Casper, R.F., Meriano, J.S., Jarvi, K.A. et al. (1996) The hypo-osmotic swelling test for selection of viable sperm for intracytoplasmic sperm injection in men with complete asthenozoospermia. Fertil. Steril., 65, 972 976. Coetzee, K., Kruger, T.F., Menkveld, R. et al. (1989) Hypoosmotic swelling test in the prediction of male infertility. Arch. Androl., 23, 131 138. Desmet, B., Joris, H., Nagy, Z. et al. (1994) Selection of vital immotile spermatozoa for intracytoplasmic injection by the sypoosmotic swelling test. 10th Annual Meeting of ESHRE, June 1994. Hum. Reprod., 9 (Suppl.4 ) p. 24

Comparison of hypo-osmotic swelling solutions to select viable spermatozoa 203 Fuse, H., Ohta, S., Sakamoto, M. et al. (1993) Hypoosmotic swelling test with a medium of distilled water. Arch. Androl., 30, 111 116. Jeyendran, R.S., Van der Ven, H.H., Perez-Pelaez, M. et al. (1984) Development of an assay to assess the functional integrity of the human sperm membrane and its relationship to other semen characteristics. J. Reprod. Fertil., 70, 219 228. Kahraman, S., Ozgur, S., Alatas, C. et al. (1996) High implantation and pregnancy rates with testicular sperm extraction and intracytoplasmic sperm injection in obstructive and non-obstructive azoospermia. Hum. Reprod., 11, 673 676. Liu, D.Y., du Plessis, Y.P., Nayudu, P.L. et al. (1988) The use of in vitro fertilization to evaluate putative tests of human sperm function. Fertil. Steril., 49, 272 277. Liu, D.Y. and Baker, H.W.G. (1992) Tests of human sperm function and fertilization in vitro. Fertil. Steril., 8, 465 483. Lomeo, A.M. and Giambersio, A.M. (1991) Water test : a simple method to assess sperm-membrane integrity. Int. J. Androl., 14, 278 282. Mansour, R., Aboulghar, M., Serour, G. et al. (1995) The effect of sperm parameters on the outcome of intracytoplasmic sperm injection. Fertil. Steril., 64, 982 986. Nagy, Z., Liu, J., Janssenswillen, C. et al. (1995a) Using ejaculated, fresh and frozen thawed epididymal and testicular spermatozoa gives rise to comparable results after ICSI. Fertil. Steril., 63, 808 15. Nagy, Z., Liu, J., Joris, H. et al. (1995b) The result of intracytoplasmic sperm injection is not related to any of the three basic sperm parameters. Hum. Reprod., 10, 1123 1129. Oehninger, S., Veeck, L., Lanzendorf, S. et al. (1995) Intracytoplasmic sperm injection: achievement of high pregnancy rates in couples with severe male factor infertility is dependent primarily upon female and not male factors. Fertil. Steril., 64, 977 981. Palermo, G., Joris, H., Devroey, P. and Van Steirteghem, A.C. (1992) Pregnancies after intracytoplasmic sperm injection of single spermatozoa into an oocyte. Lancet, 340, 17 18. Palermo, G., Joris, H., Derde, M.-P. et al. (1993) Sperm characteristics and outcome of human assisted fertilization by subzonal insemination and intracytoplasmic sperm injection. Fertil. Steril., 59, 826 35. Sathananthan, A.H. (1994) Functional competence of abnormal spermatozoa. In Fishel, S. (ed.), Baillière s Clinical Obstetrics and Gynaecology Micromanipulation Techniques. Baillière Tindal, London, pp. 141 156. Sathanathan, A.H., Ratnam, S.S., Ng, S.C. et al. (1996) The sperm centriole: its inheritance, replication and perpetuation in early human embryos. Hum. Reprod., 11, 345 356. Silber, S., Devroey, P., Tournaye, H. and Van Steirteghem, A.C. (1995) Fertilization capacity of epididymal and testicular sperm using intracytoplasmic sperm injection (ICSI). Reprod. Fertil. Dev., 7, 281 292. Tournaye, H., Devroey, P., Liu, J. et al. (1994) Microsurgical epididymal sperm aspiration and intracytoplasmic sperm injection: a new effective appraoch to infertility as a result of congenital bilateral absence of the vas deferens. Fertil. Steril., 61, 1045 1051. Tournaye, H., Liu, J., Nagy, Z. et al. (1996a) Correlation between testicular histology and outcome after intracytoplasmic sperm injection using testicular spermatozoa. Hum.Reprod., 11, 127 132. Tournaye, H., Liu, J., Nagy, Z. et al. (1996b) The use of testicular sperm for intracytoplasmic sperm injection in patients with necrozoospermia. Fertil. Steril., 66, 331 334. Tsirigotis, M., Pelekanos, M., Beski, S. et al. (1996) Cumulative experience of percutaneous epididymal sperm aspiration (PESA) with intracytoplasmic sperm injection. J. Assist. Reprod. Genet., 13, 315 319. Van Blerkom, J. (1996) Sperm centrosome dysfunction: a possible new class of male factor infertility in the human. Mol. Hum. Reprod., 2, 349 354. Van Steirteghem, A.C., Liu, J., Joris, H. et al. (1993a) Higher success rate by intracytoplasmic sperm injection than subzonal insemination. Report of a second series of 300 consecutive treatment cycles. Hum. Reprod., 8, 1055 1060. Van Steirteghem, A.C., Nagy, Z, Joris, H. et al. (1993b) High fertilization and implantation rates after intracytoplasmic sperm injection. Hum. Reprod., 8, 1061 1066. World Health Organization (1992) WHO Laboratory Manual for the Examination of Human Semen and Sperm Cervical Mucus Interaction. Cambridge University Press, Cambridge, UK. Received on October 25,1996; accepted on February 14, 1997