FERTILITY AND STERILITY
Vol. 58, No.3, September 1992
Copyright © 1992 The American Fertility Society
Printed on acid·free paper in U.S.A.
Prediction of in vitro fertilization outcome by sperm penetration assay with TEST-yolk buffer preincubation*
Yigal Soffer, M.D. t Abraham Golan, M.D. Arie Herman, M.D.
Mordechai Pansky, M.D. Eliahu Caspi, M.D. Raphael Ron-EI, M.D.
Department of Obstetrics and Gynecology, Tel Aviv University, Sackler School of Medicine, Assaf Harafe Medical Center, Zerifin, Israel
Objective: To evaluate sperm penetration assay (SPA) value as a screening tool before in vitro fertilization (IVF). Design: Follow-up study comparing sperm variables and IVF outcome. Setting: Infertile couples in an academic research environment. Patients, Participants: Two hundred forty-one infertile couples scheduled for IVF. Interventions: Sperm penentrating assay with cold Tes-TRIS (TEST) Yolk buffer semen preincubation and IVF. Main Outcome Measures: Percent of egg penetration recoded into poor and good category (0% to 20%, 21 % to 100%) and compared with fertilization, embryo transfer, and pregnancy rate (PR) in IVF, as well as sperm count motility and morphology. Results: Sperrn penetrating assay predicted fertilization with a high negative (74%) and positive (82%) predictive rate and good specificity value (0.96). One of 31 patients in the poor SPA category (3%) fertilized >50% of eggs; no pregnancy occurred. In the good SPA category, 87 of 210 patients (41 %) fertilized 2':50% of eggs with 34.3 % PRo Sperm penetrating assay correlated better with fertilization rate than did sperm count and motility but not morphology. Conclusions: Sperm penetrating assay is a useful screening assay before IVF together with sperm morphology Fertil Steril 1992;58:556-62 Key Words: Sperm egg penetration, TEST -yolk buffer enhanced capacitation, in vitro fertilization, sperm motility, count, morphology
In the array of tests aimed at the evaluation of male fertilizing capability, sperm penetrating assay (SPA) stands out as a unique bioassay (1, 2). This assay directly addresses four of the most decisive physiological processes in fertilization mechanism: [1] the capacitation, [2] acrosome reaction, [3] fusion of sperm cell with the oolema, and [4] chromatin decondensation with head swelling in the egg cytoplasma. Attempts to assess the significance of this bioassay in the diagnosis of male infertility have
Received June 19, 1991; revised and accepted May 1, 1992. the 7th World Congress on IVF and Assisted Procreations, Paris, France, June 30 to July 3, 1991. t Reprint requests: Yigal Soffer, M.D., Infertility Clinic, Department of Obstetrics and Gynecology, Assaf Harafe Medical Center, Zerifin 70300, Israel.
* Presented at
556
Soffer et al. SPA predicts IVF outcome
been confounded by the multifactorial nature of the fertilization process and potential associated female factors (2-4). The advent of in vitro fertilization and embryo transfer (IVF -ET) has contributed in eliminating various male and female factors by bringing human spermatozoa and oocytes into direct physical contact. In a somewhat similar manner, the SPA is capable of directly detecting faults in sperm penetrating capability. Most (5-9) but not all studies (10) that evaluated SPA's capability to predict IVF outcome presented supportive evidence for the use of this bioassay as a reliable detector of a male factor, and as such, as a screening tool for IVF. However, the technical difficulties, lack of standardized assay protocols, and even differences in the measured variables (percent of penetration [2,4,6,9-11] verFertility and Sterility
sus penetration index [8]) have hindered the acceptance of this important assay for routine use in many IVF programs and infertility centers. This study presents, to our knowledge, the largest ever reported series in conjunction with IVF. We have undertaken first, to examine the predictive value of SPA under conditions that enhance sperm capacitation, as a reliable predictor of poor male fertilization ability in IVF and second, to evaluate quantitatively the correlation between SPA result and IVF outcome in terms of fertilized oocytes ET and pregnancy rates (PRs). MATERIALS AND METHODS
Two hundred forty-one couples participating in IVF program entered this study. The female partner's age was 31.92 ± 4.49 (range 21 to 41 years). All these women underwent a routine infertility work-up including menstrual cycle evaluation, hysterosalpingography, and eventually hysteroscopy and laparoscopy. Female diagnosis was as follows: infertility of unknown origin 53.5%, tubal factor 38.2%, endometriosis 3.1 %, and other causes (polycystic ovarian disease, ovarian failure, female antisperm isoantibodies, and uterine pathological conditions) 5.2%. All males underwent a clinical examination and repeated semen analysis (12), which was performed in a 10-tIm-deep counting chamber (Makler Chamber, Sefi Medical Instruments, Haifa, Israel). Sperm quality was assessed by a motility index. % of motile cells X grade of motility (0 to 4), total sperm count, sperm concentration X semen volume and normal sperm form percentage on air-dried eosin - nigrosin stained smears examined at high-power field (HPF) (X1,000). The normal forms included the spermatozoa with an oval form according to World Health Organization criteria (13) and those with a well-developed long acrosome region. Sperm penetrating assay was performed 3-6 weeks before IVF. Sperm Penetration Assay
Sperm penetrating assay procedure was performed as previously described (14) with slight modifications. Ejaculates were allowed to liquefy for a maximum of 10 to 30 minutes. A small aliquot (0.2 to 0.4 mL) was kept for analysis, and all the remaining volume immediately diluted according to the semen viscosity in one or two volumes of Tes-TRIS (TEST: Sigma Chemical Co., St. Louis, MO) yolk buffer and kept in vertical tubes at 4°C for 18 to 22 hours. At the end of incubation, supernatant was carefully reVol. 58, No.3, September 1992
moved. Then, in addition to the above protocol (14), a sperm layering and rise-up procedure was done as follows. The sedimented sperm was resuspended in 0.25 mL of Biggers, Whitten, and Whittingham (BWW) buffer (15) enriched with 1.75% wt/vol bovine serum albumin, fraction V (Sigma Chemical Co.) overlaid with 1 mL of same medium in 30°C inclinated tubes, and sperm was allowed to rise up for 1 hour in a CO 2 incubator (at 37°C under 5% CO 2 atmosphere and saturated humidity). The supernatant containing motile spermatozoa was centrifuged, and sperm pellet was resuspended in a small amount of BWW and finally adjusted to a concentration of 5 to 10 X 106 motile cells/mL. A O.l-mL drop of this sperm suspension was prepared and kept in the CO 2 incubator until addition of eggs. Female golden hamsters were superovulated as already described (14). In this study, the females were 7 to 9 weeks old and operated on not later than 16 hours after human chorionic gonadotropin administration. The recovered eggs were kept at room temperature until insemination and successively treated with hyaluronidase (0.1%) and trypsin (0.05%) for the removal of cumulus and zona pellucida, respectively. After thorough washing in 4 to 6 BWW drops, 20 to 30 zona-free eggs were immediately transferred for insemination to each already prepared sperm drop. The drops were covered with paraffin oil and further incubated for 3 hours. If final sperm concentration was :i). Conversely, SPA was good in the remaining 13 cases, and fertilization did occur in 10 cases. Thus, in this group of 26 patients with the lowest sperm quality, the difference between poor and good SPA was significant (X 2 = 7.54, df = 1, P < 0.007). Of 73 cases with a high sperm quality, total sperm count> 150 millions, motility index> 150, no fertilization in IVF cycle was demonstrated only in 6 cases (8.2%). Sperm penetrating assay was poor in only 1 of these cases and good in all 5 remaining cases. Thus, in cases with a high sperm quality, no false SPA negative result was observed, but the negative predictive value of SPA was not demonstrated. Power of SPAin Predicting IVF Outcome, as Compared with Routine Sperm Variables
To select the most contributory variables in a multifactorial correlation with IVF results, simple correlations and multivariate stepwise regression analysis were made. The simple correlations between all routine sperm variables and IVF oocyte fertilization were significant (total sperm count r = 0.263, motility index r = 0.302, normal form percent r = 0.349, P < 0.0002 for all variables) but lower than SPA. A two-step significant multiple regression was found between SPA, sperm variables, and IVF oocyte fertilization (r2 = 0.234, P < 0.0002). Sperm penetrating assay entered this equation at the first step, yielding the highest coefficient of determination contribution (r2 = 0.181). Normal form percent entered the equation at the second and last step, increasing r2 by 0.051. Motility index and total sperm count did not enter the IVF equation. If these variables were forced in a multiple regression, SPA and percent of normal form were still the only significant variables in the equation. This analysis shows that SPA result correlates better with IVF outcome than both sperm motility and count but is almost equal with the percent of normal form. Thus, SPA and sperm morphology, statistically independent, are complementary in predicting IVF outcome. DISCUSSION
The data presented in this study clearly indicate that the SPA, when enhanced sperm capacitation protocol is employed, using TEST -yolk buffer cold sperm preincubation, is a valuable prognostic tool in IVF having satisfactory predictive values in each 560
Soffer et al.
SPA predicts IVF outcome
category. Indeed, the large number of patients included in this study enabled detailed analysis of the data and accurate determination of fertilization probabilities in IVF across the 0% to 100% penetration spectrum. A reversal of fertilization probabilities occurs in the 20% to 35% of penetration. Males who scored in the range of 21 % to 100% penetration (good SPA category) had a markedly higher probability of fertilizing at least one of the eggs in IVF. The cutoff point between 20% and 21% delineates a shift from 26% success rate in IVF in the poor SPA category to 82% in the good penetration category. This cutoff line was found quite similar to the conventional cutoff of 10% to 15% (2.4) and identical to the one employed by Margalioth et al. (5) without enhanced capacitation. A recent study (19) reporting on 42 fertile males who underwent SPA with TEST -yolk preincubation also suggested the lower cutoff line of 20%. Using this cutoff point, a satisfactory level of specificity but not sensitivity was achieved; this issue is strongly debated in the literature (20, 21). As in other reports (7, 8), PR in IVF cycles improved as did the fertilization rate. Thus, although 8 of the 31 patients in the poor SPA category fertilized in IVF, only 1 of them had a fertilization rate >50% of the eggs. No pregnancies resulted from any of these males in subsequent IVF cycles. In this respect, although these 8 patients can be apparently classified as 26% false-negative, in clinical practice and in light of the dependence of pregnancy on fertilization rate, the poor SPA category appears as a powerful predictor for the identification of males with severely reduced fertilization capability. In contrast, fertilization rate was markedly increased in the good SPA category. In an additional group of 80 couples who entered the IVF program after the closure of the present data set, a similar trend was found (Y Soffer et al., unpublished data, 1992). The capability of SPA to identify males with unexplained infertility was reported recently by Lamb et al. (8). However, the low negative predictive value obtained in their study (46%) precludes any accurate prediction regarding the prospective outcome of fertilization in normospermic males who score less than five penetrations per hamster egg, a criterion employed by this group for the definition of subfertility (8, 9). Scanning of the various reports dealing with the use of SPA as a prescreening tool before IVF attempts (2, 5-9, 19) reveals wide variations in SPA protocols and even differences in the criteria for SPA interpretation. Because of the wide variety of poFertility and Sterility
tential variables in SPA, there is a need to calibrate each assay system. In these studies, it appears that the most fundamental variable is the sperm capacitation protocol before the sperm oocyte coincubation. Wide variations ranging from a short (2.3 hours) at 37°C to 42 hours at 4°C sperm capacitation protocols were reported (7, 8). In an assay aimed at assessing the capacitation process, these variations have significant bearing on SPA results and their interpretation. In light of the intended use of the assay as a predictor of fertilizing probability, it is clear that the SPA has to be evaluated in terms of its negative and positive predictive values. In the two reports in which short capacitation at 37°C were employed, poor negative predictive values of 0% (11) and 75% (10) were reported. The group that practices extended capacitation period of 42 hours in the cold (7) reported high positive predictive value of 98% but weaker negative value of 46%. Moreover, the extreme capacitation conditions employed by the latter group resulted in enhanced polyspermia (7, 8, 22), which led this group to employ the penetration index (number of sperm per penetrated egg) as a criterion. Possibly because of the use of a shorter capacitation period, we found a lower penetration index and the criterion of penetration score was more diagnostic than penetration index. Similar to this study, Margalioth et al. (5) also employed overnight capacitation period, but in BWW at 37°C and not in TEST-yolk buffer at 4°C. These authors failed to find any correlation between SPA penetration score and fertilization rate in IVF. Our data also show that SPA cannot predict the rate of fertilization but only if it will occur. However, a poor SP A result, if repeated, predicts not only a high fertilization failure rate but also very poor chances of a high fertilization rate and almost no chances of a subsequent pregnancy. Supportive evidence that enhanced sperm capacitation markedly reduces false-negative rate (increases negative predictive value and assay sensitivity rate) was recently reported (9, 19,23). In this study, no evidence of any increase of the false-positive scores was found. This negative finding is noteworthy because the artificial enhancement of sperm capacitation used here would be susceptible, as recently said (20), to false high scores, in contrast to the standard SPA. Moreover, focusing SPA in the low sperm quality group underlined its value. Performing SPA yields not only the specific SPA results but also all the routine semen variables always checked in the fresh ejaculate. The close correlation found between SPA results and sperm variVol. 58, No.3, September 1992
ables, mainly total sperm count, is in contradiction to a previous study (24) in which a correlation was found with sperm morphology only but not with sperm motility. However, these authors did not use TEST -yolk preincubation. This correlation may be used as an argument against SPA, which is obviously cumbersome. This may be true in patients with a high sperm quality but not in those with a low sperm quality. Because sperm morphology and SPA are both significantly related to oocyte fertilization in IVF cycles, they should both be checked in the latter patients before entering IVF cycles. In the poor category (31 cases), with 8 ETs (26%) and no pregnancy (0%) observed, the expected PR is difficult to assess. If in the next one or two cases. ET and pregnancy does occur, their rate might rise to 28% to 30% and 3% to 6%, respectively. So, in this category, the expected PR reasonably stays somewhere between 0% and 6%. The couples should be informed of this expected low but possibly not nil PR. If in spite of these poor prospects, these couples choose to try IVF before referring to other solutions, one should accept their decision. However, if IVF fails in spite of a good SPA with TEST -yolk buffer sperm preincubation, it is very tempting to try this sperm enhanced capacitation in the next IVF cycles, mainly if morphology is good. Thus, fertilization was recently achieved in a small group of cases (25). This would also increase SPA sensitivity in predicting IVF. Larger studies are required to validate and recommend such modification in sperm preparation for IVF. In conclusion, it appears that by performing SPA with enhanced capacitation, this assay can effectively predict fertilization and PR. The SPA does also identify, mainly among men with a low semen quality, those infertile males who specifically suffer from a concealed diminished penetration ability. Acknowledgments. We are deeply grateful to Mrs. Sarita Kaufman, B.Sc., for her skillful technical assistance. We are also indebted to Peri Kedem, Ph.D., Bar Han University, Ramat Gan, Israel, for her statistical reviewing. REFERENCES 1. Yanagimachi R, Yanagimachi H, Rogers JJ. The use of zona free animal ova as a test system for the assessment of the fertilizing capacity of human spermatozoa. Bioi Reprod 1976;15:471-6. 2. Rogers BJ. The sperm penetration assay: its usefulness reevaluated. Fertil SteriI1985;43:821-40. 3. Yanagimachi R. Mammalian fertilization. In: Knobil E, Neill J, editors. The physiology of reproduction. New York: Raven, 1988:135-85.
Soffer et al.
SPA predicts IVF outcome
561
4. Martin RH, Taylor PJ. Reliability and accuracy of the zonafree hamster ova assay in the assessment of male fertility. Br J Obstet GynaecoI1982;89:951-6. 5. Margalioth EJ, Navot D, Laufer N, Lewin A, Rabinovitz R, Schenker JG. Correlation between the zona free hamster egg sperm penetration assay and in-vitro fertilization. Fertil Steril 1986;45:665-70. 6. Hirsch I, Gibbons WE, Lipshultz LI, Rossavik KK, Young RL, Poindexter AN, et al. In vitro fertilization in couples with male factor infertility. Fertil SteriI1986;45:659-64. 7. Smith RG, Johnson AR, Lamb DJ, Lipshultz LI. Functional tests of spermatozoa; sperm penetration assay. Urol Clin North Am 1987;14:451-8. 8. Lamb DJ, Johnson AR, Bassham BL, Lipshultz LI. How sperm penetration assay diagnoses infertility. Contemp Obstet GynecoI1989;31:108-14. 9. McClure RD, Tom RA, Dandekar FV. Optimizing the sperm penetration assay with human follicular fluid. Fertil Steril 1990;53:546-50. 10. Ausmanas M, Tureck RW, Blasco L, Kopf GS, Ribas J, Mastroianni L. The zona-free hamster egg penetration assay as a prognostic indicator in a human in vitro fertilization program. Fertil Steril1985;43:433-7. 11. WolfDP, Sokoloski JE, Quigley MM. Correlation of human in vitro fertilization with the hamster egg bioassay. Fertil Steril 1983;40:53-9. 12. Soffer Y, Ron-El R, Golan A, Herman A, Caspi E, Samra Z. Male genital mycoplasmas and Chlamydia trachomatis cultures; its relationship with accessory gland function, sperm quality and autoimmunity. Fertil Steril1990;53:331-6. 13. World Health Organization. WHO laboratory manual for the examination of human semen and semen-cervical mucus interaction. Cambridge: The Press Syndicate of the University of Cambridge, 1987:3-15. 14. Samuel T, Soffer Y, Caspi E. Sperm egg penetration of human spermatozoa treated with various rabbit antisera to human sperm antigens. Clin Exp ImmunoI1987;67:454-9. 15. Bigger JD, Whitten WK, Wittingham DG. The culture of mouse embryos in vitro. In: Daniel JC, editor. Methods in
562
Soffer et al.
SPA predicts IVF outcome
16.
17.
18. 19.
20.
21.
22.
23.
24.
25.
mammalian embryology. San Francisco: Freeman, 1971:86116. Caspi E, Ron-El R, Golan A, Nachum H, Herman A, Soffer Y, et al. Results of in vitro fertilization and embryo transfer by combined long-acting gonadotropin-releasing hormone DTrp-6-luteinizing hormone releasing hormone and gonadotropins. Fertil SteriI1989;51:95-9. Ron-El R, Nachum H, Herman A, Golan A, Caspi E, Soffer Y. Delayed fertilization and poor embryonic development associated with impaired semen quality. Fertil Steril 1991;55: 338-44. Eliasson R. Analysis of semen. In: Burger H, de Kretser D, editors. The testis. New York: Raven, 1981:381-5. Falk RM, Silverberg KM, Fetterolf PM, Kirchner FK, Rogers BJ. Establishment of TEST-yolk buffer enhanced sperm penetration assay limits for fertile males. Fertil SteriI1990;54: 121-6. Muller CH, Zarutskie PW, Stenchever MA, Soules MR. The sperm penetration assay; one of the best methods we have. Obstet Gynecol Rep 1990;2:412-20. Grimes DA. The sperm penetration assay; a biological phenomenon in search of a clinical use. Obstet Gynecol Rep 1990;2:413-24. Johnson AR, Syms AJ, Lipshultz LI, Smith RG. Conditions influencing human sperm capacitation and penetration of zona-free hamster ova. Fertil SteriI1984;41:603-8. Cohen J, Fehilly CB, Walters DE. Prolonged storage of human spermatozoa at room temperature or in a refrigerator. Fertil Steril 1985;44:254-62. Kruger TF, Swanson RJ, Hamilton M, Simmons KF, Acosta AA, Matta JF, et al. Abnormal sperm morphology and other semen parameters related to the outcome of the hamster oocyte human sperm penetration assay. Int J Androl 1988;11: 107-13. Katayama KP, Stehlik E, Roesler M, Jeyendran RS, Holmgren WJ, Zaneveld LJD. Treatment of human spermatozoa with an egg yolk medium can enhance the outcome of in vitro fertilization. Fertil Steril 1989;52:1077-9.
Fertility and Sterility
Vol. 58, No.3, September 1992
Copyright © 1992 The American Fertility Society
Printed on acid·free paper in U.S.A.
Prediction of in vitro fertilization outcome by sperm penetration assay with TEST-yolk buffer preincubation*
Yigal Soffer, M.D. t Abraham Golan, M.D. Arie Herman, M.D.
Mordechai Pansky, M.D. Eliahu Caspi, M.D. Raphael Ron-EI, M.D.
Department of Obstetrics and Gynecology, Tel Aviv University, Sackler School of Medicine, Assaf Harafe Medical Center, Zerifin, Israel
Objective: To evaluate sperm penetration assay (SPA) value as a screening tool before in vitro fertilization (IVF). Design: Follow-up study comparing sperm variables and IVF outcome. Setting: Infertile couples in an academic research environment. Patients, Participants: Two hundred forty-one infertile couples scheduled for IVF. Interventions: Sperm penentrating assay with cold Tes-TRIS (TEST) Yolk buffer semen preincubation and IVF. Main Outcome Measures: Percent of egg penetration recoded into poor and good category (0% to 20%, 21 % to 100%) and compared with fertilization, embryo transfer, and pregnancy rate (PR) in IVF, as well as sperm count motility and morphology. Results: Sperrn penetrating assay predicted fertilization with a high negative (74%) and positive (82%) predictive rate and good specificity value (0.96). One of 31 patients in the poor SPA category (3%) fertilized >50% of eggs; no pregnancy occurred. In the good SPA category, 87 of 210 patients (41 %) fertilized 2':50% of eggs with 34.3 % PRo Sperm penetrating assay correlated better with fertilization rate than did sperm count and motility but not morphology. Conclusions: Sperm penetrating assay is a useful screening assay before IVF together with sperm morphology Fertil Steril 1992;58:556-62 Key Words: Sperm egg penetration, TEST -yolk buffer enhanced capacitation, in vitro fertilization, sperm motility, count, morphology
In the array of tests aimed at the evaluation of male fertilizing capability, sperm penetrating assay (SPA) stands out as a unique bioassay (1, 2). This assay directly addresses four of the most decisive physiological processes in fertilization mechanism: [1] the capacitation, [2] acrosome reaction, [3] fusion of sperm cell with the oolema, and [4] chromatin decondensation with head swelling in the egg cytoplasma. Attempts to assess the significance of this bioassay in the diagnosis of male infertility have
Received June 19, 1991; revised and accepted May 1, 1992. the 7th World Congress on IVF and Assisted Procreations, Paris, France, June 30 to July 3, 1991. t Reprint requests: Yigal Soffer, M.D., Infertility Clinic, Department of Obstetrics and Gynecology, Assaf Harafe Medical Center, Zerifin 70300, Israel.
* Presented at
556
Soffer et al. SPA predicts IVF outcome
been confounded by the multifactorial nature of the fertilization process and potential associated female factors (2-4). The advent of in vitro fertilization and embryo transfer (IVF -ET) has contributed in eliminating various male and female factors by bringing human spermatozoa and oocytes into direct physical contact. In a somewhat similar manner, the SPA is capable of directly detecting faults in sperm penetrating capability. Most (5-9) but not all studies (10) that evaluated SPA's capability to predict IVF outcome presented supportive evidence for the use of this bioassay as a reliable detector of a male factor, and as such, as a screening tool for IVF. However, the technical difficulties, lack of standardized assay protocols, and even differences in the measured variables (percent of penetration [2,4,6,9-11] verFertility and Sterility
sus penetration index [8]) have hindered the acceptance of this important assay for routine use in many IVF programs and infertility centers. This study presents, to our knowledge, the largest ever reported series in conjunction with IVF. We have undertaken first, to examine the predictive value of SPA under conditions that enhance sperm capacitation, as a reliable predictor of poor male fertilization ability in IVF and second, to evaluate quantitatively the correlation between SPA result and IVF outcome in terms of fertilized oocytes ET and pregnancy rates (PRs). MATERIALS AND METHODS
Two hundred forty-one couples participating in IVF program entered this study. The female partner's age was 31.92 ± 4.49 (range 21 to 41 years). All these women underwent a routine infertility work-up including menstrual cycle evaluation, hysterosalpingography, and eventually hysteroscopy and laparoscopy. Female diagnosis was as follows: infertility of unknown origin 53.5%, tubal factor 38.2%, endometriosis 3.1 %, and other causes (polycystic ovarian disease, ovarian failure, female antisperm isoantibodies, and uterine pathological conditions) 5.2%. All males underwent a clinical examination and repeated semen analysis (12), which was performed in a 10-tIm-deep counting chamber (Makler Chamber, Sefi Medical Instruments, Haifa, Israel). Sperm quality was assessed by a motility index. % of motile cells X grade of motility (0 to 4), total sperm count, sperm concentration X semen volume and normal sperm form percentage on air-dried eosin - nigrosin stained smears examined at high-power field (HPF) (X1,000). The normal forms included the spermatozoa with an oval form according to World Health Organization criteria (13) and those with a well-developed long acrosome region. Sperm penetrating assay was performed 3-6 weeks before IVF. Sperm Penetration Assay
Sperm penetrating assay procedure was performed as previously described (14) with slight modifications. Ejaculates were allowed to liquefy for a maximum of 10 to 30 minutes. A small aliquot (0.2 to 0.4 mL) was kept for analysis, and all the remaining volume immediately diluted according to the semen viscosity in one or two volumes of Tes-TRIS (TEST: Sigma Chemical Co., St. Louis, MO) yolk buffer and kept in vertical tubes at 4°C for 18 to 22 hours. At the end of incubation, supernatant was carefully reVol. 58, No.3, September 1992
moved. Then, in addition to the above protocol (14), a sperm layering and rise-up procedure was done as follows. The sedimented sperm was resuspended in 0.25 mL of Biggers, Whitten, and Whittingham (BWW) buffer (15) enriched with 1.75% wt/vol bovine serum albumin, fraction V (Sigma Chemical Co.) overlaid with 1 mL of same medium in 30°C inclinated tubes, and sperm was allowed to rise up for 1 hour in a CO 2 incubator (at 37°C under 5% CO 2 atmosphere and saturated humidity). The supernatant containing motile spermatozoa was centrifuged, and sperm pellet was resuspended in a small amount of BWW and finally adjusted to a concentration of 5 to 10 X 106 motile cells/mL. A O.l-mL drop of this sperm suspension was prepared and kept in the CO 2 incubator until addition of eggs. Female golden hamsters were superovulated as already described (14). In this study, the females were 7 to 9 weeks old and operated on not later than 16 hours after human chorionic gonadotropin administration. The recovered eggs were kept at room temperature until insemination and successively treated with hyaluronidase (0.1%) and trypsin (0.05%) for the removal of cumulus and zona pellucida, respectively. After thorough washing in 4 to 6 BWW drops, 20 to 30 zona-free eggs were immediately transferred for insemination to each already prepared sperm drop. The drops were covered with paraffin oil and further incubated for 3 hours. If final sperm concentration was :i). Conversely, SPA was good in the remaining 13 cases, and fertilization did occur in 10 cases. Thus, in this group of 26 patients with the lowest sperm quality, the difference between poor and good SPA was significant (X 2 = 7.54, df = 1, P < 0.007). Of 73 cases with a high sperm quality, total sperm count> 150 millions, motility index> 150, no fertilization in IVF cycle was demonstrated only in 6 cases (8.2%). Sperm penetrating assay was poor in only 1 of these cases and good in all 5 remaining cases. Thus, in cases with a high sperm quality, no false SPA negative result was observed, but the negative predictive value of SPA was not demonstrated. Power of SPAin Predicting IVF Outcome, as Compared with Routine Sperm Variables
To select the most contributory variables in a multifactorial correlation with IVF results, simple correlations and multivariate stepwise regression analysis were made. The simple correlations between all routine sperm variables and IVF oocyte fertilization were significant (total sperm count r = 0.263, motility index r = 0.302, normal form percent r = 0.349, P < 0.0002 for all variables) but lower than SPA. A two-step significant multiple regression was found between SPA, sperm variables, and IVF oocyte fertilization (r2 = 0.234, P < 0.0002). Sperm penetrating assay entered this equation at the first step, yielding the highest coefficient of determination contribution (r2 = 0.181). Normal form percent entered the equation at the second and last step, increasing r2 by 0.051. Motility index and total sperm count did not enter the IVF equation. If these variables were forced in a multiple regression, SPA and percent of normal form were still the only significant variables in the equation. This analysis shows that SPA result correlates better with IVF outcome than both sperm motility and count but is almost equal with the percent of normal form. Thus, SPA and sperm morphology, statistically independent, are complementary in predicting IVF outcome. DISCUSSION
The data presented in this study clearly indicate that the SPA, when enhanced sperm capacitation protocol is employed, using TEST -yolk buffer cold sperm preincubation, is a valuable prognostic tool in IVF having satisfactory predictive values in each 560
Soffer et al.
SPA predicts IVF outcome
category. Indeed, the large number of patients included in this study enabled detailed analysis of the data and accurate determination of fertilization probabilities in IVF across the 0% to 100% penetration spectrum. A reversal of fertilization probabilities occurs in the 20% to 35% of penetration. Males who scored in the range of 21 % to 100% penetration (good SPA category) had a markedly higher probability of fertilizing at least one of the eggs in IVF. The cutoff point between 20% and 21% delineates a shift from 26% success rate in IVF in the poor SPA category to 82% in the good penetration category. This cutoff line was found quite similar to the conventional cutoff of 10% to 15% (2.4) and identical to the one employed by Margalioth et al. (5) without enhanced capacitation. A recent study (19) reporting on 42 fertile males who underwent SPA with TEST -yolk preincubation also suggested the lower cutoff line of 20%. Using this cutoff point, a satisfactory level of specificity but not sensitivity was achieved; this issue is strongly debated in the literature (20, 21). As in other reports (7, 8), PR in IVF cycles improved as did the fertilization rate. Thus, although 8 of the 31 patients in the poor SPA category fertilized in IVF, only 1 of them had a fertilization rate >50% of the eggs. No pregnancies resulted from any of these males in subsequent IVF cycles. In this respect, although these 8 patients can be apparently classified as 26% false-negative, in clinical practice and in light of the dependence of pregnancy on fertilization rate, the poor SPA category appears as a powerful predictor for the identification of males with severely reduced fertilization capability. In contrast, fertilization rate was markedly increased in the good SPA category. In an additional group of 80 couples who entered the IVF program after the closure of the present data set, a similar trend was found (Y Soffer et al., unpublished data, 1992). The capability of SPA to identify males with unexplained infertility was reported recently by Lamb et al. (8). However, the low negative predictive value obtained in their study (46%) precludes any accurate prediction regarding the prospective outcome of fertilization in normospermic males who score less than five penetrations per hamster egg, a criterion employed by this group for the definition of subfertility (8, 9). Scanning of the various reports dealing with the use of SPA as a prescreening tool before IVF attempts (2, 5-9, 19) reveals wide variations in SPA protocols and even differences in the criteria for SPA interpretation. Because of the wide variety of poFertility and Sterility
tential variables in SPA, there is a need to calibrate each assay system. In these studies, it appears that the most fundamental variable is the sperm capacitation protocol before the sperm oocyte coincubation. Wide variations ranging from a short (2.3 hours) at 37°C to 42 hours at 4°C sperm capacitation protocols were reported (7, 8). In an assay aimed at assessing the capacitation process, these variations have significant bearing on SPA results and their interpretation. In light of the intended use of the assay as a predictor of fertilizing probability, it is clear that the SPA has to be evaluated in terms of its negative and positive predictive values. In the two reports in which short capacitation at 37°C were employed, poor negative predictive values of 0% (11) and 75% (10) were reported. The group that practices extended capacitation period of 42 hours in the cold (7) reported high positive predictive value of 98% but weaker negative value of 46%. Moreover, the extreme capacitation conditions employed by the latter group resulted in enhanced polyspermia (7, 8, 22), which led this group to employ the penetration index (number of sperm per penetrated egg) as a criterion. Possibly because of the use of a shorter capacitation period, we found a lower penetration index and the criterion of penetration score was more diagnostic than penetration index. Similar to this study, Margalioth et al. (5) also employed overnight capacitation period, but in BWW at 37°C and not in TEST-yolk buffer at 4°C. These authors failed to find any correlation between SPA penetration score and fertilization rate in IVF. Our data also show that SPA cannot predict the rate of fertilization but only if it will occur. However, a poor SP A result, if repeated, predicts not only a high fertilization failure rate but also very poor chances of a high fertilization rate and almost no chances of a subsequent pregnancy. Supportive evidence that enhanced sperm capacitation markedly reduces false-negative rate (increases negative predictive value and assay sensitivity rate) was recently reported (9, 19,23). In this study, no evidence of any increase of the false-positive scores was found. This negative finding is noteworthy because the artificial enhancement of sperm capacitation used here would be susceptible, as recently said (20), to false high scores, in contrast to the standard SPA. Moreover, focusing SPA in the low sperm quality group underlined its value. Performing SPA yields not only the specific SPA results but also all the routine semen variables always checked in the fresh ejaculate. The close correlation found between SPA results and sperm variVol. 58, No.3, September 1992
ables, mainly total sperm count, is in contradiction to a previous study (24) in which a correlation was found with sperm morphology only but not with sperm motility. However, these authors did not use TEST -yolk preincubation. This correlation may be used as an argument against SPA, which is obviously cumbersome. This may be true in patients with a high sperm quality but not in those with a low sperm quality. Because sperm morphology and SPA are both significantly related to oocyte fertilization in IVF cycles, they should both be checked in the latter patients before entering IVF cycles. In the poor category (31 cases), with 8 ETs (26%) and no pregnancy (0%) observed, the expected PR is difficult to assess. If in the next one or two cases. ET and pregnancy does occur, their rate might rise to 28% to 30% and 3% to 6%, respectively. So, in this category, the expected PR reasonably stays somewhere between 0% and 6%. The couples should be informed of this expected low but possibly not nil PR. If in spite of these poor prospects, these couples choose to try IVF before referring to other solutions, one should accept their decision. However, if IVF fails in spite of a good SPA with TEST -yolk buffer sperm preincubation, it is very tempting to try this sperm enhanced capacitation in the next IVF cycles, mainly if morphology is good. Thus, fertilization was recently achieved in a small group of cases (25). This would also increase SPA sensitivity in predicting IVF. Larger studies are required to validate and recommend such modification in sperm preparation for IVF. In conclusion, it appears that by performing SPA with enhanced capacitation, this assay can effectively predict fertilization and PR. The SPA does also identify, mainly among men with a low semen quality, those infertile males who specifically suffer from a concealed diminished penetration ability. Acknowledgments. We are deeply grateful to Mrs. Sarita Kaufman, B.Sc., for her skillful technical assistance. We are also indebted to Peri Kedem, Ph.D., Bar Han University, Ramat Gan, Israel, for her statistical reviewing. REFERENCES 1. Yanagimachi R, Yanagimachi H, Rogers JJ. The use of zona free animal ova as a test system for the assessment of the fertilizing capacity of human spermatozoa. Bioi Reprod 1976;15:471-6. 2. Rogers BJ. The sperm penetration assay: its usefulness reevaluated. Fertil SteriI1985;43:821-40. 3. Yanagimachi R. Mammalian fertilization. In: Knobil E, Neill J, editors. The physiology of reproduction. New York: Raven, 1988:135-85.
Soffer et al.
SPA predicts IVF outcome
561
4. Martin RH, Taylor PJ. Reliability and accuracy of the zonafree hamster ova assay in the assessment of male fertility. Br J Obstet GynaecoI1982;89:951-6. 5. Margalioth EJ, Navot D, Laufer N, Lewin A, Rabinovitz R, Schenker JG. Correlation between the zona free hamster egg sperm penetration assay and in-vitro fertilization. Fertil Steril 1986;45:665-70. 6. Hirsch I, Gibbons WE, Lipshultz LI, Rossavik KK, Young RL, Poindexter AN, et al. In vitro fertilization in couples with male factor infertility. Fertil SteriI1986;45:659-64. 7. Smith RG, Johnson AR, Lamb DJ, Lipshultz LI. Functional tests of spermatozoa; sperm penetration assay. Urol Clin North Am 1987;14:451-8. 8. Lamb DJ, Johnson AR, Bassham BL, Lipshultz LI. How sperm penetration assay diagnoses infertility. Contemp Obstet GynecoI1989;31:108-14. 9. McClure RD, Tom RA, Dandekar FV. Optimizing the sperm penetration assay with human follicular fluid. Fertil Steril 1990;53:546-50. 10. Ausmanas M, Tureck RW, Blasco L, Kopf GS, Ribas J, Mastroianni L. The zona-free hamster egg penetration assay as a prognostic indicator in a human in vitro fertilization program. Fertil Steril1985;43:433-7. 11. WolfDP, Sokoloski JE, Quigley MM. Correlation of human in vitro fertilization with the hamster egg bioassay. Fertil Steril 1983;40:53-9. 12. Soffer Y, Ron-El R, Golan A, Herman A, Caspi E, Samra Z. Male genital mycoplasmas and Chlamydia trachomatis cultures; its relationship with accessory gland function, sperm quality and autoimmunity. Fertil Steril1990;53:331-6. 13. World Health Organization. WHO laboratory manual for the examination of human semen and semen-cervical mucus interaction. Cambridge: The Press Syndicate of the University of Cambridge, 1987:3-15. 14. Samuel T, Soffer Y, Caspi E. Sperm egg penetration of human spermatozoa treated with various rabbit antisera to human sperm antigens. Clin Exp ImmunoI1987;67:454-9. 15. Bigger JD, Whitten WK, Wittingham DG. The culture of mouse embryos in vitro. In: Daniel JC, editor. Methods in
562
Soffer et al.
SPA predicts IVF outcome
16.
17.
18. 19.
20.
21.
22.
23.
24.
25.
mammalian embryology. San Francisco: Freeman, 1971:86116. Caspi E, Ron-El R, Golan A, Nachum H, Herman A, Soffer Y, et al. Results of in vitro fertilization and embryo transfer by combined long-acting gonadotropin-releasing hormone DTrp-6-luteinizing hormone releasing hormone and gonadotropins. Fertil SteriI1989;51:95-9. Ron-El R, Nachum H, Herman A, Golan A, Caspi E, Soffer Y. Delayed fertilization and poor embryonic development associated with impaired semen quality. Fertil Steril 1991;55: 338-44. Eliasson R. Analysis of semen. In: Burger H, de Kretser D, editors. The testis. New York: Raven, 1981:381-5. Falk RM, Silverberg KM, Fetterolf PM, Kirchner FK, Rogers BJ. Establishment of TEST-yolk buffer enhanced sperm penetration assay limits for fertile males. Fertil SteriI1990;54: 121-6. Muller CH, Zarutskie PW, Stenchever MA, Soules MR. The sperm penetration assay; one of the best methods we have. Obstet Gynecol Rep 1990;2:412-20. Grimes DA. The sperm penetration assay; a biological phenomenon in search of a clinical use. Obstet Gynecol Rep 1990;2:413-24. Johnson AR, Syms AJ, Lipshultz LI, Smith RG. Conditions influencing human sperm capacitation and penetration of zona-free hamster ova. Fertil SteriI1984;41:603-8. Cohen J, Fehilly CB, Walters DE. Prolonged storage of human spermatozoa at room temperature or in a refrigerator. Fertil Steril 1985;44:254-62. Kruger TF, Swanson RJ, Hamilton M, Simmons KF, Acosta AA, Matta JF, et al. Abnormal sperm morphology and other semen parameters related to the outcome of the hamster oocyte human sperm penetration assay. Int J Androl 1988;11: 107-13. Katayama KP, Stehlik E, Roesler M, Jeyendran RS, Holmgren WJ, Zaneveld LJD. Treatment of human spermatozoa with an egg yolk medium can enhance the outcome of in vitro fertilization. Fertil Steril 1989;52:1077-9.
Fertility and Sterility
