Prediction of the in-vitro fertilization (IVF) potential of human spermatozoa using sperm function tests: the effect...

To examine the diagnostic significance of several criteria of semen quality and to determine whether their prognostic value is eroded by the time interval between assessment and the attempt at in-vitro fertilization (IVF) with embryo transfer, 73 couples undergoing IVF and embryo transfer therapy were studied. The ability of human spermatozoa to achieve fertilization in vitro was examined in relation to the conventional semen profile, sperm morphology, the computer-aided assessment of sperm movement, ionophore-induced acrosome reaction, acridine orange staining, and chemiluminescent signals induced by phorbol ester and N-formyl-methionyl-leucyl-phenylalanine (FMLP). Spermatozoa were examined both in semen and after preparation on Percoll, some weeks prior to IVF. Fertilization rates were noted to be significantly correlated with elements of sperm movement characteristics, sperm morphology, and reactive oxygen species generation. Prediction of fertilization rates in a stepwise multiple regression analysis was obtained using four variables: sperm morphology, FMLP-induced chemi-luminescence and sperm movement characteristics (beat cross frequency and straightness) (r approximately 0.5). When multiple logistic regression analysis was used to predict which samples would achieve fertilization rates above and below a 50% threshold, three variables of predictive value including linearity, average path velocity and FMLP-induced chemiluminescence were selected. Combination of these variables classified the samples achieving good or poor fertilization with an overall accuracy of 83.6%. The time interval between semen assessment and IVF had little effect on the predictive value of these tests. In conclusion, the fertilizing ability of human spermatozoa is related to sperm morphology, attributes of sperm movement and reactive oxygen species production. The time delay between testing and IVF did not appear to affect predictive accuracy.