CHARACTERIZATION AND FREQUENCY-DISTRIBUTION OF SPERM ACROSOME REACTION AMONG NORMAL AND INFERTILE MEN

We studied the acrosome reaction (AR) in human spermatozoa from 58 normal fertile men and 232 infertile patients. The median AR in the normal population was 17%; AR was >9% in 83% of the subjects studied and only 2% of fertile individuals had a failed AR (< 5%). We calculated the within person variability using multiple specimens from each subject; the 95% confidence interval for each subject was 4 AR units above and 4 AR units below his mean value. In contrast, the median AR for the infertile population was 3%. AR values were below the normal range (< 5%) in 60% of the patients and only 26% of the patients had AR values >9%. Because patients entered the study sequentially, without regard to the aetiology of infertility, the patient population comprised a wide variety of subjects ranging from individuals with good to extremely poor semen quality. There was a progressively greater number of individuals with a failed AR as semen quality deteriorated (P < 0.01). However, among patients with good count, motility and morphology, 20% unexpectedly had a failed AR, and among patients with severely impaired semen quality, 29% had a positive AR. We also analysed the relationship of AR values with other semen measurements. The frequency of acrosomal reactions was significantly correlated (Spearman, P < 0.001) with morphology (r = 0.509), concentration (r = 0.524), progression (r = 0.416) and percentage motility (r = 0.356), but not with the percentage of sperm moving at a curvilinear velocity of less than or equal to 40 mu m/s. When we constructed a quadrant analysis by plotting morphology (percentage normal forms) against the AR value for each subject, values of AR and morphology for fertile men were uniformly distributed in all four quadrants, while 83% of infertile patients fell in the quadrant characterized by morphology <10% and/or AR <20%. Our data show that the frequency distributions of AR among fertile men and infertile patients differ markedly and that AR provides additional information on the quality of a semen specimen, not derived from conventional semen analysis. We propose that AR measurement might be a clinically useful diagnostic tool and address this issue in a companion paper where we analyse prospectively the correlation of AR with fertilization success at in-vitro fertilization.