Glass wool column filtration, an advantageous way of preparing semen samples for intracytoplasmic sperm injection: an...

The characteristics of the glass wool filtration technique for the preparation of semen samples in an intracytoplasmic sperm injection programme (ICSI) were compared with a two-layer Percoll density gradient procedure. Half of each of 25 semen samples were prepared by each technique. The oocytes from each patient were randomly injected, half with spermatozoa prepared by glass wool filtration and half with Percoll-separated spermatozoa from the same semen sample. The percentage of recovered motile spermatozoa, the total motile sperm count, the percentage of morphologically normal forms, the fertilization rate, the cleavage rate and embryo quality obtained with both preparations were analysed. The 95% confidence intervals obtained through the Altman-Bland analysis showed that the percentage of motile spermatozoa recovered was 3.5-9.9% higher with the glass wool filtration technique, and this was highly correlated (r = 0.92, P = 0.0001) to the method. Similarly, the total number of spermatozoa available for ICSI was 0.18 x 10(6)-2.44 x 10(6) higher with the glass wool column and was highly correlated to the method (r = 0.956, P = 0.0001). Also, the percentage of normal forms was 1.25-3.31% higher after glass wool filtration but was poorly correlated to the method (r = 0.47, P = 0.017). Out of 100 metaphase II oocytes injected with glass wool-extracted spermatozoa, 77 fertilized and 72 cleaved. Out of 97 metaphase II oocytes injected with Percoll-selected spermatozoa, 71 fertilized and 69 cleaved. These results were not statistically different. The mean +/- SEM embryo quality score for the glass wool group (2.90 +/- 0.27) was the same as that for the Percoll group (2.80 +/- 0.24). No fertilization failures occurred and 11 patients (44% per oocyte retrieval) became pregnant. It was concluded that glass wool filtration for semen preparation in an ICSI programme offers higher sperm recovery and sperm morphology of superior quality than with the classic two-layer Percoll gradient method, without affecting the fertilization rate and embryo quality.