Evaluation of flow cytometric methods to measure human sperm concentration.

Abstract

Human sperm concentration is usually assessed using hemocytometry (HM). However, external and internal quality control schemes have shown that the accuracy of this method is low overall. Flow cytometry (FC) is a rapid, accurate, and reproducible technology for the quantification of various cell populations. We used 3 FC methods for human sperm counting from a 1:1 mixture of a diluted semen sample with a suspension of fluorospheres of known concentrations. The events that represented sperm cells were detected according to 1) gating on size and granularity (FCM1), 2) gating on DNA staining by propidium iodide (FCM2), and 3) a combination of FCM1 and FCM2 (FCM3). Sperm concentration was calculated from the ratio of detected events to fluorosphere counts and fluorosphere concentration. A pilot study undertaken by 12 technicians from different laboratories to compare FCM1 with HM showed a general agreement between both methods, despite wide variations in sperm concentration exhibited by HM due to the use of unoptimized procedures. A second experiment indicated that the overall variability in sperm concentration assessment by FCM1 was lower than that produced by HM when performed by 2 technicians using optimal procedures for 3 preparations of the same semen samples. The overall mean coefficients of variation were 3.9% for FCM1 vs 8.0% for technician 1, 12.3% for technician 2 (P < .05), and 15.7% for both technicians (P < .05). FCM1, FCM2, and FCM3 were compared with HM performed by a single trained technician for 39 semen samples (triplicates) of various quality. Compared with HM, FCM1 and FCM2 overestimated the sperm concentration by 14% and 8%, respectively, against only 4% per million sperm for FCM3, which was effective for the full spectrum of sperm concentrations (except azoospermia). In conclusion, this study demonstrates that human sperm concentration can be accurately assessed by the FC method combining gating on cell size, granularity, and DNA staining.