Computer-aided evaluation of assessment of grade a spermatozoa by experienced technicians

Transcription

1 Computer-aided evaluation of assessment of grade a spermatozoa by experienced technicians Trevor G. Cooper, Ph.D., and Ching-Hei Yeung, Ph.D. Institute of Reproductive Medicine of the University of Münster, Münster, Germany Objective: To determine how well technicians assess progressively motile spermatozoa. Design: Comparison of computerized and technician assessments of grade a spermatozoa. Setting: University fertility clinic. Patient(s): Infertile men of barren marriages. Intervention(s): None. Main Outcome Measure(s): Threshold velocities of spermatozoa categorized as grade a. Result(s): The ability of andrology technicians to distinguish grade a from grade b spermatozoa was examined monthly. Computer-aided sperm analysis (CASA) provided the threshold velocities used by technicians to categorize grade a spermatozoa. The mean cutoff value for grades a/b (25.5 m/s) was close to the designated velocity, but the cutoff for grades b/c sperm (9.9 m/s) was higher. Cutoff values for grade a spermatozoa were significantly, positively, and linearly related with the mean velocity of spermatozoa obtained by CASA, and deviated from it in a systematic way. The percentage of grade a spermatozoa judged by technicians was poorly correlated with the percentage grade a indicated by CASA and the cutoff velocity of grade a spermatozoa depending on the overall sperm quality of the semen sample. Conclusion(s): Consideration should be given to replacing the subdivision of progressing spermatozoa into two categories with one of progressive spermatozoa, or the use of CASA to distinguish grade a sperm reliably should be insisted upon. (Fertil Steril 6;85: by American Society for Reproductive Medicine.) Key Words: CASA, WHO motility grades, subjective assessment, quality control Sperm motility is routinely monitored in the andrology laboratory because it is crucial in the assessment of the infertile man. The World Health Organization (WHO) Laboratory Manual for the Examination of Human Semen and Sperm-Cervical Mucus Interaction (1) recommends assessing this parameter by categorizing spermatozoa as immotile (grade d), locally motile (grade c: flagellating but nonprogressive motion), slowly progressive (grade b), and rapidly progressive (grade a). A grade a spermatozoon is classified as swimming at 25 m/s at 37 C and thus moves at more than five head lengths, or half a tail length, per second. However, it is difficult for technicians to assess the velocity of moving spermatozoa, despite these clues from sperm size, and variations between estimates from different laboratories are large (2, 3). Computer-aided sperm analysis (CASA) can objectively measure sperm velocity, and it is possible to estimate the threshold speed of spermatozoa considered by technicians as grade a or grade b spermatozoa (4). In this study, thevariation in assigning grade a and grade b spermatozoa by technicians is reported over a 43-month period, and the comparison to the computer-derived values Received February 3, 5; revised and accepted July 8, 5. Reprint requests: Trevor G. Cooper, Ph.D., University of Münster, Institute of Reproductive Medicine, Domagkstrasse 11, Münster D-48129, Germany (FAX: ; TrevorG.Cooper@ukmuenster.de). for each sample was analyzed. The CASA estimates of the percentage of grade a spermatozoa in these samples was compared with those of the technicians. MATERIALS AND METHODS As part of the routine internal quality control (IQC) of semen analysis, a single semen sample was chosen each month to reflect the range of samples routinely assessed in the institute and analyzed for all semen parameters by three to five experienced andrology technicians. All patients provided written consent that excess semen could be used for research purposes. In this study, the sperm motility assessments of these IQC samples from April 1 October 4 (43 months) were analyzed in detail. The percentages of WHO (1) grade a ( 25 m/s at 37 C), grade b ( 25 5 m/s), grade c (locally motile but nonprogressive: 5 m/s), and grade d (immotile) spermatozoa monitored at 37 C were recorded by each technician (the mean of duplicate counts of 1 spermatozoa). Aliquots of the same semen samples were analyzed by CASA (Hamilton-Thorne Version 1.8, Beverly, MA), and the straight-line velocities (VSL) considered by each technician to distinguish grade a from grade b and grade b from grade c spermatozoa were calculated from the single tracks as described by Yeung et al. (4). This involved first recalculating the percentage of sper- 2 Fertility and Sterility Vol. 85, No. 1, January /6/$32. Copyright 6 American Society for Reproductive Medicine, Published by Elsevier Inc. doi:1.116/j.fertnstert

2 FIGURE 1 The fluctuation in cutoff velocities (ordinate) for the categorization of motile spermatozoa as grade a (black circles) and grade b (gray circles) by technicians over a four year period (abscissa). Average values over the period are given as a broken line ( grade a ) and a solid line ( grade b ). 7 Cutoff value (VSL, µm/s) mean ± SD 1 Grade a Grade b Mean a (25.5 µm/s) Mean b (9.9 µm/s) Cooper. Technician assessment of grade a sperm. Fertil Steril 6. matozoa initially classified as grades a, b, and c (on the basis of the total sperm population) as a percentage of the motile (a b c) sperm population. The VSL values from the individual track data of all the motile sperm in the corresponding sample analyzed by the CASA (accuracy of the system in detection and tracking of motile sperm validated previously with errors 2%) (5) were listed in descending order, so that the grades a, b, and c motile sperm equivalent to the technician s values could be identified in the list. For example, if the recalculated grades a, b, and c were %, %, and % respectively, and there were individual VSL values, the 11st value on the list would be the cutoff for grade a/b and the 161st value for b/c assumed by the technician for that sample. The mean cutoff value from all the technicians was calculated for each semen sample, and the intertechnician coefficient of variation (CV) for the a/b and b/c cutoffs was calculated for each month. The percentage grade a sperm in the samples was also determined from the CASA data and compared to that of the technicians. April 1 October 4 were m/s, and m/s, respectively (Fig. 1). A large variation occurred over the years in the assessments of cut-off velocities of both grade a and grade b spermatozoa, with maximum values of 62.6 m/s and 29.5 m/s, respectively; minimum values of 7.2 m/s and 2.8 m/s, respectively; and ranges of 45 m/s and 21 m/s, respectively (Table 1). The technicians estimates of the grade a/b cutoff velocities, targeting 25 m/s, were dependent upon the overall motility of the sperm in the sample as determined by CASA (Fig. 2). Estimates of the percentage of grade a spermatozoa made by both technicians and computer indicated correlations but did not agree well: in samples with a high percentage ( 35%) of grade a sperm as assessed by CASA, technicians underestimated this percentage, whereas in samples with 35% grade a spermatozoa, technicians made overestimates (Fig. 3). For three of the four technicians, statistically significant (P.5) positive linear correlations with CASA values were obtained (CV.4.558). RESULTS The mean SD cutoff velocities of grades a/b and grades b/c spermatozoa determined each month over the period DISCUSSION The mean value from the technicians for the threshold speed of a grade a spermatozoon over the 43-month Fertility and Sterility 221

3 TABLE 1 Variation among technicians with time in the assessment of velocities for grade a, b, and c grade spermatozoa. Technician Mean cutoff ( m/s) SD ( m/s) CV (%) Max ( m/s) Min ( m/s) Range ( m/s) Grades a/b A B C D Average Grades b/c A B C D Average Note: SD standard deviation; CV coefficient of variation. Cooper. Technician assessment of grade a sperm. Fertil Steril 6. FIGURE 2 Cutoff values determined by each technician (A, B, C, D) for the categorization of grade a spermatozoa (ordinate) plotted against the mean velocity of the spermatozoa in the same semen samples determined by CASA (abscissa). Broken lines are the regression lines for each technician. Technicians' mean a/b cutoff (VSL µm/s) 1 A B C D Regression CASA (VSL µm/s) Cooper. Technician assessment of grade a sperm. Fertil Steril Cooper and Yeung Technician assessment of grade a sperm Vol. 85, No. 1, January 6

4 FIGURE 3 The percentage of grade a spermatozoa in each semen sample determined by each technician (A, B, C, D) (ordinate) plotted against the percentage of grade a spermatozoa determined by CASA from the cutoff value of 25 m/s (abscissa). Broken lines are the regression lines for each technician and the solid line is the line of identity. 8 Technicians' mean grade a (%) A B C D Regression Identity CASA grade a (%) Cooper. Technician assessment of grade a sperm. Fertil Steril 6. period from April 1 October 4 (26 m/s) closely approximated that designated by WHO (1) for grade a spermatozoa (25 m/s at 37 C). However, the cutoff for grade b spermatozoa (1 m/s) was twice that designated. The variability in the estimates of the cutoff velocities for both grade a and grade b spermatozoa by all technicians was large and varied above and below the designated values. That there appeared to be no improving trend with time, but rather a hunting around the mean, is a clear indication of how difficult it is for even highly experienced technicians to maintain a constant, objective concept of the velocity of spermatozoa. Confirmation of this view came from the dependence of the grade a/b cutoff value on the overall quality of the semen sample in terms of fast-moving spermatozoa. Thus, for samples expressing poor motility, the technicians tended to be more lenient in assigning grade a spermatozoa (i.e., the cutoff velocity for grade a spermatozoa was set lower), whereas with good semen samples presenting many fast spermatozoa, the mental cutoff for a grade a spermatozoon was increased in the subjective assessment (and fewer grade a sperm registered). This was confirmed when the percentages of grade a spermatozoa provided by technicians and computer were directly compared, and a clear trend to underestimation, or overestimation, of grade a sperm in samples with, respectively, higher, or lower, percentages of grade a spermatozoa assessed by CASA. The difficulty that technicians have in distinguishing consistently and reproducibly between grade a and grade b spermatozoa makes such an exercise impossible to control. Consideration should be given to replacing the subdivision of progressing spermatozoa into two categories with one of progressive spermatozoa, or insisting on the use of CASA to distinguish grade a sperm reliably. Acknowledgments: We thank Barbara Hellenkemper for calculating the cutoff values from semen samples analyzed by Ele Kürten, Sabine Rehr, Daniele Hanke, and Daniele Schmidt. REFERENCES 1. World Health Organization. Laboratory manual for the examination of human semen and sperm-cervical mucus interaction. 4th ed. Cambridge: Cambridge University Press, 1999:9 1. Fertility and Sterility 223

5 2. Cooper TG, Atkinson, AD, Nieschlag E. Experience with external quality control in spermatology. Hum Reprod 1999;14: Cooper TG, Björndahl L Vreeburg J Nieschlag E. Semen analysis and external quality control schemes for semen analysis need global standardization. Int J Androl 2;25: Yeung CH, Cooper TG, Nieschlag E. A technique for standardization and quality control of subjective sperm motility assessments in semen analysis. Fertil Steril 1997;67: Yeung CH, Nieschlag, E. Performance and comparison of CASA systems equipped with different phase-contrast optics. J Androl 1993;14: Cooper and Yeung Technician assessment of grade a sperm Vol. 85, No. 1, January 6