The design of the Micro-Cell chamber was based on the physical properties of capillary flow, MATERIALS AND METHODS The Micro-Cell Chamber

Transcription

1 FERTILITY AND STERILITY Copyright " 1990 The American Fertility Society Vol. 53, No.5, May 1990 Printed on ocid-free poper in U.S.A. The influence of chamber characteristics on the reliability of sperm concentration and movement measurements obtained by manual and videomicrographic analysis* Kenneth A. Ginsburg, M.D. D. RandallArmant, Ph.D.t The Division of Reproductive Endocrinology and Infertility, Department of Obstetrics and Gynecology, Wayne State University School of Medicine, C.S. Matt Center for Human Growth and Development, Detroit, Michigan To assess the influence of chamber design and depth on the accuracy and precision of sperm measurements, mal)ual counting of a standardized latex bead solution and automated sperm motility measurements were made using a Makler chamber (Sefi Medical Industries, Haifa, Israel), Neubauer hemocytometer (American Optical Company, Buffalo, NY), and a new, disposable device (Micro-Cell; Cyto Fluidics, Inc., Silver Spring, MD). Bead counts obtained with the Micro-Cell chamber or hemocytometer were not statistically different from those determined by electronic particle counting, whereas Makler chamber counts were 62% higher. Makler counts had a significantly higher standard deviation, suggesting that counts made with this device are less reproducible. Analyzing live sperm samples, the percentage of motile sperm determined using Micro-Cell and Makler chambers were similar. However, significant differences in sperm concentration and mean velocity were found. The Micro-Cell disposable chamber provided consistent and accurate data on sperm concentration, percent motility, and mean velocity. These differences in sperm measurements emphasize the importance of sampling chamber characteristics on data reliability. Fertil Steril53:882, 1990 Contemporary methods for human semen analysis, both in the research setting and the clinical laboratory, have advanced greatly in recent years. Videomicrographic systems employing digital image processing allow examination of both individual and large numbers of spermatozoa with computation of average measurements for populations of sperm. The vigor and pattern of sperm movement can be described with measurements such as percentage motile sperm, curvilinear velocity, head beat amplitude and frequency, and path linearity or "straightness." Received October 18, 1989: revised and accepted January 10, * Presented at the 45th Annual Meeting of The American Fertility Society, San Francisco, California, November 13 to 16, t Reprint requests: D. Randall Armant, Ph.D., The Division of Reproductive Endocrinology and Infertility, Department of Obstetrics and Gynecology, Wayne State University School of Medicine, C.S. Mott Center for Human Growth and Development, 275 E. Hancock St., Detroit, Michigan For acceptable reproducibility and accuracy, standardized assay conditions must be rigorous 1-3 yet easily obtained. Various chambers 4 5 are employed for computer-assisted sperm movement analysis (CASMA), as well as manual determination of sperm concentration and percent motility. However, potential sources of error in using these chambers derive from their repeated handling, cleaning, and reassembly. A disposable, self-filling analysis chamber has recently been introduced. We tested the operating characteristics of the Micro Cell chamber (Cyto Fluidics, Inc., Silver Spring, MD) by determining its counting accuracy and precision, as well as biocompatibility. We report these results and compare its performance with other laboratory counting devices. MATERIALS AND METHODS The Micro-Cell Chamber The design of the Micro-Cell chamber was based on the physical properties of capillary flow, 882 Ginsburg and Armant Reliability of semen analysis chambers Fertility and Sterility

2 Figure 1 Photograph of the Micro-Cell chamber. A glass microscope slide was patterned with a 20-l'm layer of fluorocarbon material (white area) and epoxy (stippled appearing areas). When overlaid with a glass coverslip, the pattern left two clean glass (dark) regions in the center of the slide, separated by the epoxy-filled channels. These areas of uncoated glass constitute the two 20-l'm sample chambers, each bounded by the glass slide below, the affixed coverslip above, and the hydrophobic fluorocarbon emulsion on the sides. In each chamber a small area of uncoated glass protrudes past the coverslip, serving as a loading area. Two venting channels lead to the edge of the coverslip in each chamber, allowing air to exit during filling. wherein an aqueous solution will flow freely over a glass surface but not over a hydrophobic one. Micro-Cell chambers were constructed by coating a 25 X 75 mm glass slide with a nonwettable (hydrophobic) fluorocarbon material (Fig. 1). An ultravioletcuring epoxy was applied to certain areas of the slide, and a cover slip bonded to the center. The pattern of this fluorocarbon coating and the epoxy created two separate wettable chambers with a depth of 20 ± 2 ~m (tolerance determined by laser interferometry). Each of the chambers was separately filled by placing a 10 ~L sample on the uncoated surface of the chamber protruding slightly beyond the coverslip at either side of the pattern, and allowing the liquid to enter by capillary flow. As the fluid entered the chamber over a period of 60 seconds, displaced air was vented to the sides through hydrophobic regions created by the fluorocarbon coating. This bead solution was manually counted in three devices: (1) the Micro-Cell chamber; (2) a new, unused Makler chamber (Sefi Medical Industries, Haifa, Israel); and (3) a new, unused Neubauer hemocytometer (American Optical Company, Buffalo, NY). In the Micro-Cell and Makler chambers the stock bead solution was counted manually using a 1 X 1 em net eyepiece reticle that was divided into 100 squares and calibrated using a stage micrometer. A chamber thickness of 20 ~m Bead Count Manual Bead Counting To calibrate the counting chambers, a stock solution of 5 ~m latex beads (range 4.6 to 5.5 ~m; Coulter Electronics, Inc., Hialeah, FL) was prepared and the concentration. determined using a Coulter Counter Model ZM with a Channelizer. Four dilutions of the stock bead solution were prepared and each counted in quadruplicate (Fig. 2), yielding a linear dilution curve. From these data the concentration of the stock solution was determined to be 16.8 ± 0.2 X 106 beads/ml (mean ± SD). Vol. 53, No.5, May Volume of Beads Added to Diluent (Ill) Figure 2 Standardization of latex bead solution using the Coulter Counter. A solution of 5!'ill latex beads was prepared and its concentration determined by electronic particle counting. Samples were analyzed by diluting a volume of the bead solution in 20 ml and counting 0.5 ml in the Coulter Counter. Average bead counts obtained from four dilutions of the stock solution are shown, each point representing the mean of four readings (SD error bars are not shown because of their small magnitude). The line drawn through the points was computed from a linear regression analysis. Ginsburg and Armant Reliability of semen analysis chambers 883

3 was assumed for calculation of bead concentrations in the Micro-Cell chamber, whereas the Makler chamber was assumed to have a thickness of 10 #LID. The Makler chamber was loaded with only 5 #LL to prevent overfilling artifacts. Hemocytometric counting was performed using the smallest (50 X 50 #Lm) squares in the central area of the chamber, assuming a depth of 100 #LID. Each time a chamber was loaded, four fields were counted. The mean count from these four fields was used to calculate the density of the bead solution. Between loadings, Makler chambers and hemocytometers were rinsed thoroughly with distilled water, lightly dried with lens paper, and dusted with compressed gas. Automated Motion Analysis Videomicrographic semen analysis was performed using the Hamilton-Thorn IJTM-2030 Motility Analyzer (Hamilton-Thorn Research, Danvers, MA) employing internal dark field illumination. Semen samples from seven men were obtained by masturbation into sterile plastic containers and allowed to liquify at room temperature. Makler and Micro-Cell chambers were loaded four times with undiluted semen from each sample, and analyzed (at a frequency of 18.6 Hz and sampling interval of 0.32 seconds) at 37 c. Quadruplicate measurements of sperm concentration, percent motility, and mean curvilinear velocity were obtained for each of the seven ejaculates in each chamber and averages were computed. At least 225 motile sperm were counted and analyzed from each sample. One of the Micro-Cell chambers from each ejaculate was reanalyzed as above 30 and 60 minutes after its initial loading to determine the viability of sperm during prolonged exposure to the chamber. Chambers were maintained over this time period at room temperature. Statistical Analysis Bead concentrations obtained for the three chambers were compared using repeated measures analysis of variance (ANOVA) and the Scheffe Posthoc test. Bartlett and Cochrans tests were employed for assessing homogeneity of variances. Mean sperm concentration, percent motility, and track velocity were compared between Micro-Cell and Makler chambers using paired t-tests. Investigation of the effect of time in the Micro-Cell chamber on sperm viability was carried out using Table 1 Latex Bead Concentrations Determined Manually using Micro-Cell, Makler, and Hemocytometer Chambers Bead concentration a Micro-Cell Makler Hemocytometer Xl0 6 /ml ± 1.6b 27.6 ± 2~b,c 19.2 ± 1.3b a Values represent the mean of counts obtained from four different fields for 10 Micro-cell, Makler, or hemocytometer chamber loadings. b Values are means± SD. Mean concentration significantly higher (P < 0.001) than Micro-cell; Hemocytometer, and Coulter Counter determinations by Scheffe Posthoc test. ANOV A on motility and velocity data normalized to initial measurements. All statistical tests were performed with the BMDP software package. 6 RESULTS Accuracy of the Micro-Cell Chamber Determined by Bead Concentration Measurements To compare the accuracy and reproducibility of manual concentration determinations, the standardized bead solution was counted and the density computed using three chambers: the Micro-Cell chamber, Makler Chamber, and hemocytometer (Table 1). The bead concentrations calculated from data obtained using the hemocytometer and Micro-Cell chamber did not differ significantly from each other or the value obtained by electronic particle counting. However, the average concentration obtained using the Makler chamber was 62% higher than that obtained using _the Coulter Counter, a statistically significant difference (P < 0.01). Both Bartlett and Cochran's tests for homogeneity of variances confirmed that the variance associated with Makler chamber concentration measurements was significantly higher than the variance of the Micro-Cell and hematocytometer chambers (Bartlett test, P < 0.02; Cochran's test, P < 0.05). 884 Ginsburg and Armant Reliability of semen analysis chambers Fertility and Sterility

4 Table 2 Computer-Assisted Sperm Movement Analysis Data Obtained with the Micro-Cell and Makler Chambers on Seven Semen Samples Concentration Mean velocity Motility Sample Micro-Cell Makler Micro-Cell Makler Micro-Cell Makler spfmlx 10-6 p.m/s % 22.3 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 3.4 Mean±SD 55.4 ± 26.2b 79.4 ± 44.4b 42.0± 8.4c 48.0 ± 12.3c 57.4 ± 26.6d 58.7 ± 24.3d Data for each sample represent the mean± SD of quadruplicate loadings of each chamber. b Mean concentrations were significantly different by paired t-test, P = Motion Analysis of Live Spermatozoa in the Micro-Cell Chamber Automated semen analysis of seven samples comparing concentration, mean velocity, and motility data for the Micro-Cell and Makler chambers are shown in Table 2. These results demonstrate a significant (P < 0.02) difference between Makler and Micro-Cell chambers in sperm concentration, with the Makler chamber values averaging 43% higher than Micro-Cell values. In addition, Makler chamber mean sperm velocities were 14% higher than Micro-Cell velocities, again a statistically significant difference (P < 0.05). There was no difference in percent motility of semen samples between chambers. Viability of Sperm in the Micro-Cell Chamber Data on sperm motility and track velocity obtained 30 and 60 minutes after filling and initial analysis of the Micro-Cell chambers are shown in Table 3. Measurements at 30 and 60 minutes are c Mean velocities were significantly different by paired t-test, P= d Mean motilities were not significantly different by paired t test. expressed as a percentage of initial values. There was no significant difference in motility or mean velocity measurements at 30 minutes compared with initial measurements. At 60 minutes, the percentage of motile spermatozoa dropped significantly (mean decrease in motility 31%, P < 0.01) in the Micro-Cell chamber, although mean velocity remained essentially unchanged. DISCUSSION Determination of the Micro-Cell chambers' accuracy was based on a comparison between it and two other laboratory standards for cell counting, ie., the Coulter Countee 8 and the hemocytometer. 5 Using the latex bead size profile obtained with the Coulter Channelizer, it was possible to set the Coulter Counter threshold limits to ensure that all beads were counted whereas artifacts were excluded. Counts obtained with the Coulter instrument are therefore precise and accurate, as evi- Table 3 Normalized Motility and Mean Velocity as a Function of Time in the Micro-Cell Chamber Motility Mean velocity Sample Omin 30min 60min" Omin 30min 60min % % % p.mfs p.mfs p.mfs Mean±SD 85.3± ± ± ± 14.8 Significantly different from percent motility at 0 minutes, P = Vol. 53, No.5, May 1990 Ginsburg and Armant Reliability of semen analysis chambers 885

5 denced by their low variability compared with the counting chambers studied. Concentration measurements obtained with the Micro-Cell chamber were similar to those obtained by the Coulter Counter and the hemocytometer, with no significant difference in mean values detectable between these three counting methods. The fact that the variances of hemocytometer and Micro-Cell data were essentially equal confirms that the Micro-Cell chamber provides consistent, reliable measurements of concentration by manual bead counting. Contrary to these findings, the bead concentrations determined using the Makler chamber were significantly higher than those calculated from the other devices. Furthermore, the variance of the Makler chamber data was significantly higher. Overall, these results suggest that the measurements obtained with either the Micro-Cell or hemocytometer chambers are more accurate and precise than those obtained usirtg the Makler chamber. During CASMA, the Micro-Cell and Makler devices provided equivalent measurements of sperm percent motility. Using the optical system within the Hamilton-Thorne analyzer, there was no apparent problem tracking sperm related to the larger chamber volume as might have been anticipated. 4 However, concentration data differed significantly between chambers (Table 2). It was interesting to note that similar systematic overestimates of sperm density (between 40% and 60%) were seen both with beads and live sperm samples. This discrepancy between concentration determinations may be related to inaccurate Makler chamber depth, resulting in an actual volume that is larger than expected. 9 Such alteration of the 10 JLm gap might occur if the pillars supporting the cover glass were too high, if surfaces were not planar, if dust particles or dried semen were to raise the cover glass off of the pillars, or if the chamber were overfilled with semen, floating the cover glass. By design the Micro-Cell device has a larger chamber depth than the Makler chamber, which may account for the differences in sperm velocity that were observed (Table 2). The sperm head traverses a complex, three.dimensional path in space as it progresses. It is conceivable that the smaller Makler chamber depths could artificially constrain sperm head movements to two dimensions and increase the frequency of collisions with chamber surfaces, thereby increasing the apparent swimming speed. Larger chamber depths may thus provide better approximations of true movement characteristics. Alternatively, sperm movement in the Micro-Cell chamber may result in z-axis motion that cannot be resolved optically, thereby resulting in a shorter sperm track in the x-y plane and lower apparent velocities. These data emphasize the influence of chamber dimensions on the acquisition of reliable sperm concentration and motion data, since variations in these measurements can be expected when chambers of different design and depth are employed. Caution should therefore be used in comparing data among laboratories using various chambers. Biocompatability of the Micro-Cell chambers was tested by repeated videomicrographic motion analysis of samples over a 1 hour period. There was no apparent reduction in the percent motility or mean velocity of semen samples over 30 minutes, a period sufficient to perform most analytical procedures. The reduction in percent motility noted after 60 minutes is probably not indicative of a toxic effect on sperm survival, since mean velocity was maintained during this interval. This reduction in sperm motility may have been due to drying of the sample within the device. We conclude that the Micro-Cell chamber allows acquisition of statistically reliable information both in manual sperm counting and during digital processing of sperm trajectories for motion analysis. The chamber performed at least as well as others with regard to reproducibility and accuracy of results. Over 30 minutes, sperm samples showed no loss of viability based on stability of percent motility and mean velocity measurements. In addition, there are several distinct advantages inherent in the Micro-Cell device. Because it is preassembled and disposable, there is no potential for loss or breakage of delicate parts as the chamber is handled during washing, drying, and reassembly. The fixed cover slip would prevent alterations in chamber depth that could result in systematic measurement errors. User error during loading is eliminated since the chamber self-fills by capillary flow. Finally, since the chamber is not reused, medical laboratory personnel would reduce their exposure to and possible direct contact with human semen, which is considered a potentially infectious body fluid. Acknowledgments. The authors acknowledge with gratitude the expert assistance of Michael Kruger, M.A., for his help in statistical analysis, and the secretarial assistance of Ms. Debra Paduchowski in the preparation of the manuscript. 886 Ginsburg and Armant Reliability of semen analysis chambers Fertility and Sterility

6 REFERENCES 1. Knuth UA, Yeung C-H, Nieschlag E: Computerized semen analysis: objective measurement of semen characteristics is biased by subjective parameter setting. Fertil Steril48:118, Vantman D, Koukoulis G, Dennison L, Zinaman M, Sherins R: Computer-assisted semen analysis: evaluation of method and assessment of the influence of sperm concentration on linear velocity determination. Fertil Steril 49: 510, Ginsburg KA, Moghissi KS, Abel EL: Computer-assisted human semen analysis sampling error and reproducibility. J Artdrol 9:82, Makler A: A new chamber for rapid sperm count and motility estimation. Fertil Steril30:313, Mortimer D, Shu MA, Tan R: Standardization and quality control of sperm concentration and sperm motility counts in semen analysis. Hum Reprod 1:299, Dixon WJ, Brown MB, Enbleman L, Frane JW, Hill MA, Jennrich RI, Toporek JD: BMDP Statistical Software. Berkeley, University of California Press, Sundquist T, Fjallbrant B, Magnusson KE: Computeraided counting with the Coulter Counter of low numbers of spermatozoa in human semen. Int J Androl4:18, Brotherton J, Barnard G: Estimation of number, mean size and size distribution of human spermatozoa in oligospermia using a Coulter Counter. J Reprod Fertil40:341, Mortimer D, Shu MA, Tan R, Mortimer ST: A technical note on diluting semen for the haemocytometric determination of sperm concentration. Hum Reprod 4:166, 1989 Vol. 53, No.5, May 1990 Ginsburg and Armant Reliability of semen analysis chambers 887