- Email: [email protected]
chronic pelvic pain syndrome
ADULT UROLOGY
COUNTING LEUKOCYTES IN EXPRESSED PROSTATIC SECRETIONS FROM PATIENTS WITH CHRONIC PROSTATITIS/CHRONIC PELVIC PAIN SYNDROME JOHN N. KRIEGER, SUSAN O. ROSS, LESLIE A. DEUTSCH, THOMAS R. FRITSCHE, DONALD E. RILEY
AND
ABSTRACT Objectives. The evaluation of WBCs in EPS is recommended for classifying patients with chronic prostatitis/ chronic pelvic pain syndrome (CP/CPPS) but no agreement has been reached on the optimal method. We sought to determine the relationship between the expressed prostatic secretions (EPS) leukocyte (WBC) count per high-power field (evaluated by a more quantitative wet mount method and the traditional gram-stained smear method used in clinical microbiology laboratories) and the EPS WBC concentration to determine whether quantitative methods are necessary for accurate patient classification. Methods. EPS collected from 94 patients with CP/CPPS were evaluated by gram-stained smear, a standardized wet mount, and a hemocytometer method. Results. The gram-stained smear detected EPS WBCs in 21 (22%) of 94 subjects compared with 78 (83%) by the standardized wet mount and 57 (60%) by the hemocytometer method. The gram-stained EPS WBC count correlated poorly with the WBC concentration by hemocytometer (R2 ⫽ 0.051, P ⫽ 0.03). Although the standardized EPS WBC count correlated better with the concentration by hemocytometer, the correlation coefficient remained low (R2 ⫽ 0.244, P ⬍0.0001). Conclusions. The standardized wet mount proved superior to the gram-stained smear, but both methods lacked precision. Quantitative determination of the EPS WBC concentration by a counting chamber method proved to be the superior evaluation for research studies of CP/CPPS. UROLOGY 62: 30–34, 2003. © 2003 Elsevier Inc.
A
resurgence of interest in prostatitis has been stimulated by the National Institutes of Health consensus classification.1 This approach emphasizes the need to define the pathophysiology of chronic prostatitis/chronic pelvic pain syndrome (CP/CPPS), the most common and difficult category. Patients with CP/CPPS present with characteristic symptoms, but they have no evidence of a prostatic focus of bacteria causing urinary tract infection. Because current treatment is empirical and often ineffective, major research efforts have foThis study was supported in part by grant RO1-DK38955 from the National Institutes of Health, Bethesda, Maryland. From the Departments of Urology and Laboratory Medicine, University of Washington School of Medicine, Seattle, Washington Reprint requests: John N. Krieger, M.D., Urology Section, 112GU, Veterans Affairs Puget Sound Health Care System, 1660 South Columbian Way, University of Washington Campus Mail Box 358280, Seattle, WA 98108-1597 Submitted: December 19, 2002, accepted (with revisions): February 14, 2003
30
© 2003 ELSEVIER INC. ALL RIGHTS RESERVED
cused on developing better methods to diagnose and treat these patients. The presence of leukocytes (WBCs) in the expressed prostatic secretions (EPS) is a readily available clinical marker in patients with chronic prostatitis. Counting EPS WBCs has long been recommended for patient classification. In the traditional classification, patients with EPS WBCs but no evidence of bacterial infection of the prostate were considered to have nonbacterial prostatitis, and similar patients without EPS WBCs were considered to have prostatodynia.2 The consensus classification also recommends EPS WBC evaluation (in addition to WBC evaluation in the postmassage urine and seminal fluid) to distinguish patients with inflammatory CP/CPPS from those with noninflammatory CP/CPPS.1 The traditional method is to prepare an EPS wet mount and then count the WBCs “per high power” microscopic field (hpf). The specified cutpoint to distinguish patients with inflammation is often 5 0090-4295/03/$30.00 doi:10.1016/S0090-4295(03)00237-1
or 10 EPS WBCs per hpf.3–5 Many published studies contain few details about the EPS volume placed on the slide, the magnification used, or the number of microscopic fields examined. Because the traditional EPS WBC count per hpf has not proved reliable for directing diagnostic testing or treatment, many research groups are working to develop better markers, such as cytokine levels,6,7 the presence of autoimmunity,8 or measures of oxidative stress.9 Other investigators have recommended more quantitative methods to determine the WBC concentration in a known EPS volume.3,10 Our earlier study showed that significantly more cases of inflammation were detected by hemocytometer than by the traditional wet mount method.11 One common question is, “How does the EPS WBC concentration relate to traditional EPS WBC counts per hpf?” A related question is, “Is it necessary to use a counting chamber?” These are important issues, because the great majority of research studies continue to use the traditional EPS wet mount. The goals of this study were to determine the relationship between the EPS WBC count per hpf (evaluated by a more quantitative wet mount method and the traditional gram-stained smear method used in clinical microbiology laboratories) and the EPS WBC concentration to determine whether quantitative methods are necessary for accurate patient classification. MATERIAL AND METHODS SUBJECTS AND CLINICAL EVALUATION Patients with CP/CPPS were evaluated at the University of Washington Prostatitis Clinic in standardized fashion according to our published methods.11,12 The study followed institutional review board procedures. All patients had stopped taking antimicrobial and anti-inflammatory medications for at least 1 month before evaluation. Potential subjects were excluded if they had evidence of urethritis, acute or chronic bacterial prostatitis, significant structural or functional genitourinary abnormalities, or any consensus exclusion criterion.1,13,14
EPS PROCESSING EPS were collected by digital rectal massage into sterile, sealed, polypropylene containers. Containers were brought to the male fertility laboratory within 30 minutes of collection. The specimens were divided for evaluation in the male fertility laboratory and clinical microbiology laboratory. EPS samples with limited volumes (less than 30 L) were diluted with sterile, modified Tyrode’s salt solution (pH 7.4 to 7.6, containing 8 g NaCl, 0.020 g KCl, 0.021 g MgCl2 䡠6 H2O, and 2.38 g HEPES pKa 7.5 buffer per liter) to at least 30 L. Results were corrected for any dilution factor. The fertility laboratory evaluated EPS WBCs using the hemocytometer and standardized wet mount methods. The microbiology laboratory evaluated EPS WBCs using a traditional gram-stained smear, and then cultured the EPS following published methods.12 Each assay was done by technologists who were unaware of the results of the other assays. UROLOGY 62 (1), 2003
HEMOCYTOMETER METHOD
Following published methods, 10 L of EPS was loaded into an Improved Neubauer phase contrast counting chamber (0.01-mm cell depth) and allowed to settle for 2 to 5 minutes.11 WBCs were counted in the entire 25 square, central 1⫻1 mm grid or until at least 100 cells were counted. The mean cell count was corrected for the number of cells per 25 squares (0.1 mm3) and multiplied by 10 to calculate the mean EPS WBC concentration per cubic millimeter.
GRAM-STAINED SMEAR One drop of undiluted EPS was placed on a slide, air-dried, heat-fixed, gram-stained, and examined under low power to evaluate clumping. At least 25 high-power (400⫻) microscopic fields were evaluated to determine the average EPS WBC count per hpf. The EPS WBC count was reported as both a number per hpf and qualitatively (usual practice for clinical microbiology laboratories). The qualitative categories were none (no WBCs observed in 25 fields), occasional (fewer than 3 WBCs per 25 fields), 1⫹ (more than 3 WBCs per slide but less than 1 WBC per hpf), 2⫹ (1 to 4 WBCs per hpf), 3⫹ (5 to 30 WBCs per hpf), and 4⫹ (more than 30 WBCs per hpf).
STANDARDIZED WET MOUNT
Using an automatic micropipette, 5 L of EPS was placed on a glass slide and covered with a No. 1 cover glass 22 mm square. The slide was examined with phase contrast using a Labophot 2 microscope (Nikon, New York, New York). At least 25 fields were examined. WBCs were counted and averaged to determine the mean EPS WBC number per 400⫻ hpf using our published criteria.11
STATISTICAL ANALYSIS The data were evaluated initially by scatterplots. Least mean squares linear regression lines, Pearson’s correlation coefficients, and t-statistics were calculated using standard techniques. The performance of different assays on the same samples was compared by calculating McNemar’s chi-square statistics.
RESULTS SUBJECTS AND SPECIMENS Of 180 consecutive patients who provided EPS in our chronic prostatitis clinic, 94 (52%) met the study criteria and had all the evaluations. Potential subjects were excluded if their EPS were not evaluated in both laboratories because of delays in transportation or were evaluated late in the afternoon after closure of the fertility laboratory (69 patients, 38%), or because they did not meet the criteria for CP/CPPS (17 patients, 10%). Only the first evaluation was included for subjects who provided multiple samples. COMPARISON OF HEMOCYTOMETER AND GRAM-STAINED SMEAR METHODS By the hemocytometer method, 57 (60%) of the 94 subjects had EPS WBCs (range 0 to 106,000 WBCs/mm3). For all 94 specimens, the median EPS WBCs concentration was 215/mm3 (mean ⫾ standard deviation 3791 ⫾ 12,000/mm3). The 57 subjects with detectable EPS WBCs had a median 31
FIGURE 1. Gram-stained smear compared with hemocytometer method for evaluating WBCs in EPS of 94 patients with CP/CPPS. See text for delineation of EPS WBC categories for wet mount method. Individual bars indicate mean EPS concentration determined by hemocytometer method for each gram-stained smear category. Error bars indicate standard deviation. No patients had 4⫹ EPS WBCs by traditional gram-stained smear. EPS WBC count correlated with WBC concentration by hemocytometer (P ⫽ 0.03), but the correlation coefficient was low (R2 ⫽ 0.051); no correlation was found among the 21 subjects with any WBCs detected by gram-stained smear (R2 ⫽ 0.007, P ⫽ 0.72).
of 1750 EPS WBCs/mm3 (mean 6251 ⫾ 14,949 WBCs/mm3). The gram-stained smear detected EPS WBCs in only 21 (22%) of the 94 subjects compared with 57 (60%) by the hemocytometer method (McNemar’s chi-square ⫽ 34.0, P ⬍0.001). The 21 subjects with EPS WBCs by the gram-stained smear included 8 with occasional WBCs, 8 with 1⫹ WBCs, 1 with 2⫹ WBCs, 4 with 3⫹ WBCs, and none with 4⫹ WBCs. For the overall group, the median was 0 EPS WBCs per hpf (range 0 to 20, mean 0.7 ⫾ 3.0). For the 21 subjects with EPS WBCs, the median was 0.5 WBC per hpf (mean 3.1 ⫾ 5.8). The gram-stained smear EPS WBC count correlated with the WBC concentration by hemocytometer (F ⫽ 4.986, P ⫽ 0.03), but the correlation coefficient was low (R2 ⫽ 0.051). Among the 21 subjects with any WBCs detected by the wet mount method, no correlation was present between the EPS WBC concentration by the hemocytometer method or the EPS WBC count per hpf (R2 ⫽ 0.007, F ⫽ 0.138, P ⫽ 0.715, Fig. 1). COMPARISON OF HEMOCYTOMETER AND STANDARDIZED WET MOUNT METHODS The standardized wet mount detected EPS WBCs in 78 (83%) of 94 subjects compared with 57 (60%) of 94 subjects who had WBCs by the hemocytometer method (McNemar’s chi-square ⫽ 14.82, P ⬍0.001). This was also greater than the 21 subjects with WBCs by the traditional wet mount 32
FIGURE 2. “Standardized” wet mount method compared with hemocytometer method for evaluating WBCs in EPS of 94 patients with CP/CPPS. “Standardized” wet mount method uses a standard amount of EPS (5 L delivered by an automatic pipette). Individual bars indicate mean EPS concentration determined by hemocytometer method for each wet mount category. Error bars indicate standard deviation. “Improved” EPS WBC count correlated with WBC concentration by hemocytometer (P ⬍0.0001), but the correlation coefficient was low (R2 ⫽ 0.244).
method (McNemar’s chi-square ⫽ 53.15, P ⬍0.001). The 78 subjects with EPS WBCs by the standardized wet mount included 54 with fewer than 4 WBCs per hpf, 10 with 5 to 9 WBCs per hpf, and 17 with more than 10 WBCs per hpf. For the overall group, the median was 1 EPS WBC per hpf (range 0 to 60, mean 5.6 ⫾ 10.5). Among the 78 subjects with WBCs by the standardized wet mount method, the median was 2 EPS WBCs per hpf (mean 6.8 ⫾ 11.2). The standardized EPS WBC count correlated with the WBC concentration by hemocytometer (F ⫽ 29.666, P ⬍0.0001), but the correlation coefficient was low (R2 ⫽ 0.244). Among the 78 subjects with any WBCs by the wet mount method, a statistically significant correlation was noted with the EPS WBC concentration by the hemocytometer method (F ⫽ 22.520, P ⬍0.0001), but the correlation coefficient was low (R2 ⫽ 0.229, Fig. 2). COMMENT The results of our study help explain a number of confusing observations. Traditional EPS WBC counts correlate poorly, or not at all, with the symptoms15 or response to treatment of patients with CP/CPPS9 or with other markers, such as EPS cytokine levels6,7 or oxidative stress.9 Such findings might be predicted given the low correlation (R2 ⫽ 0.05) between the EPS WBC count per hpf by the gram-stained smear and the EPS WBC concentration per cubic millimeter. No statistically significant correlation was found between these meaUROLOGY 62 (1), 2003
sures in the 21 patients with EPS WBCs detected by the gram-stained smear traditionally used by clinical microbiology laboratories (Fig. 1). These findings support the statement of Nickel et al.13 that the consensus classification, “suffers from the limitations imposed by an inadequate understanding of the relevance of WBCs, lack of standardization of WBC investigation techniques, and lack of comparable cutpoints for ‘elevated numbers’.” The traditional stained smear or wet mount methods using variable EPS volumes appear inadequate to address these limitations. These approaches might distinguish populations without WBCs from those with high numbers (more than 20 EPS WBCs per hpf), but such examinations offer limited discrimination in the critical range of 1 to 20 WBCs used to distinguish patients with and without inflammation in most studies. The importance of WBC evaluation was illustrated in a recent study of 488 patients, including 397 who had EPS examined by the traditional wet mount method.15 Using cutpoints of more than 5 WBCs per hpf or more than 10 WBCs per hpf, no correlation was found between EPS inflammation and patients’ symptoms. This may reflect the lack of precision of the wet mount method over the range of WBC counts per hpf used to distinguish patients with and without inflammation.4,15 It is likely that EPS WBC results are even less precise in multicenter studies involving technical differences in wet mount preparation, magnification, and methods used to examine the slides. Thus, the conclusion that EPS WBCs do not correlate with patients’ symptoms or with bacterial counts requires verification by more quantitative methods. We will investigate such possible correlations in future studies. Our results agree with previous studies emphasizing the limitations of the EPS wet mount method. Using counting chambers, Anderson and Weller3 proposed 1000 WBCs/mm3 and Thin and Simmons10 proposed 500 WBCs/mm3 as the critical values defining EPS inflammation. The latter study concluded, “Leukocyte counts of the EPS appeared to be more sensitive than simple microscopic estimation.” Stamey16 stated that, “quantitation [with a counting chamber] . . . is clearly better.” Our earlier study in a single laboratory investigated factors limiting the reliability of the EPS wet mount.11 Also, we emphasized the value of staining procedures to distinguish WBCs from other EPS cells, such as degenerating cells, isolated nuclei, glandular cells from the prostate or seminal vesicles, squamous epithelium from the urethra or skin, and transitional epithelium. The results of the current study, in a different cohort, provide a direct comparison of clinical samples evaluated in two laboratories in blinded fashion. Again, we conUROLOGY 62 (1), 2003
firm the poor reliability of the traditional EPS WBC evaluation using a variable EPS volume. Because the wet mount remains the method of choice for most researchers and clinicians, we tried to increase the accuracy of this time-honored examination. Using a standard EPS volume proved superior to the traditional gram-stained smear. However, the correlation with the quantitative hemocytometer concentration remained low (R2 ⫽ 0.229, Fig. 2). The standardized wet mount detected EPS WBCs in 83% of the 94 subjects compared with 22% who had WBCs by the gramstained method (P ⬍0.001) and 60% who had WBCs by the hemocytometer method (P ⬍0.001). Thus, neither the standardized wet mount nor the traditional stained smear proved an adequate replacement for the quantitative chamber count. This illustrates the increased possibility of systematic artifacts that could be introduced into prostatitis studies by the methods chosen for WBC evaluation. In most clinical laboratories, a counting chamber such as the hemocytometer is preferred for quantification. For example, several devices using controlled chamber sizes are used in andrology laboratories.11 Counting chambers are also recommended for diagnosis of urinary tract infection and inflammation.17 The critical question is whether different diagnostic evaluations or therapies should be recommended for patients with and without inflammation. Recent observations suggest that there may be important differences between inflammatory and noninflammatory CP/CPPS. Patients with inflammatory CP/CPPS were significantly more likely to have prostatic bacterial DNA,12 different symptom profiles,18 greater intraprostatic pressures,19 and higher endotoxin levels in their EPS and/or postmassage urine.20 Until the clinical significance of EPS inflammation is resolved, it is desirable to provide accurate EPS WBC concentrations in research studies using a counting chamber. It is also important to determine the range of EPS WBC concentrations in normal men using optimal methods. Use of a counting chamber is a well-accepted and readily available clinical laboratory procedure that can easily be adapted to office practice. CONCLUSIONS This study addressed three questions. First, how does the EPS WBC concentration relate to traditional EPS WBC counts per hpf? There was a relationship, but this correlation was not strong (R2 ⫽ 0.051), and no correlation was found over the usual range of WBC counts used to distinguish inflammatory from noninflammatory CP/CPPS. Second, should a counting chamber be recom33
mended for research studies? Our findings suggest that the hemocytometer method, or other quantitative methods, should be recommended. Finally, can we standardize the EPS wet mount to improve results? The standardized wet mount proved superior to the traditional gram-stained smear, but this method was imprecise for calculating EPS WBC concentrations (R2 ⫽ 0.244). The counting chamber method can be adapted to most clinical settings, providing accurate and rapid results with substantially less technical difficulty than use of alternative markers of disease activity, such as cytokine levels or oxidative stress. It would appear prudent to include quantitative determination of the EPS WBC concentration as a recommended evaluation in research studies of CP/CPPS. REFERENCES 1. Krieger JN, Nyberg L Jr, and Nickel JC: NIH consensus definition and classification of prostatitis. JAMA 282: 236 – 237, 1999. 2. Drach GW, Meares EM Jr, Fair WR, et al: Classification of benign disease associated with prostatic pain: prostatitis or prostatodynia? J Urol 120: 266, 1978. 3. Anderson RU, and Weller C: Prostatic secretion leukocyte studies in non-bacterial prostatitis (prostatosis). J Urol 121: 292–294, 1979. 4. Wright ET, Chmiel JS, Grayhack JT, et al: Prostatic fluid inflammation in prostatitis. J Urol 152: 2300 –2303, 1994. 5. Schaeffer AJ, Wendel EF, Dunn JK, et al: Prevalence and significance of prostatic inflammation. J Urol 125: 215–219, 1981. 6. Alexander RB, Ponniah S, Hasday J, et al: Elevated levels of proinflammatory cytokines in the semen of patients with chronic prostatitis/chronic pelvic pain syndrome. Urology 52: 744 –749, 1998. 7. Hochreiter WW, Nadler RB, Koch AE, et al: Evaluation of the cytokines interleukin 8 and epithelial neutrophil activating peptide 78 as indicators of inflammation in prostatic secretions. Urology 56: 1025–1029, 2000.
34
8. Alexander RB, Brady F, and Ponniah S: Autoimmune prostatitis: evidence of T cell reactivity with normal prostatic proteins. Urology 50: 893–899, 1997. 9. Shahed AR, and Shoskes DA: Oxidative stress in prostatic fluid of patients with chronic pelvic pain syndrome: correlation with gram positive bacterial growth and treatment response. J Androl 21: 669 –675, 2000. 10. Thin RN, and Simmons PD: Chronic bacterial and nonbacterial prostatitis. Br J Urol 55: 513–518, 1983. 11. Muller CH, Berger RE, Mohr LE, et al: Comparison of microscopic methods for detecting inflammation in expressed prostatic secretions. J Urol 166: 2518 –2524, 2001. 12. Krieger JN, Riley DE, Roberts MC, et al: Prokaryotic DNA sequences in patients with chronic idiopathic prostatitis. J Clin Microbiol 34: 3120 –3128, 1996. 13. Nickel JC, Nyberg LM, and Hennenfent M: Research guidelines for chronic prostatitis: consensus report from the first National Institutes of Health International Prostatitis Collaborative Network. Urology 54: 229 –233, 1999. 14. Schaeffer AJ, Landis JR, Knauss JS, et al: Demographic and clinical characteristics of men with chronic prostatitis: the National Institutes of Health Chronic Prostatitis Cohort Study. J Urol 168: 593–598, 2002. 15. Schaeffer AJ, Knauss JS, Landis JR, et al: Leukocyte and bacterial counts do not correlate with severity of symptoms in men with chronic prostatitis: the National Institutes of Health Chronic Prostatitis Cohort Study. J Urol 168: 1048 –1053, 2002. 16. Stamey TA: Prostatitis. J R Soc Med 74: 22–40, 1981. 17. Stamm WE: Measurement of pyuria and its relation to bacteriuria. Am J Med 75: 53–58, 1983. 18. Krieger JN, Ross SO, Penson DA, et al: Symptoms of inflammatory versus non-inflammatory CP/CPPS. Urology 60: 959 –963, 2002. 19. Mehik A, Hellstrom P, Nickel JC, et al: The chronic prostatitis-chronic pelvic pain syndrome can be characterized by prostatic tissue pressure measurements. J Urol 167: 137– 140, 2002. 20. Li LJ, Shen ZJ, Lu YL, et al: The value of endotoxin concentrations in expressed prostatic secretions for the diagnosis and classification of chronic prostatitis. BJU Int 88: 536 – 539, 2001.
UROLOGY 62 (1), 2003
COUNTING LEUKOCYTES IN EXPRESSED PROSTATIC SECRETIONS FROM PATIENTS WITH CHRONIC PROSTATITIS/CHRONIC PELVIC PAIN SYNDROME JOHN N. KRIEGER, SUSAN O. ROSS, LESLIE A. DEUTSCH, THOMAS R. FRITSCHE, DONALD E. RILEY
AND
ABSTRACT Objectives. The evaluation of WBCs in EPS is recommended for classifying patients with chronic prostatitis/ chronic pelvic pain syndrome (CP/CPPS) but no agreement has been reached on the optimal method. We sought to determine the relationship between the expressed prostatic secretions (EPS) leukocyte (WBC) count per high-power field (evaluated by a more quantitative wet mount method and the traditional gram-stained smear method used in clinical microbiology laboratories) and the EPS WBC concentration to determine whether quantitative methods are necessary for accurate patient classification. Methods. EPS collected from 94 patients with CP/CPPS were evaluated by gram-stained smear, a standardized wet mount, and a hemocytometer method. Results. The gram-stained smear detected EPS WBCs in 21 (22%) of 94 subjects compared with 78 (83%) by the standardized wet mount and 57 (60%) by the hemocytometer method. The gram-stained EPS WBC count correlated poorly with the WBC concentration by hemocytometer (R2 ⫽ 0.051, P ⫽ 0.03). Although the standardized EPS WBC count correlated better with the concentration by hemocytometer, the correlation coefficient remained low (R2 ⫽ 0.244, P ⬍0.0001). Conclusions. The standardized wet mount proved superior to the gram-stained smear, but both methods lacked precision. Quantitative determination of the EPS WBC concentration by a counting chamber method proved to be the superior evaluation for research studies of CP/CPPS. UROLOGY 62: 30–34, 2003. © 2003 Elsevier Inc.
A
resurgence of interest in prostatitis has been stimulated by the National Institutes of Health consensus classification.1 This approach emphasizes the need to define the pathophysiology of chronic prostatitis/chronic pelvic pain syndrome (CP/CPPS), the most common and difficult category. Patients with CP/CPPS present with characteristic symptoms, but they have no evidence of a prostatic focus of bacteria causing urinary tract infection. Because current treatment is empirical and often ineffective, major research efforts have foThis study was supported in part by grant RO1-DK38955 from the National Institutes of Health, Bethesda, Maryland. From the Departments of Urology and Laboratory Medicine, University of Washington School of Medicine, Seattle, Washington Reprint requests: John N. Krieger, M.D., Urology Section, 112GU, Veterans Affairs Puget Sound Health Care System, 1660 South Columbian Way, University of Washington Campus Mail Box 358280, Seattle, WA 98108-1597 Submitted: December 19, 2002, accepted (with revisions): February 14, 2003
30
© 2003 ELSEVIER INC. ALL RIGHTS RESERVED
cused on developing better methods to diagnose and treat these patients. The presence of leukocytes (WBCs) in the expressed prostatic secretions (EPS) is a readily available clinical marker in patients with chronic prostatitis. Counting EPS WBCs has long been recommended for patient classification. In the traditional classification, patients with EPS WBCs but no evidence of bacterial infection of the prostate were considered to have nonbacterial prostatitis, and similar patients without EPS WBCs were considered to have prostatodynia.2 The consensus classification also recommends EPS WBC evaluation (in addition to WBC evaluation in the postmassage urine and seminal fluid) to distinguish patients with inflammatory CP/CPPS from those with noninflammatory CP/CPPS.1 The traditional method is to prepare an EPS wet mount and then count the WBCs “per high power” microscopic field (hpf). The specified cutpoint to distinguish patients with inflammation is often 5 0090-4295/03/$30.00 doi:10.1016/S0090-4295(03)00237-1
or 10 EPS WBCs per hpf.3–5 Many published studies contain few details about the EPS volume placed on the slide, the magnification used, or the number of microscopic fields examined. Because the traditional EPS WBC count per hpf has not proved reliable for directing diagnostic testing or treatment, many research groups are working to develop better markers, such as cytokine levels,6,7 the presence of autoimmunity,8 or measures of oxidative stress.9 Other investigators have recommended more quantitative methods to determine the WBC concentration in a known EPS volume.3,10 Our earlier study showed that significantly more cases of inflammation were detected by hemocytometer than by the traditional wet mount method.11 One common question is, “How does the EPS WBC concentration relate to traditional EPS WBC counts per hpf?” A related question is, “Is it necessary to use a counting chamber?” These are important issues, because the great majority of research studies continue to use the traditional EPS wet mount. The goals of this study were to determine the relationship between the EPS WBC count per hpf (evaluated by a more quantitative wet mount method and the traditional gram-stained smear method used in clinical microbiology laboratories) and the EPS WBC concentration to determine whether quantitative methods are necessary for accurate patient classification. MATERIAL AND METHODS SUBJECTS AND CLINICAL EVALUATION Patients with CP/CPPS were evaluated at the University of Washington Prostatitis Clinic in standardized fashion according to our published methods.11,12 The study followed institutional review board procedures. All patients had stopped taking antimicrobial and anti-inflammatory medications for at least 1 month before evaluation. Potential subjects were excluded if they had evidence of urethritis, acute or chronic bacterial prostatitis, significant structural or functional genitourinary abnormalities, or any consensus exclusion criterion.1,13,14
EPS PROCESSING EPS were collected by digital rectal massage into sterile, sealed, polypropylene containers. Containers were brought to the male fertility laboratory within 30 minutes of collection. The specimens were divided for evaluation in the male fertility laboratory and clinical microbiology laboratory. EPS samples with limited volumes (less than 30 L) were diluted with sterile, modified Tyrode’s salt solution (pH 7.4 to 7.6, containing 8 g NaCl, 0.020 g KCl, 0.021 g MgCl2 䡠6 H2O, and 2.38 g HEPES pKa 7.5 buffer per liter) to at least 30 L. Results were corrected for any dilution factor. The fertility laboratory evaluated EPS WBCs using the hemocytometer and standardized wet mount methods. The microbiology laboratory evaluated EPS WBCs using a traditional gram-stained smear, and then cultured the EPS following published methods.12 Each assay was done by technologists who were unaware of the results of the other assays. UROLOGY 62 (1), 2003
HEMOCYTOMETER METHOD
Following published methods, 10 L of EPS was loaded into an Improved Neubauer phase contrast counting chamber (0.01-mm cell depth) and allowed to settle for 2 to 5 minutes.11 WBCs were counted in the entire 25 square, central 1⫻1 mm grid or until at least 100 cells were counted. The mean cell count was corrected for the number of cells per 25 squares (0.1 mm3) and multiplied by 10 to calculate the mean EPS WBC concentration per cubic millimeter.
GRAM-STAINED SMEAR One drop of undiluted EPS was placed on a slide, air-dried, heat-fixed, gram-stained, and examined under low power to evaluate clumping. At least 25 high-power (400⫻) microscopic fields were evaluated to determine the average EPS WBC count per hpf. The EPS WBC count was reported as both a number per hpf and qualitatively (usual practice for clinical microbiology laboratories). The qualitative categories were none (no WBCs observed in 25 fields), occasional (fewer than 3 WBCs per 25 fields), 1⫹ (more than 3 WBCs per slide but less than 1 WBC per hpf), 2⫹ (1 to 4 WBCs per hpf), 3⫹ (5 to 30 WBCs per hpf), and 4⫹ (more than 30 WBCs per hpf).
STANDARDIZED WET MOUNT
Using an automatic micropipette, 5 L of EPS was placed on a glass slide and covered with a No. 1 cover glass 22 mm square. The slide was examined with phase contrast using a Labophot 2 microscope (Nikon, New York, New York). At least 25 fields were examined. WBCs were counted and averaged to determine the mean EPS WBC number per 400⫻ hpf using our published criteria.11
STATISTICAL ANALYSIS The data were evaluated initially by scatterplots. Least mean squares linear regression lines, Pearson’s correlation coefficients, and t-statistics were calculated using standard techniques. The performance of different assays on the same samples was compared by calculating McNemar’s chi-square statistics.
RESULTS SUBJECTS AND SPECIMENS Of 180 consecutive patients who provided EPS in our chronic prostatitis clinic, 94 (52%) met the study criteria and had all the evaluations. Potential subjects were excluded if their EPS were not evaluated in both laboratories because of delays in transportation or were evaluated late in the afternoon after closure of the fertility laboratory (69 patients, 38%), or because they did not meet the criteria for CP/CPPS (17 patients, 10%). Only the first evaluation was included for subjects who provided multiple samples. COMPARISON OF HEMOCYTOMETER AND GRAM-STAINED SMEAR METHODS By the hemocytometer method, 57 (60%) of the 94 subjects had EPS WBCs (range 0 to 106,000 WBCs/mm3). For all 94 specimens, the median EPS WBCs concentration was 215/mm3 (mean ⫾ standard deviation 3791 ⫾ 12,000/mm3). The 57 subjects with detectable EPS WBCs had a median 31
FIGURE 1. Gram-stained smear compared with hemocytometer method for evaluating WBCs in EPS of 94 patients with CP/CPPS. See text for delineation of EPS WBC categories for wet mount method. Individual bars indicate mean EPS concentration determined by hemocytometer method for each gram-stained smear category. Error bars indicate standard deviation. No patients had 4⫹ EPS WBCs by traditional gram-stained smear. EPS WBC count correlated with WBC concentration by hemocytometer (P ⫽ 0.03), but the correlation coefficient was low (R2 ⫽ 0.051); no correlation was found among the 21 subjects with any WBCs detected by gram-stained smear (R2 ⫽ 0.007, P ⫽ 0.72).
of 1750 EPS WBCs/mm3 (mean 6251 ⫾ 14,949 WBCs/mm3). The gram-stained smear detected EPS WBCs in only 21 (22%) of the 94 subjects compared with 57 (60%) by the hemocytometer method (McNemar’s chi-square ⫽ 34.0, P ⬍0.001). The 21 subjects with EPS WBCs by the gram-stained smear included 8 with occasional WBCs, 8 with 1⫹ WBCs, 1 with 2⫹ WBCs, 4 with 3⫹ WBCs, and none with 4⫹ WBCs. For the overall group, the median was 0 EPS WBCs per hpf (range 0 to 20, mean 0.7 ⫾ 3.0). For the 21 subjects with EPS WBCs, the median was 0.5 WBC per hpf (mean 3.1 ⫾ 5.8). The gram-stained smear EPS WBC count correlated with the WBC concentration by hemocytometer (F ⫽ 4.986, P ⫽ 0.03), but the correlation coefficient was low (R2 ⫽ 0.051). Among the 21 subjects with any WBCs detected by the wet mount method, no correlation was present between the EPS WBC concentration by the hemocytometer method or the EPS WBC count per hpf (R2 ⫽ 0.007, F ⫽ 0.138, P ⫽ 0.715, Fig. 1). COMPARISON OF HEMOCYTOMETER AND STANDARDIZED WET MOUNT METHODS The standardized wet mount detected EPS WBCs in 78 (83%) of 94 subjects compared with 57 (60%) of 94 subjects who had WBCs by the hemocytometer method (McNemar’s chi-square ⫽ 14.82, P ⬍0.001). This was also greater than the 21 subjects with WBCs by the traditional wet mount 32
FIGURE 2. “Standardized” wet mount method compared with hemocytometer method for evaluating WBCs in EPS of 94 patients with CP/CPPS. “Standardized” wet mount method uses a standard amount of EPS (5 L delivered by an automatic pipette). Individual bars indicate mean EPS concentration determined by hemocytometer method for each wet mount category. Error bars indicate standard deviation. “Improved” EPS WBC count correlated with WBC concentration by hemocytometer (P ⬍0.0001), but the correlation coefficient was low (R2 ⫽ 0.244).
method (McNemar’s chi-square ⫽ 53.15, P ⬍0.001). The 78 subjects with EPS WBCs by the standardized wet mount included 54 with fewer than 4 WBCs per hpf, 10 with 5 to 9 WBCs per hpf, and 17 with more than 10 WBCs per hpf. For the overall group, the median was 1 EPS WBC per hpf (range 0 to 60, mean 5.6 ⫾ 10.5). Among the 78 subjects with WBCs by the standardized wet mount method, the median was 2 EPS WBCs per hpf (mean 6.8 ⫾ 11.2). The standardized EPS WBC count correlated with the WBC concentration by hemocytometer (F ⫽ 29.666, P ⬍0.0001), but the correlation coefficient was low (R2 ⫽ 0.244). Among the 78 subjects with any WBCs by the wet mount method, a statistically significant correlation was noted with the EPS WBC concentration by the hemocytometer method (F ⫽ 22.520, P ⬍0.0001), but the correlation coefficient was low (R2 ⫽ 0.229, Fig. 2). COMMENT The results of our study help explain a number of confusing observations. Traditional EPS WBC counts correlate poorly, or not at all, with the symptoms15 or response to treatment of patients with CP/CPPS9 or with other markers, such as EPS cytokine levels6,7 or oxidative stress.9 Such findings might be predicted given the low correlation (R2 ⫽ 0.05) between the EPS WBC count per hpf by the gram-stained smear and the EPS WBC concentration per cubic millimeter. No statistically significant correlation was found between these meaUROLOGY 62 (1), 2003
sures in the 21 patients with EPS WBCs detected by the gram-stained smear traditionally used by clinical microbiology laboratories (Fig. 1). These findings support the statement of Nickel et al.13 that the consensus classification, “suffers from the limitations imposed by an inadequate understanding of the relevance of WBCs, lack of standardization of WBC investigation techniques, and lack of comparable cutpoints for ‘elevated numbers’.” The traditional stained smear or wet mount methods using variable EPS volumes appear inadequate to address these limitations. These approaches might distinguish populations without WBCs from those with high numbers (more than 20 EPS WBCs per hpf), but such examinations offer limited discrimination in the critical range of 1 to 20 WBCs used to distinguish patients with and without inflammation in most studies. The importance of WBC evaluation was illustrated in a recent study of 488 patients, including 397 who had EPS examined by the traditional wet mount method.15 Using cutpoints of more than 5 WBCs per hpf or more than 10 WBCs per hpf, no correlation was found between EPS inflammation and patients’ symptoms. This may reflect the lack of precision of the wet mount method over the range of WBC counts per hpf used to distinguish patients with and without inflammation.4,15 It is likely that EPS WBC results are even less precise in multicenter studies involving technical differences in wet mount preparation, magnification, and methods used to examine the slides. Thus, the conclusion that EPS WBCs do not correlate with patients’ symptoms or with bacterial counts requires verification by more quantitative methods. We will investigate such possible correlations in future studies. Our results agree with previous studies emphasizing the limitations of the EPS wet mount method. Using counting chambers, Anderson and Weller3 proposed 1000 WBCs/mm3 and Thin and Simmons10 proposed 500 WBCs/mm3 as the critical values defining EPS inflammation. The latter study concluded, “Leukocyte counts of the EPS appeared to be more sensitive than simple microscopic estimation.” Stamey16 stated that, “quantitation [with a counting chamber] . . . is clearly better.” Our earlier study in a single laboratory investigated factors limiting the reliability of the EPS wet mount.11 Also, we emphasized the value of staining procedures to distinguish WBCs from other EPS cells, such as degenerating cells, isolated nuclei, glandular cells from the prostate or seminal vesicles, squamous epithelium from the urethra or skin, and transitional epithelium. The results of the current study, in a different cohort, provide a direct comparison of clinical samples evaluated in two laboratories in blinded fashion. Again, we conUROLOGY 62 (1), 2003
firm the poor reliability of the traditional EPS WBC evaluation using a variable EPS volume. Because the wet mount remains the method of choice for most researchers and clinicians, we tried to increase the accuracy of this time-honored examination. Using a standard EPS volume proved superior to the traditional gram-stained smear. However, the correlation with the quantitative hemocytometer concentration remained low (R2 ⫽ 0.229, Fig. 2). The standardized wet mount detected EPS WBCs in 83% of the 94 subjects compared with 22% who had WBCs by the gramstained method (P ⬍0.001) and 60% who had WBCs by the hemocytometer method (P ⬍0.001). Thus, neither the standardized wet mount nor the traditional stained smear proved an adequate replacement for the quantitative chamber count. This illustrates the increased possibility of systematic artifacts that could be introduced into prostatitis studies by the methods chosen for WBC evaluation. In most clinical laboratories, a counting chamber such as the hemocytometer is preferred for quantification. For example, several devices using controlled chamber sizes are used in andrology laboratories.11 Counting chambers are also recommended for diagnosis of urinary tract infection and inflammation.17 The critical question is whether different diagnostic evaluations or therapies should be recommended for patients with and without inflammation. Recent observations suggest that there may be important differences between inflammatory and noninflammatory CP/CPPS. Patients with inflammatory CP/CPPS were significantly more likely to have prostatic bacterial DNA,12 different symptom profiles,18 greater intraprostatic pressures,19 and higher endotoxin levels in their EPS and/or postmassage urine.20 Until the clinical significance of EPS inflammation is resolved, it is desirable to provide accurate EPS WBC concentrations in research studies using a counting chamber. It is also important to determine the range of EPS WBC concentrations in normal men using optimal methods. Use of a counting chamber is a well-accepted and readily available clinical laboratory procedure that can easily be adapted to office practice. CONCLUSIONS This study addressed three questions. First, how does the EPS WBC concentration relate to traditional EPS WBC counts per hpf? There was a relationship, but this correlation was not strong (R2 ⫽ 0.051), and no correlation was found over the usual range of WBC counts used to distinguish inflammatory from noninflammatory CP/CPPS. Second, should a counting chamber be recom33
mended for research studies? Our findings suggest that the hemocytometer method, or other quantitative methods, should be recommended. Finally, can we standardize the EPS wet mount to improve results? The standardized wet mount proved superior to the traditional gram-stained smear, but this method was imprecise for calculating EPS WBC concentrations (R2 ⫽ 0.244). The counting chamber method can be adapted to most clinical settings, providing accurate and rapid results with substantially less technical difficulty than use of alternative markers of disease activity, such as cytokine levels or oxidative stress. It would appear prudent to include quantitative determination of the EPS WBC concentration as a recommended evaluation in research studies of CP/CPPS. REFERENCES 1. Krieger JN, Nyberg L Jr, and Nickel JC: NIH consensus definition and classification of prostatitis. JAMA 282: 236 – 237, 1999. 2. Drach GW, Meares EM Jr, Fair WR, et al: Classification of benign disease associated with prostatic pain: prostatitis or prostatodynia? J Urol 120: 266, 1978. 3. Anderson RU, and Weller C: Prostatic secretion leukocyte studies in non-bacterial prostatitis (prostatosis). J Urol 121: 292–294, 1979. 4. Wright ET, Chmiel JS, Grayhack JT, et al: Prostatic fluid inflammation in prostatitis. J Urol 152: 2300 –2303, 1994. 5. Schaeffer AJ, Wendel EF, Dunn JK, et al: Prevalence and significance of prostatic inflammation. J Urol 125: 215–219, 1981. 6. Alexander RB, Ponniah S, Hasday J, et al: Elevated levels of proinflammatory cytokines in the semen of patients with chronic prostatitis/chronic pelvic pain syndrome. Urology 52: 744 –749, 1998. 7. Hochreiter WW, Nadler RB, Koch AE, et al: Evaluation of the cytokines interleukin 8 and epithelial neutrophil activating peptide 78 as indicators of inflammation in prostatic secretions. Urology 56: 1025–1029, 2000.
34
8. Alexander RB, Brady F, and Ponniah S: Autoimmune prostatitis: evidence of T cell reactivity with normal prostatic proteins. Urology 50: 893–899, 1997. 9. Shahed AR, and Shoskes DA: Oxidative stress in prostatic fluid of patients with chronic pelvic pain syndrome: correlation with gram positive bacterial growth and treatment response. J Androl 21: 669 –675, 2000. 10. Thin RN, and Simmons PD: Chronic bacterial and nonbacterial prostatitis. Br J Urol 55: 513–518, 1983. 11. Muller CH, Berger RE, Mohr LE, et al: Comparison of microscopic methods for detecting inflammation in expressed prostatic secretions. J Urol 166: 2518 –2524, 2001. 12. Krieger JN, Riley DE, Roberts MC, et al: Prokaryotic DNA sequences in patients with chronic idiopathic prostatitis. J Clin Microbiol 34: 3120 –3128, 1996. 13. Nickel JC, Nyberg LM, and Hennenfent M: Research guidelines for chronic prostatitis: consensus report from the first National Institutes of Health International Prostatitis Collaborative Network. Urology 54: 229 –233, 1999. 14. Schaeffer AJ, Landis JR, Knauss JS, et al: Demographic and clinical characteristics of men with chronic prostatitis: the National Institutes of Health Chronic Prostatitis Cohort Study. J Urol 168: 593–598, 2002. 15. Schaeffer AJ, Knauss JS, Landis JR, et al: Leukocyte and bacterial counts do not correlate with severity of symptoms in men with chronic prostatitis: the National Institutes of Health Chronic Prostatitis Cohort Study. J Urol 168: 1048 –1053, 2002. 16. Stamey TA: Prostatitis. J R Soc Med 74: 22–40, 1981. 17. Stamm WE: Measurement of pyuria and its relation to bacteriuria. Am J Med 75: 53–58, 1983. 18. Krieger JN, Ross SO, Penson DA, et al: Symptoms of inflammatory versus non-inflammatory CP/CPPS. Urology 60: 959 –963, 2002. 19. Mehik A, Hellstrom P, Nickel JC, et al: The chronic prostatitis-chronic pelvic pain syndrome can be characterized by prostatic tissue pressure measurements. J Urol 167: 137– 140, 2002. 20. Li LJ, Shen ZJ, Lu YL, et al: The value of endotoxin concentrations in expressed prostatic secretions for the diagnosis and classification of chronic prostatitis. BJU Int 88: 536 – 539, 2001.
UROLOGY 62 (1), 2003