CHRONIC PROSTATITIS: CHARLOTTESVILLE TO SEATTLE

0022-5347/04/1726-2557/0 THE JOURNAL OF UROLOGY® Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 172, 2557–2560, December 2004 Printed in U.S.A.

DOI: 10.1097/01.ju.0000144291.05839.a0

CHRONIC PROSTATITIS: CHARLOTTESVILLE TO SEATTLE JOHN N. KRIEGER*

AND

DONALD E. RILEY

From the Department of Urology, University of Washington School of Medicine, Seattle, Washington

ABSTRACT

Purpose: Since few men with chronic prostatitis/chronic pelvic pain syndrome (CP/CPPS) have culturable bacteria by traditional approaches, we used sensitive molecular methods to determine presence of fastidious microorganisms. Materials and Methods: We evaluated 135 men with CP/CPPS by standardized clinical evaluation, and by lower tract localization cultures and chamber counts of expressed prostatic secretions of leukocytes. We excluded from study patients with bacteriuria, bacterial prostatitis, urethritis or positive urethral cultures. Prostate biopsy was obtained using a double-needle technique to limit contamination. We chose molecular approaches because previous studies had used culture antigen detection in urine, urethral swabs and expressed prostatic secretions. However, interpretation of such studies is complicated because urogenital samples often acquire bacteria while passing through the urethra. We used specific and broad-spectrum polymerase chain reaction (PCR) assays. Results: Only 10 (8%) of the 135 subjects had positive specific PCR assays, including Mycoplasmia genitalium, Chlamydia trachomatis and Trichomonas vaginalis. Our findings suggested that C. trachomatis, T. vaginalis and M. genitalium may be identified in some patients with CP/CPPS, even among men with no evidence of urethritis and with negative urethral cultures and other assays. The broad-spectrum PCR assays provided the most provocative findings. DNA encoding tetracycline resistance was identified in 25% of subjects, and 77% of subjects had evidence of 16S rDNAs. The white blood cell concentration in the prostatic secretions correlated with identification of 16S rDNAs in prostate tissue (p ⬍0.01). Conclusions: Delineating the precise role of these organisms in the etiology of CP/CPPS may help define better diagnostic and treatment algorithms. KEY WORDS: prostatitis, pelvic pain, infection, DNA WHAT I LEARNED FROM JAY GILLENWATER

After completing my training in urology at The New York Hospital-Cornell Medical Center, I came to the University of Virginia as an American Urological Association fellow. There were 2 considerations behind this decision. The first was the outstanding opportunity to work with the world-class faculty in infectious diseases at Mr. Jefferson’s University (also known as “The University”). The second may have been even more important. That was my assessment of Dr. Jay Gillenwater as an outstanding teacher and mentor. I was not disappointed. The following was among the first, and most important, lessons I learned from Jay, “Before you do any experiments, go to the library. See what other people have done. Then talk to the experts.” At the time I did not realize what a long speech this was for Jay. Then, to assure that I learned where the library was he assigned me to write a long, review article for Investigative Urology.1 I was especially interested in urethritis and prostatitis, and spent considerable time studying every article I could find, especially the work of the preeminent Stanford group. Jay kindly arranged for Dr. Thomas Stamey to visit The University as a guest professor to explain his approach to urinary tract infections and prostatitis. I finished at Virginia and moved to Seattle as a new Assistant Professor at the University of Washington. My plan was to pursue my interest in genitourinary infections and prostatitis. Shortly after I arrived, Doctor Stamey kindly

advised me, “Find another problem. How about prostate cancer? The great infection problems are solved. We know how to treat prostatitis.” The new consensus classification recognizes the 4 prostatitis syndromes of acute bacterial prostatitis, chronic bacterial prostatitis, chronic prostatitis/chronic pelvic pain syndrome (CP/CPPS) and asymptomatic inflammatory prostatitis. Infection with standard uropathogenic bacteria is responsible for acute and chronic bacterial prostatitis. However, few patients presenting with chronic prostatitis are in these categories. Chronic prostatitis/chronic pelvic pain syndrome comprises the largest group of symptomatic patients. Recent studies suggest that fastidious bacteria, ie those not culturable on standard media but detectable using molecular techniques, may have a role in chronic prostatitis/chronic pelvic pain syndrome. Few data are available concerning the role of infection and other agents in asymptomatic inflammatory prostatitis. Infection clearly has a role in prostatitis but defining the proportion of prostatitis related to genitourinary tract infection will require a broad-based approach using new molecular methods.

CLASSIFICATION OF PROSTATITIS

Traditional approach. The traditional classification of prostatitis included the 4 categories of acute bacterial prostatitis, chronic bacterial prostatitis, nonbacterial prostatitis and prostatodynia.2 This classification was based on presence of symptoms, history of bacteriuria, physical examination and presence of white cells in the prostatic secretions. The problem was that this traditional approach did not work too well for most patients I was seeing in my clinic.

Supported by Grant DK38955 from the National Institutes of Health. * Correspondence: Department of Urology, University of Washington, VAPSHCS (112-GU), 1600 South Columbian Way, Seattle, Washington 98108 (telephone: 206-764-2265; FAX: 206-764-2239; e-mail: [email protected]). 2557

2558

CHRONIC PROSTATITIS

Practical problems were that only a small proportion of patients had acute or chronic bacterial prostatitis and that few clinicians conducted the careful localization studies necessary for patient evaluation. The great majority of men presenting with prostatitis had pain complaints, sexual dysfunction and voiding symptoms but little or no evidence of bacterial infection in the prostatic fluid.3, 4 Consensus classification. Recognizing this problem, the National Institutes of Health (NIH) convened a conference to develop a new, more clinically relevant classification.5 Like the traditional classification, the consensus classification has 4 categories of acute bacterial prostatitis, chronic bacterial prostatitis, chronic prostatitis/chronic pelvic pain syndrome and asymptomatic inflammatory prostatitis. The acute and chronic bacterial prostatitis categories were unchanged from the previous definitions. A new category of chronic prostatitis/chronic pelvic pain syndrome included inflammatory and noninflammatory subgroups. Finally, there was a group for asymptomatic inflammatory prostatitis with no symptoms but evidence of inflammation in either the prostatic parenchyma or lower genitourinary tract samples. An important distinction from the traditional classification is that the consensus classification categorizes patients based on presence or absence of leukocytes in the expressed prostatic secretions, post-massage urine or seminal fluid analysis, rather than the prostatic secretions alone. Although this new classification has revived considerable academic interest in the field of prostatitis, to date there are few evidence based recommendations for rational therapy for most patients. The great majority of men presenting with chronic prostatitis have pain complaints, sexual dysfunction and voiding symptoms with little or no evidence of bacterial infection in the prostate.3, 4 Nonetheless, antibiotics remain the most commonly prescribed therapy. In a controlled study many patients benefited from antimicrobial therapy,6 providing empirical support for the role of infection in CP/CPPS. Many men relate the onset to sexual activity, which is often an episode of urethritis. However, the problem is that common uropathogens are localized to the prostate in only a small minority of patients. It is crucial to distinguish such patients with bacteriuria because they may have acute or chronic bacterial prostatitis. We have rational approaches for treatment of such unusual cases, potential role of uncommon microorganisms in CP/CPPS is summarized.

nongonococcal urethritis (NGU) more than 20 years ago. It is difficult to culture. Early immunological assays for this organism had mixed results in patients with urethritis.8 More recent studies using polymerase chain reaction (PCR) based assays provided stronger evidence that this organism causes acute NGU and probably chronic NGU also.9 A recent study identified M. genitalium by PCR in 22% of NGU cases (27 of 121) compared to 4% of controls (4 of 117, p ⬍0.01).10 It is interesting to compare populations of patients with NGU or persistent/recurrent NGU and CP/CPPS. Both populations are difficult to treat and are composed of men with a mean age in the 30s. Both conditions respond somewhat to antimicrobial therapy. Patients with NGU tend to be more sexually active than controls and are more likely to have a history of urethritis. These findings are also similar to men with CP/CPPS, particularly if they have evidence of white cells. Urethral white cells have been identified in all patients with NGU and in 50% of the NIH CP/CPPS cohort.11 When comparing symptoms of NGU (especially persistent or recurrent NGU) with those of CP/CPPS, some remarkable similarities become apparent. Of the 7 symptoms evaluated in the NIH chronic prostatitis symptom index12 3 symptoms (penile pain/discomfort, urinary frequency and dysuria) are common in both populations. The other 4 symptoms (pain/ discomfort in the perineum, pain/discomfort in the testicles, pain/discomfort in the pubic or bladder area, and pain/discomfort during or after ejaculation) are not typically described in studies of patients with NGU. In contrast, urethral discharge (which can be scant) is characteristic of NGU but is not specifically elucidated in studies of CP/CPPS. The “urethritis issue.” Various approaches may be used for studies of CP/CPPS in patient populations in which evidence of urethritis is common. One approach is to emphasize careful clinical evaluation, specifically eliciting symptoms and signs of urethritis that may be subtle. A second approach involves careful laboratory evaluation of urethral white cells as well as the recognized urethral pathogens, with specific testing. A third approach is to use a case definition that excludes patients with symptoms, signs or laboratory evidence of urethritis. For example, the consensus definition specifically excludes such patients.5 Another approach would be to say that, “Urethritis is not a problem (or confounding variable).” This approach uses a convenience sample of symptomatic patients in the hope that treatment with effective agents will include sufficient subjects to show a statistically significant benefit.

FASTIDIOUS MICROORGANISMS IN CP/CPPS

Definition of fastidious microorganisms. We define “fastidious” microorganisms as those that are not isolated on standard media for culturing the urine. These organisms are identified in patients who do not fit the criteria for acute or chronic bacterial prostatitis. In the literature there are data for a number of fastidious organisms in chronic prostatitis, including Chlamydia trachomatis, the genital mycoplasmas (Ureaplasma urealyticum, Mycoplasma hominis and Mycoplasma genitalium), the protozoan Trichomonas vaginalis, as well as other organisms such as Neisseria gonorrhoeae, genital tract viruses (particularly herpes simplex virus types 1 and 2, and cytomegalovirus), fungi, anaerobic bacteria and gram-positive bacteria. Many of these fastidious organisms bear similarities to the common uropathogens that cause urinary tract infections. These similarities include an ascending route of infection in most cases, the observation that many patients respond to antimicrobial therapy and that patients often have inflammation. Many of these fastidious organisms are also associated with urethritis. We review M. genitalium as an example. Mycoplasma genitalium. Currently there is considerable interest in M. genitalium in lower genitourinary tract syndromes.7 This organism was first isolated from 2 men with

UNIVERSITY OF WASHINGTON EXPERIENCE

Based on these considerations, we excluded patients from our studies if they had any evidence of urethritis identified on clinical or laboratory evaluation. Most clinicians would agree that this patient population had prostatitis.13 (We recognize that excluding cases with any evidence of urethritis may also exclude some prostatitis cases.) We started by doing prostate tissue cultures from 85 men with CP/CPPS. Urethral, urine and perineal prostate biopsies were cultured in an anaerobic research laboratory.13 Men with white cells in prostatic secretions were more likely to have any bacteria isolated (p ⫽ 0.01), positive cultures for anaerobic bacteria (p ⫽ 0.03), higher total bacterial counts (p ⫽ 0.02) and more bacterial species identified (p ⫽ 0.02) on prostate biopsies. The observations led us to hypothesize that bacterial colonization/ invasion of the prostate may occur more commonly than appreciated by standard techniques and that such uncommon microorganisms might be associated with CP/CPPS. Molecular approaches. We chose molecular approaches because previous studies had used culture antigen detection in urine, urethral swabs and expressed prostatic secretions.14 –16 However, interpretation of such studies is complicated because urogenital samples often acquire bacteria

2559

CHRONIC PROSTATITIS

while passing through the urethra. Therefore, we evaluated urethra and prostate tissue. Then we compared the prostate tissue findings with the inflammatory response in the prostatic secretions. We used 2 complementary approaches, both involving PCR methods. With the first approach we used specific PCRs that were developed to test for all pathogens implicated in chronic prostatitis. With the second approach we used broadspectrum PCRs to test for bacterial DNAs, including common tetracycline resistance encoding genes and bacterial ribosomal encoding genes (16S rDNA). Bacterial 16S rDNA identification has been used to identify and define roles for microorganisms in other idiopathic diseases. The conserved sequences where the amplifications are primed are read as positive or negative. Variable sequences can then be used to determine bacterial taxonomy. We evaluated 135 men with CP/CPPS by standardized clinical evaluation and thorough microbiological studies for fastidious organisms, lower tract localization cultures and chamber counts of EPS leukocytes.15 The protocol involved more than 1,000 physician visits (at least 4 visits for each of the 260 men evaluated to select the 135 subjects). We excluded patients with bacteriuria, bacterial prostatitis, urethritis, or positive urethral culture or other tests for urethral pathogens. Prostate biopsy was obtained using a double needle technique to limit contamination, a problem that has proven difficult in some studies.17, 18 RESULTS

Only 10 (8%) of the 135 subjects had positive specific PCRs, including M. genitalium (positive 4), C. trachomatis (positive 3) and T. vaginalis (positive 2). One case was positive for M. genitalium and C. trachomatis. These observations fit with prior findings that C. trachomatis and T. vaginalis may be identified in prostate tissue. To our knowledge, this was the first demonstration of M. genitalium in the prostate. Our findings suggested that C. trachomatis, T. vaginalis and M. genitalium may be identified in some patients with CP/ CPPS, even among men with no evidence of urethritis and with negative urethral cultures in other assays. The majority of specific PCR assays were negative, including 99.5% of 2,552 assays. We had no positives for the general mycoplasma probe, U. urealyticum probe, herpes virus probes or cytomegalovirus probe. The broad-spectrum PCR assays provided the most provocative findings. DNA encoding tetracycline resistance was identified in 25% of subjects and 77% of subjects had evidence of 16S rDNAs. These PCR assays were done at different laboratories but the results were highly correlated. The white blood cell count in the prostatic secretions correlated with identification of 16S rDNAs in prostate tissue (p ⬍0.01). There was strong evidence against contamination in this study. The majority of the PCR tests were negative. We had taken extreme care in procuring and handling the clinical samples. The 16S rDNA data were supported by finding tetracycline resistance in a high proportion of the 16S rDNA positive cases. Furthermore, patients with 16S rDNAs were significantly more likely to have inflammation in prostatic secretions. DNA cloning and sequencing revealed multiple sources of 16S rDNAs in the prostate. Many of the sequences were novel, meaning less than 95% related to known bacterial rDNAs. DNA sequence data revealed that the rDNAs were distinct from known gut, skin and laboratory contaminations. The correlation with tetracycline resistance fit with the observation that antibiotics provide transitory, if any, relief for many patients. Our findings were criticized because we had compared patients with and without inflammation in expressed prostatic secretions but we had no controls.17, 19 To address this issue, we compared prostate tissue biopsies from patients

with CP/CPPS to biopsies from patients undergoing radical prostatectomy for cancer.14 Overall, 21 (20%) of 117 patients with prostate cancer had 16S rDNAs compared to 79 (47%) of 170 subjects with CP/CPPS (p ⬍0.01). These observations support a potential role for uncommon organisms in CP/ CPPS. Issues. To evaluate the potential role of fastidious microorganisms in chronic prostatitis, we evaluated a carefully selected patient population using optimal testing for all recognized pathogens, including culture, immunological and molecular testing of the urethra. We then used traditional microbiological approaches, incorporating careful localization studies to rule out prostatic infection and optimal white blood cell evaluation. Our results suggest that uncommon organisms may be identified in prostate tissue from many patients with CP/CPPS. These molecular data do not establish that bacteria cause CP/CPPS. Detecting bacteria in prostate tissue may prove to be important in unraveling the pathophysiology of this syndrome. Additional sequencing is clearly necessary to define the microbiology of the prostate gland and to determine the precise role of bacteria in chronic prostatitis. Such studies should include many more subjects with chronic prostatitis, as well as other populations. Establishing a role for bacteria in CP/CPPS will likely involve many steps but in our opinion detecting bacteria in the prostate is an important step in this process. Delineating the precise role of these organisms in the etiology of CP/CPPS may help define better diagnostic and treatment algorithms. FROM THE UNIVERSITY TO THE UNIVERSITY OF WASHINGTON

Jay Gillenwater’s lessons continue to serve me well. At the University of Virginia I accomplished my original goals by acquiring a strong theoretical and technical background. Jay encouraged me to pursue a third goal, developing the compulsion to read a wide range of scientific articles. In Seattle this background provided the confidence to pursue a career that has been a bit iconoclastic, often disregarding the best expert opinion and advice. But before making these choices I had studied almost every article in the field and had a comprehensive understanding of the relevant literature, especially the strengths and limitations of critical studies. Awareness of such nuances led me to pursue new approaches to clinical issues that did not fit into the then current conceptual framework. Now, I advise our residents and fellows to adapt similar attitudes and approaches to a wide range of urological issues. My only problem is that it takes me many paragraphs to reiterate lessons that Jay could teach in 1 or 2 sentences. REFERENCES

1. Krieger, J. N.: Urologic aspects of trichomoniasis. Invest Urol, 18: 411, 1981 2. Drach, G. W., Fair, W. R., Meares, E. M. and Stamey, T. A.: Classification of benign disease associated with prostatic pain: prostatitis or prostatodynia? J Urol, 120: 266, 1978 3. Alexander, R. B. and Trissel, D.: Chronic prostatitis: results of an Internet survey. Urology, 48: 568, 1996 4. Krieger, J. N., Egan, K. J., Ross, S. O., Jacobs, R. and Berger, R. E.: Chronic pelvic pains represent the most prominent urogenital symptoms of “chronic prostatitis.” Urology, 48: 715, 1996 5. Krieger, J., Nyberg, L., Jr. and Nickel, J.: NIH consensus definition and classification of prostatitis. JAMA, 282: 236, 1999 6. Nickel, J. C., Downey, J., Johnston, B., Clark, J., Group, T. C. and Canadian Prostatitis Research Group: Predictors of patient response to antibiotic therapy for the chronic prostatitis/ chronic pelvic pain syndrome: a prospective multicenter clinical trial. J Urol, 165: 1539, 2001 7. Yoshida, T., Deguchi, T., Ito, M., Maeda, S., Tamaki, M. and Ishiko, H.: Quantitative detection of Mycoplasma genitalium

2560

8. 9. 10.

11.

12.

13.

CHRONIC PROSTATITIS

from first-pass urine of men with urethritis and asymptomatic men by real-time PCR. J Clin Microbiol, 40: 1451, 2002 Hooton, T., Roberts, M. and Kenny, G.: Mycoplasma genitalium and non-gonococcal urethritis. Lancet, 343: 69, 1994 Taylor-Robinson, D.: Mycoplasma genitalium—an up-date. Int J STD AIDS, 13: 145, 2002 Totten, P. A., Schwartz, M. A., Sjostrom, K. E., Kenny, G. E., Handsfield, H. H., Weiss, J. B. et al: Association of Mycoplasma genitalium with nongonococcal urethritis in heterosexual men. J Infect Dis, 183: 269, 2001 Schaeffer, A. J., Knauss, J. S., Landis, J. R., Propert, K. J., Alexander, R. B., Litwin, M. S. 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, 2002 Litwin, M. S., McNaughton-Collins, M., Fowler, F. J., Jr., Nickel, J. C., Calhoun, E. A., Pontari, M. A. et al: The National Institutes of Health chronic prostatitis symptom index: development and validation of a new outcome measure. Chronic Prostatitis Collaborative Research Network. J Urol, 162: 369, 1999 Berger, R. E., Krieger, J. N., Rothman, I., Muller, C. H. and Hillier, S. L.: Bacteria in the prostate tissue of men with

idiopathic prostatic inflammation. J Urol, 157: 863, 1997 14. Krieger, J. N., Riley, D. E., Vessella, R. L., Miner, D. C., Ross, S. O. and Lange, P. H.: Bacterial DNA sequences in prostate tissue from patients with prostate cancer and chronic prostatitis. J Urol, 164: 1221, 2000 15. Krieger, J., Riley, D., Roberts, M. and Berger, R. E.: Prokaryotic DNA sequences in patients with chronic idiopathic prostatitis. J Clin Microbiol, 34: 3120, 1996 16. Riley, D. E., Berger, R. E., Miner, D. and Krieger, J. N.: Diverse and related 16S rRNA-encoding DNA sequences in prostate tissues of men with chronic prostatitis. J Clin Microbiol, 36: 1646, 1998 17. Keay, S., Zhang, C. O., Baldwin, B. R. and Alexander, R. B.: Polymerase chain reaction amplification of bacterial 16s rRNA genes in prostate biopsies from men without chronic prostatitis. Urology, 53: 487, 1999 18. Lee, J. C., Muller, C. H., Rothman, I., Agnew, K. J., von Eschenbach, D., Ciol, M. A. et al: Prostate biopsy culture findings of men with chronic pelvic pain syndrome do not differ from those of healthy controls. J Urol, 169: 584, 2003 19. Hochreiter, W. W., Duncan, J. L. and Schaeffer, A. J.: Evaluation of the bacterial flora of the prostate using a 16S rRNA gene based polymerase chain reaction. J Urol, 163: 127, 2000