Detection of Escherichia coli 16S RNA and Cytotoxic Necrotizing Factor 1 Gene in Benign Prostate Hyperplasia

Transcription

1 european urology 51 (2007) available at journal homepage: Benign Prostatic Hyperplasia Detection of Escherichia coli 16S RNA and Cytotoxic Necrotizing Factor 1 Gene in Benign Prostate Hyperplasia Johanna Bergh a, Ingrid Marklund a, Camilla Thellenberg-Karlsson b, Henrik Grönberg c, Fredrik Elgh a,d, Oleg A. Alexeyev a, * a Department of Medical Biosciences/Pathology, Umeå University, Umeå, Sweden b Department of Radiation Sciences/Oncology, Umeå University, Umeå, Sweden c Department of Medical Epidemiology and Biostatistics, Karolinska Institute, Stockholm, Sweden d Department of Clinical Microbiology/Virology, Umeå University, Umeå, Sweden Article info Article history: Accepted June 7, 2006 Published online ahead of print on June 23, 2006 Keywords: Escherichia coli Cytotoxic necrotizing factor Benign prostate hyperplasia Abstract Objectives: Inflammation and occasionally necrosis is observed in the prostate tissue from patients with benign prostate hyperplasia (BPH). The etiology of prostatic inflammation/necrosis is unknown, but bacteria may be involved. Materials and methods: Retrospective analysis of archival prostate tissue samples collected during was undertaken. Three hundred fifty-two specimens from patients with BPH obtained via transurethral resection of prostate (TURP) were studied for the presence of Escherichia coli 16S RNA and E. coli virulence factor genes: cytotoxic necrotizing factor (cnf1), alpha-hemolysin (hly), an autotransported protein (sat), and P fimbriae ( papc). Results: E. coli 16S RNA was detected in 12 (3%) samples and cnf1 gene in six (1.5%) samples, with two samples being positive for both markers. hly, sat, papc genes were not detected. Of 6 cnf1-positive samples, severe inflammation and necrosis were present in four and three samples, respectively. Of eight E. coli-positive/cnf1-negative samples, five showed signs of severe inflammation and two showed severe necrosis. Conclusions: A small proportion of patients with BPH undergoing TURP are positive for E. coli 16S RNA and cnf1 gene in the prostate tissue. Further studies are needed to show that particular E. coli genotypes are involved in the development of prostatic inflammation/necrosis in BPH. # 2006 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Department of Medical Biosciences/Pathology, Umeå University, S-90185, Umeå, Sweden. Tel ; Fax: address: oleg.alexeyev@medbio.umu.se (O.A. Alexeyev) /$ see back matter # 2006 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi: /j.eururo

2 458 To increase the likelihood for cnf1, sat, hly, and papc detection in archival material, we designed PCRs amplifying short fragments of DNA (91, 91, 96 and 91 bp, respectively). The published sequence of E. coli SFT073 [10] was used to design primers for sat, hly, and papc. cnf1 primers were designed by using the published sequence [11]. Primers ( ) were as follows: papc (GCCATCAACCGGTACACCTC and AAAAGGTGeuropean urology 51 (2007) Introduction Uropathogenic strains of Escherichia coli are believed to display a variety of virulence properties that help them colonize host mucosal surfaces and circumvent host defenses to allow invasion of the normally sterile urinary tract. Among these factors, siderophores, toxins, capsules, fimbriae, and others have been described [1]. Some of these virulence factors, such as cytotoxic necrotizing factor (cnf1), alpha-hemolysin (hly), or an autotransported protein (sat), which acts as a proteolytic toxin, have been found to be located in pathogenicity islands. Studies have linked cnf1 with E. coli strains that cause prostatitis, as well as uncomplicated urinary tract infection (UTI) in women. Specifically, a significant proportion of prostatitis patient isolates have been reported to be cnf1 positive [2 5]. The significance of cnf1 in prostatitis development is further corroborated by animal studies. Infection of the rat prostate with cnf1-positive uropathogenic E. coli caused more inflammation-mediated morphologic and histologic tissue damage than did infection with isogenic cnf1- negative mutants [6]. papc, hly, and probably sat represent the other virulence factors reported to be more common among E. coli prostatitis isolates [1]. Inflammation is an almost universal finding (>90%) in the prostate of patients with benign prostate hyperplasia (BPH) [7], but little is known on association of inflammation and the presence of E. coli and its virulence factors in this condition. In this study we evaluated whether E. coli 16S RNA and E. coli virulence factors were present in the prostate of patients with BPH undergoing transurethral resection of prostate (TURP) and if their presence was associated with histologic signs of inflammation or necrosis. 2. Materials and methods 2.1. Patients Prostate tissue specimens from 402 residents (median age, 64 years; range, 51 71) in the Västerbotten county of Sweden, who were diagnosed with BPH and underwent TURP between 1982 and 1997, were retrospectively identified and studied. No patient had clinical symptoms of prostatitis or clinical and laboratory signs of urinary tract infection at the time of TURP. Prostate tissue obtained during the TURP treatment was promptly fixed in formalin, embedded in paraffin and stored at room temperature until tested. When tissue specimens from different TURP treatments from the same patient were available, the specimen from the first TURP was selected for the study. Histologic inflammation and the necrosis score were examined by board-certified uropathologists who were blinded to the polymerase chain reaction (PCR) results. Histologic inflammation was characterized as a dichotomous variable: severe versus moderate/minimal inflammation. There were no samples devoid of inflammation. Severe inflammation fulfilled the following criteria: inflammatory foci on a majority of the slides and three or more single foci on the slides, or a single focus that took up one third or more of the slide. As most prostate samples contained foci of or diffuse accumulation of inflammatory cells (lymphocytes and macrophages) in the prostate stroma, the samples were screened for the presence of polymorphonuclear leukocytes in the glandular lumen and among the glandular epithelial cells (categorized as acute intraglandular inflammation). The prostate glands were then scored for the presence of epithelial cell damage/necrosis. The grades were 0 = absent, + = discrete or ++ = severe. The study was approved by the local institutional ethic committee DNA extraction Archival samples were deparaffinized as described [8]. DNA was purified by using the QIAamp DNA Blood mini kit (Qiagen, Hilden, Germany) according to the manufacturer s instructions S rrna nested PCR assay We PCR amplified 16S rrna-encoding DNA, called 16S rrna sequences. These sequences constitute molecular signatures of prokaryotes. A first-step PCR included 29 cycles and was performed as described previously [9]. In the second step the DNA was further amplified with a set of inner primers. A PCR mix comprised of 1 PCR buffer, 1.5 mmol/l MgCl 2, 0.1 mmol/l dntp, 1 U Taq polymerase, 0.4 mmol/l of each of the following primers: 16SFac: GCTCAGATTGAACGCTGG, 16SFbc: GCT- CAGGAYGAACGCTGG, 16SRc: 59-TACTGCTGCCTCCCGTA and sterile water to give a final volume of 45 ml. Five microlitres of recombinant DNA were added to aliquots of this mixture, and the reactions were heated to 94 8C for 3 minutes, followed by 26 cycles of 94 8C for 30 seconds, 63 8C for 1 minute and 72 8C for 1 minute. PCR reactions were electrophoresed through a 2% agarose gel containing ethidium bromide and 310-bp bands were visualized by ultraviolet (UV) transillumination. To assess the limit of sensitivity of this PCR system, we performed the following experiment. After harvesting E. coli, bacterial DNA was extracted and quantified, and serial dilutions of DNA served as templates for the PCR. The limit of sensitivity of the 16S DNA PCR used in this study was determined to be approximately 300 colony- forming units per PCR reaction PCR detection of specific gene sequences

3 european urology 51 (2007) CGGCGTTACATG), sat (ATACCAGGAGTGGGAGCTGTAGTC and CCTCATCGGTAAGTACGTTCACC), hly (GCTCAGCCAATATC- TACGCAGG and CGGTTGCTTTTGTGCCATC), and cnf1 (TGCTG- TTCTGTATGGCATAGCC and GAGCATCTCCAGTGTTCCGAC). For sat, hly, papc, a PCR mix comprised of 1 PCR buffer, 0.1 mmol/l dntp, 1 U Taq polymerase, 0.2 mmol/l of each primer and sterile water to give a final volume of 35 ml. Five microlitres of DNA were added to aliquots of this mixture, and the reactions were heated at 95 8C for 3 minutes, followed by 35 cycles of 94 8C for 30 seconds, 55 8C for 30 seconds, and 72 8C for 1 minute. For cnf1, PCR conditions were essentially the same except for higher (0.4 mmol/l) primer concentrations. Serial dilutions of DNA extracted from J96 and CFT073 strains of E. coli were used to estimate sensitivity limits for cnf1, hly, papc, and sat PCRs, respectively. The sensitivity limits for the cnf1, hly, papc PCRs were approximately 40 genomes per reaction, and for the sat PCR, five genomes per reaction PCR controls As a control for the integrity of all template DNA samples, two PCR amplifications of the human beta-globin gene were performed on every specimen yielding 110- and 268-bp products [8] Anticontamination procedures The PCR laboratories have complied or exceeded standard recommendations for PCR research, including isolated setup areas with dedicated pipettes, small reagent and primer aliquots, and use of aerosol-resistant pipette tips. A separate room where positive controls were added was also used. Working areas were UV irradiated for 2 hours prior to their usage. Negative and positive controls were performed with PCR assays to monitor potential reagent or laboratory contamination Sequence analysis PCR products were cloned using the pt7blue Perfectly Blunt cloning kit (Novagen, Madison, WI, USA), and plasmids were transformed into Nova Blue Shingles competent cells (Novagen) according to the manufacturer s instructions. Colony PCR analysis was performed on 5 10 randomly selected white colonies with the use of primers U19 and P7 to verify that the colonies contained inserts. Chimeric plasmids containing cloned 16S rrna gene inserts were selected for sequencing. A median of two sequences (range, one to seven) was analyzed for each sample. A 6-ml aliquot of each of the overnight cultures was used to extract plasmid DNA by using the QIAprep Spin Miniprep kit (Qiagen). Sequences for both strands were obtained by using the BigDye Terminator Cycle Sequencing kit 1.1 (Applied Biosystems, Foster City, CA, USA) and analysed in ABI PRISM 3700 DNA Analyzer (AME Bioscience, Toroed, Norway). Approximately 320 bp of each 16S rrna gene sequence were obtained by this approach. Chimeric 16S rrna sequences were identified by the program Chimera Check, accessed through the Ribosomal Database Project and then discarded. Homology searches were performed with the alignment search tool BLAST, an Internet public database. Sharing >98% nucleotide identity with a known species or with an uncultured bacterium sequence was used as the criteria for defining the species Statistical analysis The Fisher exact test was used for statistical analysis. 3. Results Of 402 samples tested, 352 were positive for the human beta-globin gene. These samples were considered to have sufficient DNA quality and were therefore used for subsequent analysis with 16S DNA PCR. In total, 96 of 352 (27%) samples were positive for the 16S RNA gene. Of the 96 samples positive for the 16S RNA gene, 12 (12%) were harboring sequences sharing >98% similarity to known sequences of E. coli. Of the same 96 samples, 6 (6%) were positive for cnf1. Two samples were positive for both E. coli 16S RNA and cnf1. The remaining four cnf1-positive samples harbored the following bacterial DNA sequences: 1, uncultured bacteria and Pseudomonas species; 2 and 3, uncultured bacteria; 4, Streptococcus mutans and Propionibacterium acnes. No cnf1 was detected in 100 randomly selected 16S RNA-negative samples. No hly-, sat-, or papc-positive samples were found in 96 16S RNA-positive samples and 100 randomly selected 16S RNA-negative samples. Table 1 Histological inflammation and necrosis in the prostate tissue from patients with BPH positive for E. coli 16s RNA and/or cnf1 gene E. coli 16S RNA cnf1 Mild inflammation Severe inflammation Necrosis NA + NA NA NA + NA NA NA NA: non-available. Necrosis grade: 0-absent, (+)-discrete, (++)-severe.

4 460 european urology 51 (2007) Fig. 1 Sections from human prostate tissue at 200 magnification. A C, In the cnf1-positive patient with benign prostatic hyperplasia (BPH), there is intraglandular accumulation of polymorphonuclear leukocytes, epithelial cell necrosis and intraglandular accumulation of cellular debris. D, In the E. coli/cnf1-negative patient with BPH, there is low-grade chronic inflammation with accumulation of lymphocytes seen around some of the glands, but there are no signs of acute inflammation or tissue damage. The relationship between the presence of cnf1 and E. coli and histologic signs of inflammation and the presence of necrosis in prostate tissues is shown in Table 1 and Fig. 1. Of 16 E. coli- and/or cnf1-positive samples, 14 and 13 were available for evaluation of inflammation and necrosis, respectively. Severe histologic inflammation was detected in 4 (66%) of 6 cnf1-positive, 5 (62%) of 8 E. coli-positive/cnf1- negative samples and 147 (43%) of 336 E. coli/cnf1- negative samples. In total, necrosis was present in 5 of 6 (83%) cnf1-positive samples, 7 of 8 (87%) E. colipositive/cnf1-negative and 6 (46%) of 13 E. coli/cnf1- negative samples. Severe necrosis accompanied by acute intraglandular inflammation was present in 3 of 6 (50%) cnf1-positive samples, 2 of 8 (25%) E. colipositive/cnf1-negative samples and 3 (23%) of 13 E. coli/cnf1-negative samples ( p > 0.05, Fisher exact probability test). 4. Discussion To the best of our knowledge, this is the first study to report the presence of E. coli virulence factors along with E. coli 16S RNA directly in prostate tissue taken from patients with BPH. We demonstrate that a small proportion of patients (12 of 352 [3%]) with BPH harbour the 16S RNA gene of E. coli in the prostate at the time of TURP. cnf1 another marker for E. coli was the only virulence factor found in the prostate tissue (6 of 352 [1.5%]), whereas papc, sat and hly were not detected. Archival formalin-fixed, paraffin-embedded tissue is a suitable target for 16S PCR and is comparable to fresh tissue if small fragments are amplified [12]. In our material nearly 90% of the samples studied were positive for the beta-globin gene, indicating sufficient DNA quality and the absence of PCR inhibitors in archival prostate tissue. There was a 100% concordance between the presence of 16S RNA, a ubiquitous bacterial gene, and the presence of cnf1. However, only two cnf1-positive samples were E. coli positive, whereas in the four remaining cnf1-positive samples, sequence analysis revealed no E. coli 16S RNA. In these samples, the E. coli 16S RNA concentration may have been below the detection limit of the 16S PCR used. In addition, analysing only 5 10 clones from each 16S RNApositive sample may have failed to detect E. coli 16S RNA sequences present at a low frequency [13]. Therefore, the true incidence of E. coli in prostate tissue in patients with BPH may be underestimated in this study. Summing up data on the presence of E. coli 16S RNA and cnf1 as two markers of E. coli in the material studied would increase the incidence of E. coli to 16 of 352 (4.5%) samples, with 6 of these

5 european urology 51 (2007) (37.5%) being cnf1 positive. E. coli has been implicated as an etiologic factor in more than 50% of nearly 1500 chronic prostatitis cases [14].InBPHthe E. coli isolation rates are much lower with 3 9% of patients having positive cultures from the prostate [15,16] and 19% of patients having positive urine cultures [17]. Terai et al. [5] have shown that E. coli strains that cause prostatitis and possess multiple virulence factors have cnf1 and hly as a common (48 60% of all strains) combination [5]. In our study, despite repetitive attempts, cnf1-positive samples tested negative for other virulence factors studied. The reason for this observation is not known but could not be attributed to differences in sensitivity of the PCRs used to detect cnf1, papc, sat and hly. Virulence genotypes of E. coli isolates from patients with prostatitis have been well documented [2 5,18]. However, the isolates studied were cultured mainly from urine and linked to prostate disease on clinical grounds. This observation is important to bear in mind, since direct comparison of bacterial isolates from the urethra and the prostate has revealed identical organisms in only 30% of patients [19]. In the present study, the prostate specimens were directly obtained through the resectoscope sheath, thus minimizing but not eliminating the potential for penile urethral contact and subsequent contamination of the prostate by E. coli. Because of the retrospective design of our study, we cannot address the issue on whether the presence of indwelling catheters affected detection rates of E.coli/cnf1 directly in the prostate. Nonetheless, our finding of cnf1 gene directly in the prostate tissue supports the earlier observations that cnf1 is especially relevant in E. coli isolates linked to prostate inflammation [2 5,18]. The prevalence of cnf1 in E. coli isolates from patients with acute prostatitis has been reported to be in the range of 60 80% [3,4] versus the predicted prevalence of 37.5% in our study. It is important to bear in mind that patients with BPH enrolled in this study did not have clinical signs of prostatitis or urinary tract infection at the time of TURP. The small numbers of E. coli- and cnf1-positive samples detected in this study have limited the power to detect a statistically meaningful difference in intensity of prostate inflammation/necrosis in positive versus negative specimens. More then 60% of the samples positive for E. coli 16S RNA and/or cnf1 had severe inflammation, compared with 43% of negative samples. In cnf1-positive specimens, the frequency of severe necrosis (50%) tended to be more common than in E. coli-positive/cnf1-negative samples (25%) and E. coli/ cnf1-negative samples (23%). In the current study we cannot establish whether inflammation observed was deep or superficial; this issue must be also addressed. Finally, it is not known to what extent concomitant microorganisms (Pseudomonas, streptococci and Propionibacterium) or microorganisms possibly undetected in this study may contribute to inflammation/necrosis seen in BPH. 5. Conclusions Analysis of 352 prostate tissue samples from patients with BPH undergoing TURP showed the presence of E. coli 16S RNA and/or cnf1 gene. Additional studies are needed to clarify the role of cnf1-positive E. coli in the development of histologic inflammation and necrosis seen in patients with BPH. Follow-up studies are also needed to demonstrate the utility that E. coli/cnf1 detection may serve as early markers for severe complications (e.g., urinary tract infection/sepsis after TURP treatment) and thus contribute to assessment of BPH [20]. Conflicts of interest The Kempe Foundation (JCK-2531), the Swedish Cancer Society, the Cancer Research Foundation of Northern Sweden, the Percy Falk Foundation for prostate cancer research (Sweden), the Maud & Birger Gustavsson Foundation, and the medical faculty at Umeå University, Sweden (AMP & LP ), provided financial support for this study. Acknowledgements We thank Professor Bernt Eric Uhlin and Marianne Rönnmark, Department of Bacteriology, Umeå University, for providing bacterial strains. References [1] Oelschlaeger TA, Dobrindt U, Hacker J. Virulence factors of uropathogens. Curr Opin Urol 2002;12:33 8. [2] Andreu A, Stapleton AE, Fennell C, et al. Urovirulence determinants in Escherichia coli strains causing prostatitis. J Infect Dis 1997;176: [3] Mitsumori K, Terai A, Yamamoto S, Ishitoya S, Yoshida O. Virulence characteristics of Escherichia coli in acute bacterial prostatitis. J Infect Dis 1999;180: [4] Ruiz J, Simon K, Horcajada JP, et al. Differences in virulence factors among clinical isolates of Escherichia coli causing

6 462 european urology 51 (2007) cystitis and pyelonephritis in women and prostatitis in men. J Clin Microbiol 2002;40: [5] Terai A, Yamamoto S, Mitsumori K, et al. Escherichia coli virulence factors and serotypes in acute bacterial prostatitis. Int J Urol 1997;4: [6] Rippere Lampe KE, Lang M, Ceri H, Olson M, Lockman HA, O Brien AD. Cytotoxic necrotizing factor type 1-positive Escherichia coli causes increased inflammation and tissue damage to the prostate in a rat prostatitis model. Infect Immun 2001;69: [7] Bedalov G, Vuckovic I, Fridrih S, Bruk M, Puskar D, Bartolin Z. Prostatitis in benign prostatic hyperplasia: a histological, bacteriological and clinical study. Acta Med Croatica 1994;48: [8] Greer CE, Wheeler CM, Manos MM. Sample preparation and PCR amplification from paraffin-embedded tissues. PCR Methods Appl 1994;3:S [9] Harris KA, Hartley JC. Development of broad-range 16S rdna PCR for use in the routine diagnostic clinical microbiology service. J Med Microbiol 2003;52: [10] Welch RA, Burland V, Plunkett 3rd G, et al. Extensive mosaic structure revealed by the complete genome sequence of uropathogenic Escherichia coli. Proc Natl Acad Sci U S A 2002;99: [11] Falbo V, Pace T, Picci L, Pizzi E, Caprioli A. Isolation and nucleotide sequence of the gene encoding cytotoxic necrotizing factor 1 of Escherichia coli. Infect Immun 1993;61: [12] De Groote D, Haesebrouck F, van Doorn LJ, Vandamme P, Ducatelle R. Evaluation of a group-specific 16S ribosomal DNA-based PCR for detection of Helicobacter bizzozeronii, Helicobacter felis, and Helicobacter salomonis in fresh and paraffin-embedded gastric biopsy specimens. J Clin Microbiol 2001;39: [13] Fredricks DN, Fiedler TL, Marrazzo JM. Molecular identification of bacteria associated with bacterial vaginosis. N Engl J Med 2005;353: [14] Weidner W, Schiefer HG, Krauss H, Jantos C, Friedrich HJ, Altmannsberger M. Chronic prostatitis: a thorough search for etiologically involved microorganisms in 1,461 patients. Infection 1991;19:S [15] Gorelick JI, Senterfit LB, Vaughan Jr ED. Quantitative bacterial tissue cultures from 209 prostatectomy specimens: findings and implications. J Urol 1988;139: [16] Nickel JC, Downey J, Young I, Boag S. Asymptomatic inflammation and/or infection in benign prostatic hyperplasia. BJU Int 1999;84: [17] Genster HG, Madsen PO. Urinary tract infections following transurethral prostatectomy: with special reference to the use of antimicrobials. J Urol 1970;104: [18] Johnson JR, Kuskowski MA, Gajewski A, et al. Extended virulence genotypes and phylogenetic background of Escherichia coli isolates from patients with cystitis, pyelonephritis, or prostatitis. J Infect Dis 2005;191: [19] Shiina H, Himeno Y, Ishibe T. Organisms in the prostate and antibiotics in the treatment of postoperative infections. Urol Int 1992;48: [20] Madersbacher S, Alivizatos G, Nordling J, Sanz CR, Emberton M, de la Rosette JJ. EAU 2004 guidelines on assessment, therapy and follow-up of men with lower urinary tract symptoms suggestive of benign prostatic obstruction (BPH guidelines). Eur Urol 2004;46: Editorial Comment Magnus Grabe, Associate Professor of Urology, Department of Urology, Malmö University Hospital, Malmö, Sweden Magnus.grabe@skane.se The authors detect E. Coli in a few percent of patients who have undergone TURP for benign prostatic hyperplasia. The patients had no documented risk factors. The results must be viewed in the light of two observations. The first is the culture of prostate chips that has revealed bacterial growth in as many as 1/4 of the samples with E. Coli as the most frequently isolated organism [1]. More recently, Nickel et al. [2] detected bacterial growth in 28% of non-catheterized men of which 3.6% were E. Coli. The second is the finding of various grades and extent of inflammation of the prostate in almost all removed tissues [2,3]. Bacterial and viral invasions, urine reflux, stagnation of prostatic secretion, lower urinary tract obstruction have, among others, been incriminated. Which are then the factors responsible for the inflammation that launch the combat between the inflammatory cells, the pro- and anti-inflammatory cytokines [4], the reactive oxygen and nitrogen species and other enzymes that meet on the battlefield in the immediate surrounding of the glandular structure? Which are the other prokaryote species presenting a 16S rrna sequence involved in this warfare possibly responsible for the proliferative inflammatory atrophy that, eventually, could turn into prostatic intraepithelial neoplasia [5], thought to be the precursor of invasive prostate cancer? This study raises the question of the lack of virulence factors. Could it be so simple that any bacterial strain could easily invade the prostate ducts of the aging male and initiate an inflammatory response? A weakness in retrospective studies as the present one is the lack of reliable information on risk factors such as bacteriuria, indwelling catheter treatment, history of genitourinary infections and concomitant diseases. Further research must incorporate information on those assumed risk factors.

7 european urology 51 (2007) The questions are many. The new techniques open a fascinating field of research that could eventually lead to a better understanding of the causes of prostate diseases. References [1] Gorelick JI, Senterfit LB, Vaughan ED. Quantitative bacterial tissue cultures from 209 prostatectomy specimens. J Urol 1988;139: [2] Nickel JC, Downey IY, abd Boag S. Asymptomatic inflammation and/or infection in benign hyperplasia. BJU Int 1999;84: [3] Di Silverio F, Gentile V, De Matteis A, et al. Distribution of inflammation, pre-malignant lesions, incidental carcinoma in histological confirmed benign prostatic hyperplasia: a retrospective analysis. Eur Urol 2003;43: [4] Hochreiter WW. Male accessory gland infection: standardization of inflammatory parameters including cytokines. Andrologia 2003;35: [5] Nelson WG, De Marzo AM, Isaacs WB. Prostate cancer. N Engl J Med 2003;349: