European Urology

1. Introduction

Prostate biopsy (P-Bx) is currently an essential procedure for prostate cancer diagnosis [1] and a frequently performed procedure in urology with an estimated 1 million cases each year in Europe. Following the introduction of transrectal ultrasound-guided P-Bx [2], this procedure has been widely accepted and optimised in recent years, continuously following the transrectal route [1]. A systematic review on randomised controlled studies (RCTs) on antibiotic prophylaxis in transrectal P-Bx showed a significant reduction in the risk of bacteriuria compared with placebo [3]. Reductions in symptomatic urinary tract infections (UTIs) and bacteremia were also seen but did not reach statistical significance [3]. Clinical guidelines therefore recommend antibiotic prophylaxis, typically with a oral fluoroquinolone, prior to transrectal P-Bx [4] and [5]. Several reports have recently suggested an increased rate of infective complications following transrectal P-Bx in both North America [6] and [7] and Europe [8]. The reasons for this increase and the factors associated with a higher susceptibility to infection remain largely unknown.

The aim of this international study was to evaluate the worldwide prevalence of infective complications after P-Bx and determine whether putative risk factors are associated with a higher rate of infection.

2. Material and methods

3. Results

A total of 84 centres in Africa (n = 2), Asia (n = 11), Europe (n = 67), and South-America (n = 4) participated in the GPIU prostate biopsy study (see appendix). The centres included a median of 4 evaluable patients for each year of participation (range: 1–25; standard deviation [SD]: ±6.7). A total of 702 men were included across the two study years (2010, n = 385; 2011, n = 317). Median patient age was 66 yr (range: 40–90; SD: ±8.0). Table 1 shows the biopsy performance parameters.

Outcome at 2 wk was available for 521 of 702 men (74.2%) (2010, n = 290; 2011, n = 231) (Table 2). Symptomatic UTIs were recorded in 27 of 521 patients (5.2%), with dysuria (n = 19), frequency (n = 18), urgency (n = 10), and prostate pain (n = 9) the most frequent symptoms. Urine culture was positive in 10 of 27 cases (Escherichia coli, n = 8; Pseudomonas species, n = 1; Klebsiella species, n = 1). Fluoroquinolone resistance was observed in 6 of 10 isolates. Febrile UTIs were present in 18 of 521 patients (3.5%), and 16 of 521 patients (3.1%) required hospitalisation due to infection. This included one patient admitted to the intensive care unit with uroseptic shock 1 d after P-Bx. In two men the infection had not resolved by the end of the 2-wk follow-up period. A total of 494 of 521 men (94.8%) showed no evidence of infective complications at up to 2 wk after biopsy. Symptomatic UTI was reported in 1 of the 2 patients submitted by centres in Africa, 3 of 57 submissions (5.3%) from centres in Asia, 21 of 441 submissions (4.8%) from European centres, and 2 of 21 submission (9.5%) from centres in South America. There was no statistical significant difference (p = 0.282) comparing infectious complications from Europe versus Africa, Asia, and South America.

To assess factors for increased infectious complications after P-Bx, patients with infectious complications (n = 27) were compared with patients without infectious complications (n = 494) (Table 2). In multivariate analysis no significant association between UTI and any investigated variable was found (for all variables p > 0.05).

4. Discussion

Millions of men undergo P-Bx throughout the world each year as part of the accepted diagnostic pathway for prostate cancer [6]. Recently retrospective studies from Canada [7] and the United States [6] and one prospectively performed study from Europe [8] reported an apparent increase in the incidence of infective complications after transrectal P-Bx. In this study we found that just over 5% of men experienced symptomatic UTI after P-Bx, which resulted in serious morbidity in about 70% of cases with systemic symptoms and admission to hospitals. The rate of systemic infection found in our study is higher than earlier reports in the literature where a rate of 1% was quoted in one large study [11]. This finding is important because the low specificity of PSA testing means that cancer detection rates for initial P-Bx range between 22.8% and 42.0% [1]. A diagnostic procedure with a generally <50% positive detection rate should be as safe as possible.

EAU guidelines classify transrectal P-Bx as a contaminated procedure and, if a urinary catheter or bacteriuria is present, as a dirty procedure [12] warranting antibiotic prophylaxis in all patients. Systematic review and meta-analysis of RCTs confirm that prophylaxis reduces the rate of bacteriuria following P-Bx, although the results for more clinically oriented outcomes such as symptomatic UTI are less certain [3]. These studies have so far focused on the use of fluoroquinolones and trimethoprim/sulfamethoxazole as suitable antibiotics for prophylaxis [3]. This is in line with pharmacokinetic studies in prostate tissue and prostatic secretions that confirm most orally administered fluoroquinolones achieve concentrations in the prostate that should be sufficient for the treatment of chronic bacterial prostatitis caused by susceptible pathogens [13]. High prostatic antibiotic concentrations are equally important in the prophylaxis of P-Bx. The superiority of fluoroquinolones was suggested by the results of a clinical pharmacokinetic study comparing ciprofloxacin (moderate prostatic tissue penetration) versus gentamicin (low prostatic tissue penetration) in prophylaxis against bacteremia following transrectal P-Bx. The study found that use of ciprofloxacin was associated with a lower incidence of bacteremia and clinical bloodstream infection, and this response correlated with higher prostate tissue levels of the drug [14]. There is as yet no consensus on the optimal duration of antibiotic prophylaxis for P-Bx. Our multinational study showed a median duration of 3 d, but there was no association between durations >1 d and the incidence of symptomatic UTI. EAU guidelines recommend single-dose prophylaxis for low-risk patients and prolonged courses of prophylaxis only in high-risk patients [5]; American Urological Association guidelines generally recommend prophylaxis for <24 h [4]. Because most men undergoing P-Bx are likely to have low levels of comorbidity, reducing the prophylaxis regimens to a single appropriately administered dose as recommended by the EAU guidelines would seem reasonable. This is important as part of the worldwide drive to reduce antibiotic selection pressure and thus decrease the emergence of antibiotic resistance.

Procedure-specific factors for a higher risk of infective complications after P-Bx are still not well defined, and our study was unable to add any further clarity on this issue [12]. It would seem appropriate to check a midstream urine culture prior to the procedure to rule out asymptomatic bacteriuria, which is recommended in the EAU guideline [5]. It is also recommended that in the presence of asymptomatic bacteriuria the prophylactic antibiotic regimen should include an agent tested susceptible to cover the urinary bacteria [5]. In our study, recommended preoperative patient evaluation by urine culture prior to biopsy [5] was performed in only 164 of 702 patients (23.4%) (Table 1). Adherence to guidelines could certainly be improved in this respect.

All other evaluated possible factors for increased infective complications were not significantly different between patients having symptomatic UTI and those who did not; however, diabetes as a risk factor was not investigated in this study. In a previous study it was found that a larger prostate and diabetes were significantly associated with infective complications after P-Bx [8].

The observational epidemiologic design of our study collecting data at two time points in 2 yr prevented us from looking at changing patterns of infective complications and bacterial resistance over time. Further data collections at annual intervals over the coming years will allow us to achieve this goal. Our study also had insufficient power to be certain of identifying whether or not some of the risk factors examined were associated with the incidence of symptomatic UTI. The small sample size per participating site and in comparison with other studies therefore may have limited the conclusions from our study. Possible risk factors evaluating comorbidities such as diabetes, steroid use, or heart disease have not been evaluated, which is a limitation and will be taken up in subsequent study years. Our data on bacterial isolates do provide provisional support for the faecal carriage of fluoroquinolone-resistant bacteria as a risk factor for infective complications. A previous prospective study in consecutive patients undergoing P-Bx found that 22.0% of patients harboured ciprofloxacin-resistant E coli strains [15]. The study also found that faecal carriage of fluoroquinolone-resistant E coli strains was a significant risk factor for infective complications after P-Bx [15]. Specific fluoroquinolone-resistant E coli strains (eg, E coli ST131) could be detected as an important cause of sepsis after P-Bx [16]. It has also been shown that faecal fluoroquinolone-resistant E coli can even be selected by a single oral dose of 500 mg ciprofloxacin given for the prophylaxis of transrectal P-Bx [17]. The use of fluoroquinolones in the 6 mo before biopsy was also shown to be associated with an increased risk of faecal carriage of fluoroquinolone-resistant E coli strains [15]. Our study adds further credence to the emerging possibility that faecal carriage of fluoroquinolone-resistant E coli strains represents a significant risk factor classifying patients into a high-risk group for infective complications after P-Bx, particularly if, as seems likely, fluoroquinolones remain the most frequently used agents for prophylaxis. The rate of fluoroquinolone resistance amongst isolates from men with symptomatic UTI after P-Bx in the present study was particularly high (60%) compared with rates seen in other hospital settings that ranged from 22.7% to 30.8% [18]. Faecal fluoroquinolone-resistant bacteria as a risk factor could certainly explain the increase of infective complications after P-Bx, paralleling the worldwide increase of fluoroquinolone resistance in enterobacteria [10], [19], and [20].

Strategies capable of decreasing infective complications after P-Bx must therefore be developed and evaluated so as not to dissuade healthy men who would benefit from early prostate cancer management from undergoing P-Bx when clinically indicated. Several strategies are currently debated, such as risk assessment to select patients at higher risk for infective complications, such as microbiological sampling of the faecal flora prior to biopsy to identify susceptibility to specific agents [21], change of biopsy route (eg, perineal P-Bx), and the evaluation of alternative antibiotics with improved susceptibility to be used for prophylaxis [22]. Antibiotic susceptibility is only one of the criteria for recommending an alternative antibiotic. The pharmacokinetic distribution of the antibiotic within the prostate is equally important, as well as the suitability for this indication demonstrated in a clinical trial.

5. Conclusions

In this prospective multinational study on infective complications after P-Bx, a significant rate of infective complications after P-Bx was observed to be present worldwide. Our study supports the findings of others in suggesting that the presence of faecal fluoroquinolone-resistant bacteria is the most important risk factor. If this is the case, strategies to identify fluoroquinolone-resistant bacteria should be sought so as to decrease infective complications after P-Bx. This important and yearly performed ongoing study will deliver more data in the future and thus be able to address some of these important issues using a global perspective.

Author contributions: Florian M.E. Wagenlehner had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Wagenlehner, van Oostrum, Çek, Naber, Bjerklund-Johansen.

Acquisition of data: Wagenlehner, van Oostrum, Çek, Grabe, Tandogdu, Tenke, Naber, Bjerklund-Johansen.

Analysis and interpretation of data: Wagenlehner, van Oostrum, Çek, Grabe, Tandogdu, Tenke, Pickard, Wullt, Pilatz, Weidner, Naber, Bjerklund-Johansen.

Drafting of the manuscript: Wagenlehner.

Critical revision of the manuscript for important intellectual content: Wagenlehner, van Oostrum, Çek, Grabe, Tandogdu, Tenke, Pickard, Wullt, Weidner, Naber, Bjerklund-Johansen.

Statistical analysis: Wagenlehner, van Oostrum, Pilatz.

Obtaining funding: Wagenlehner, van Oostrum, Çek, Grabe, Tenke, Naber, Bjerklund-Johansen.

Administrative, technical, or material support: van Oostrum.

Supervision: Wagenlehner, Naber, Bjerklund-Johansen.

Other (specify): None.

Financial disclosures: Florian M.E. Wagenlehner certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.

Funding/Support and role of the sponsor: This study was organised by the board of the ESIU and endorsed and sponsored by the EAU and performed in collaboration with the Asian Association of UTI and STD, the International Society of Chemotherapy for Infection and Cancer, and the Interregional Association of Clinical Microbiology and Antimicrobial Chemotherapy.

Acknowledgement statement: The GPIU study group is most grateful to the altruistic contributions of the GPIU investigators (see appendix).