European Urology

1. Introduction

The US Preventive Services Task Force recently reviewed the literature on prostate cancer screening and released an updated draft recommendation against prostate-specific antigen (PSA) screening [1]. Although it was demonstrated in 2009 that PSA screening reduces the incidence of metastatic prostate cancer and disease-specific mortality [2], the panel concluded that the harms outweigh the benefits. However, there are limited high-quality data on the harms of screening itself.

Because abnormal screening results lead to prostate biopsy for a histologic diagnosis, an examination of the complications from prostate biopsy is germane to the screening debate. Numerous studies have recently reported rising complication rates after prostate biopsy, particularly infectious complications.

In US Medicare participants from 1991 to 2007, Loeb et al. compared the 30-d hospital admission rate of 17 472 men who underwent prostate biopsy versus 134 977 randomly selected controls [3]. On multivariable analysis with age, race, region, year, and Charlson comorbidity score, the prostate biopsy group had more than double the odds of hospitalization with a primary diagnosis of infection compared with the control population (odds ratio [OR]: 2.26; 95% confidence interval [CI], 1.71–2.99; p < 0.0001). In addition, the rates of hospitalization for infectious reasons increased significantly over the study period in the biopsy group (p = 0.001). These results are likely attributable to increasing antimicrobial resistance, although culture data were not available in this claims-based analysis.

An earlier study from Canada found strikingly similar results in 41 682 men with negative prostate biopsy from an insurance database [4]. In 2005, infectious complications leading to hospitalization were reported in 3.6% of biopsies, a significant increase from 0.6% in 1996 (OR: 3.57; 95% CI, 1.9–4.8; p < 0.0001).

In light of these alarming reports, our objective was to perform a detailed investigation of febrile complications and hospital admissions after prostate biopsy in men from the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer (ERSPC). Specific aims of the study were (1) to ascertain whether the same concerning trends reported in the United States and Canada are also occurring in Europe, (2) to shed additional light on the risk-to-benefit ratio of prostate cancer screening through an investigation of the risks involved in the diagnostic work-up, and (3) to examine individual-level data on predictors and features of these complications to identify possible preventive strategies.

2. Patients and methods

The ERSPC was initiated in 1993 to study the effects of PSA screening on prostate cancer-specific mortality [2]. In the Rotterdam section, the screening protocol consisted of PSA testing every 4 yr. Transrectal ultrasound (TRUS)–guided biopsy was recommended for abnormal digital rectal examination/TRUS or a PSA level ≥4 (until May 1997), and thereafter for a PSA ≥3 ng/ml [5].

From 1993 to 2011, 10 474 lateralized sextant prostate biopsies were performed in the Rotterdam ERSPC. Prophylaxis consisted of trimethoprim-sulfamethoxazole until 2008, after which it was switched to ciprofloxacin given 2 h before and 4 h after PNB. A more extended duration of ciprofloxacin prophylaxis (5 additional days) was used in high-risk patients (insulin-dependent diabetes, steroid use, prosthetic device). For patients with a history of endocarditis or artificial cardiac valves, intravenous amoxicillin was given 1 h prior to the biopsy in addition to the standard prophylaxis.

Two weeks after biopsy, participants returned to the screening office to learn the pathology results. At this visit, urologists reviewed a standard questionnaire regarding biopsy-related complications with each patient, including specific questions regarding any fever or hospital admission. All patients were asked the same questions without any selection criteria, and these data were prospectively recorded in the study database, along with other demographic and clinical characteristics (eg, age, comorbid health conditions).

In the case of a hospitalization, more detailed information was obtained from the medical record, including admission to the intensive care unit and culture data (blood and urine) with resistance patterns. Because not all cultures were tested for exactly the same panel of antimicrobial sensitivities, rates of resistance are reported for the eight most commonly tested antibiotics as a fraction of the total number tested for that agent. Of note, norfloxacin was the fluoroquinolone that was primarily tested for sensitivity by our microbiology department.

The chi-square test was used to compare binary clinical covariates between participants with and without fever or hospital admission. The Shapiro-Wilk test was used to assess the normality of continuous variables. For variables with evidence of non-normal distribution (age, year), medians were presented and compared using the Wilcoxon rank-sum test for univariable analysis. Additionally, multivariable logistic regression models were used to examine predictors of fever or hospitalization after biopsy, including the following covariates: year of biopsy (continuous), age at biopsy (continuous), prostate enlargement (>40 cm³ on TRUS), diabetes (yes/no), history of prior prostatitis (yes/no), and heart disease (yes/no). SPSS v.17 and STATA v.11 were used for statistical analysis.

3. Results

Table 1 shows the clinical characteristics of the study population. Overall, fever and hospital admission were reported on 392 of 9241 questionnaires (4.2%) and 78 of 9198 questionnaires (0.8%), respectively. On univariate analysis, biopsy year and prostatic enlargement were significantly associated with febrile complications. There also tended to be a greater proportion of diabetic patients with fever after biopsy, although this did not reach statistical significance.

With respect to hospital admission after biopsy, there were no significant differences by age, diabetes, or history of prostatitis or heart disease. However, hospital admission was significantly associated with prostatic enlargement and later year of biopsy.

On multivariable analysis (Table 2), prostate enlargement and diabetes were significantly associated with an increased risk of fever after PNB, whereas later year of biopsy was the only factor significantly associated with an increased risk of hospital admission.

Although most fevers were managed on an outpatient basis, infection was the leading indication for hospital admission (n = 63; 81%). Only two patients required admission to the intensive care unit, and there were no biopsy-related deaths.

Urine cultures were available for 60 of these men, of which 38 (63.3%) ultimately came back positive. Escherichia coli was the predominant organism (n = 32; 84%), followed by Pseudomonas aeruginosa (n = 4), Klebsiella oxytoca (n = 1), and Sphingomonas paucimobilis with Enterococcus faecalis (n = 1).

Blood culture results were available for 56 men, of which 34 (60.7%) were positive. Of these men, 26 (76.5%) also had a positive urine culture. On blood culture, E coli was again the most common pathogen (n = 31; 91.2%), followed by Pseudomonas (n = 2) and Klebsiella (n = 1).

Figure 1 shows the antimicrobial susceptibility patterns for urine and blood cultures. Of positive urine and blood cultures, 78.9% and 94.1% were resistant to at least one of the eight most common antimicrobial agents tested. The highest rates of resistance were for trimethoprim-sulfamethoxazole and amoxicillin. Resistance to norfloxacin was observed in 5.6% of urine and 14.3% of blood cultures. Gentamicin and cefuroxime had the lowest frequency of resistance in blood cultures.

Fig. 1 Percentage of blood and urine cultures obtained at hospital admission after prostate biopsy with antimicrobial resistance.

4. Discussion

In the ERSPC, PSA screening was shown to reduce the incidence of metastatic disease at the time of diagnosis by 41% and prostate cancer–specific mortality by 20% in the intention-to-treat analysis at 9 yr [2]. After 14 yr of follow-up, the Gothenburg population-based screening trial reported a 44% relative decrease in prostate cancer mortality with PSA screening [6]. Despite these benefits, screening also has potential harms including overdiagnosis of potentially indolent disease [7] as well as unnecessary prostate biopsies triggered by false-positive screening results.

This is particularly concerning in light of growing evidence showing increasing hospitalizations for serious infectious complications after prostate biopsy. In this study, we examined the rates of biopsy-related febrile complications and hospital admissions in men from the ERSPC section Rotterdam to better elucidate the risk-to-benefit ratio of prostate cancer screening and detection.

Similar to prior studies from the United States and Canada [3] and [4], we observed a significant increase in the rates of hospitalization within 2 wk of prostate biopsy. Specifically, we found a 10% increase in the frequency of admissions, and most of these admissions were for infectious complications. Thus it is likely that the observed increase in hospitalizations was related to rising antimicrobial resistance.

Another interesting finding in our study was that men with prostatic enlargement and diabetes had an increased risk of febrile complications after prostate biopsy. Prior studies have shown an increased risk of biopsy-related infectious complications among men with greater comorbidities [3]. These results highlight the importance of judicious patient selection for PSA screening because these men may be less likely to benefit from early prostate cancer detection and also have a greater risk of complications from the diagnostic work-up. These results also suggest the potential value of multivariable risk assessment tools taking into account other clinical factors (eg, prostate volume) in addition to PSA for clinical decision making [8].

Conversely, it should be considered that the absolute frequency of hospital admissions was <1%. Only two cases involved admission to the intensive care unit, and there were no biopsy-related deaths. These results concur with prior studies showing that prostate biopsy is not associated with excess mortality [9]. Thus for healthy men with sufficient life expectancy to benefit from early prostate cancer detection, the risk of biopsy-related complications should not deter men from undergoing a recommended prostate biopsy.

Instead, measures should be implemented to further decrease the potential risks associated with prostate biopsy, beginning with careful patient selection for screening and biopsy. For appropriate patients undergoing biopsy, a detailed history should be performed with attention to risk factors for infectious complications (eg, diabetes) and for harboring resistant organisms (eg, recent hospitalization or antibiotic use [10] and [11]). Because the key trials on antimicrobial prophylaxis for prostate biopsy were conducted more than a decade ago and antimicrobial resistance patterns have changed [12], [13], and [14], additional studies are warranted to reevaluate the optimal regimen in the contemporary era.

Consistent with prior studies [10] and [11], E coli accounted for most of the positive cultures in our population, with high levels of antimicrobial resistance. Overall, trimethoprim-sulfamethoxazole and amoxicillin had the highest rates of resistance, suggesting that these agents should not be used for prostate biopsy prophylaxis. Conversely, the lowest rates of resistance were found for gentamicin and cefuroxime, suggesting possible utility for management of biopsy-related infectious complications. Nevertheless, culture data were only available for a limited subset of the population, and additional study of this issue is warranted in larger populations to identify the best antimicrobial agents for this setting.

For patients with risk factors for biopsy-related infectious complications, there is ongoing investigation into possible strategies to reduce these risks. One possibility is to obtain better coverage through the use of more extended antimicrobial prophylaxis [15]. An alternative possibility is the use of targeted antimicrobial prophylaxis. This paradigm was recently examined by Liss et al, who performed rectal swabs on patients scheduled for prostate biopsy [16]. Fluoroquinolone-resistant E coli were present in 22% of patients, including only one of five men with fever after biopsy. These results demonstrate that the presence of resistant organisms does not always translate into clinical infection. Therefore, additional investigation is necessary into the cost effectiveness of using culture swabs and targeted prophylaxis.

Several limitations of our study should be noted. First, sextant biopsy was performed in the Rotterdam ERSPC. It is unclear whether the rates of infectious complications would be different with a greater number of biopsy cores. However, a randomized trial of 6- versus 12-core biopsy previously reported no significant difference in febrile complications (p = 0.25) [17]. Also, because all of the biopsies were performed transrectally, data were not available to determine whether transperineal biopsy is associated with a lower risk of infectious complications.

Another limitation is that our population was primarily composed of white European men. Differences in health status among Rotterdam screening participants [18], as well as geographic differences in antimicrobial resistance patterns [14], may limit the generalizability of these findings. Furthermore, Loeb et al. previously reported an association between nonwhite race and an increased risk of infectious complications in American men [3]. Thus our findings require reevaluation in men from different ethnic backgrounds.

Finally, our method of biopsy prophylaxis was changed from trimethoprim-sulfamethoxazole to ciprofloxacin in 2008. In the subset of 390 men who have undergone biopsy since 2008, fever and hospital admission were reported in 11 (2.8%) and 5 (1.3%) men, respectively. Due to the small number of events during this period, we are not able to draw any definitive conclusions about the effect of this recent change in prophylaxis on resistance patterns.

Strengths of our study include the prospective collection of data on biopsy-related infectious complications and hospital admissions in a large population of men from a randomized screening trial. Advantages over other large studies on this issue include the standardized nature of the biopsy protocol and questionnaire used to assess complications that may reduce potential sources of confounding. Also, our study included a wider age group of men undergoing prostate biopsy compared with prior studies including Medicare participants ≥65 yr of age [3]. Finally, the availability of individual-level data on comorbid health conditions, type of antimicrobial prophylaxis, and culture results including antimicrobial susceptibility profiles provided the unique opportunity to evaluate possible strategies to reduce infectious complications.

5. Conclusions

Febrile complications occur in <5% of prostate biopsy procedures. Although most were managed on an outpatient basis, fever remained the leading cause of hospital admission within 2 wk of biopsy. Although the risk of hospitalization after prostate biopsy has increased over time, the overall frequency was 0.8%, and no biopsy-related deaths were observed. The risk-to-benefit ratio of prostate cancer screening and biopsy varies based on individual factors, highlighting the importance of individualized decision making. Although prostate biopsy may be associated with serious complications, this was a relatively rare event and should not by itself deter healthy young men who would benefit from early detection from pursuing a recommended biopsy. Ongoing investigation is warranted into strategies to continue to reduce the morbidity of PNB in contemporary screening populations.

Author contributions: Stacy Loeb 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: Loeb, van den Heuvel, Roobol.

Acquisition of data: Loeb, van den Heuvel, Zhu, Bangma, Schroder, Roobol.

Analysis and interpretation of data: Loeb, van den Heuvel, Zhu, Roobol.

Drafting of the manuscript: Loeb.

Critical revision of the manuscript for important intellectual content: van den Heuvel, Zhu, Bangma, Schroder, Roobol.

Statistical analysis: Loeb, van den Heuvel, Zhu, Roobol.

Obtaining funding: Bangma, Schroder, Roobol.

Administrative, technical, or material support: van den Heuvel, Zhu, Bangma, Schroder, Roobol.

Supervision: Bangma, Schroder, Roobol.

Other (specify): None.

Financial disclosures: I certify 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: The ERSPC Rotterdam is supported by the Dutch Cancer Society (KWF 94-869, 98-1657, 2002-277, 2006-3518, 2010-4800) and The Netherlands Organisation for Health Research and Development (ZonMW-002822820, 22000106, 50-50110-98-311, 62300035) that helped collect and manage the data. Stacy Loeb was supported by the Society of Women in Urology Elisabeth Pickett Research Award.