REVIEWS Reducing infection rates after prostate biopsy Florian M. E. Wagenlehner, Adrian Pilatz, Przemyslaw Waliszewski, Wolfgang Weidner and Truls E. Bjerklund Johansen Abstract | Over the years, prostate biopsy has become the gold-standard technique for diagnosing prostate carcinoma. Worldwide, several million prostate biopsies are performed every year, most commonly using the transrectal approach. Preoperative antibiotic prophylaxis with fluoroquinolones has been shown to be effective for reducing infection rates. However, in recent years, an increase in febrile infection rates after transrectal prostate biopsy (from 1% to 4%) has been reported in retrospective and prospective studies. The predominant risk factor for infection seems to be the presence of fluoroquinolone-resistant bacteria in faeces. Patients at risk of fluoroquinolone resistance should receive carefully selected antibiotics at sufficient concentrations to be effective. Targeted prophylaxis after rectal flora swabbing has been shown to be efficacious compared with empirical antibiotic prophylaxis. Several forms of bowel preparations are under investigation, although none have yet been shown to significantly reduce infection rates. Perineal prostate biopsy is currently being evaluated as a strategy for preventing the inoculation of rectal flora, but limited data support this approach at present. Wagenlehner, F. M. E. et al. Nat. Rev. Urol. 11, 80–86 (2014); published online 14 January 2014; doi:10.1038/nrurol.2013.322
Introduction
Clinic for Urology, Paediatric Urology and Andrology, Justus‑Liebig-University, Rudolf-Buchheim Strasse 7, 35392 Giessen, Germany (F. M. E. Wagenlehner, A. Pilatz, P. Waliszewski, W. Weidner). Urologic Clinic, Oslo University Hospital, P. O. Box 4959 Nydalen, N‑0424 Oslo, Norway (T. E. Bjerklund Johansen). Correspondence to: F. M. E. Wagenlehner [email protected]
Diagnosis of prostate cancer relies on histopathological proof of neoplastic tissue in prostate biopsies or other prostate tissue samples. Although prostate biopsy can be performed transperineally, the transrectal route is more frequently performed.1 Until recently, transrectal prostate biopsy was considered to be a very safe pro cedure, associated with only rare occurrences of infec tious complications, such as UTI, male genital infection, prostatitis, epididymo-orchitis, 2,3 and urosepsis. 1–3 However, reports published within the past few years suggest that the complication rate might be higher than previously thought.4–6 Even if the rate of infectious com plications is regarded as low, the total number of patients developing complications should be considered signifi cant, as millions of men undergo prostate biopsy every year worldwide.4 Patient-specific risk factors for infectious compli cations after transrectal prostate biopsy have not been conclusively evaluated, although factors such as increased comorbidity scores, untreated asymptomatic bacteriuria, indwelling urinary catheters, history of prostatitis or UTI, diabetes mellitus, presence of bladder stones, and prostate size have all been reported to increase infection rates to some extent.4,5,7–10 Procedure-specific risk factors that have been identified in some studies, but not others, Competing interests F. M. E. Wagenlehner declares associations with the following companies: Astellas, AstraZeneca, Bionorica, Calixa, Cerexa, Cernelle, Cubist, Glaxo-Smith Kline, Janssen-Cilag, Johnson & Johnson, Merlion, OM‑Pharma, Lilly Pharma, Pharmacia, Pierre Fabre, Rosen-Pharma, Sanofi-Aventis, Serag Wiessner, Strathmann, and Zambon. See the article online for full details of the relationships. The other authors declare no competing interests.
80 | FEBRUARY 2014 | VOLUME 11
include number of biopsy cores11,12 and number of repeat biopsy procedures.13,14 Although these risk factors are not well established, the introduction of prostatic MRI and the use of MRI–transrectal ultrasonography (MRI–TRUS)fusion-guided 3D targeted biopsies is likely to reduce the number of procedures needed, especially in patients with tumours in the anterior part of the prostate.15 The anterior prostate is difficult to biopsy with the trans rectal approach, but this region can be reached with MRI guidance and transperineal sampling. In this Review, we discuss strategies to reduce severe symptomatic infections and urosepsis after prostate biopsy.
Incidence of postbiopsy infection A retrospective study of 75,190 men in Ontario, Canada, who underwent prostate biopsies between 1996 and 2005 reported a significant increase in the infection-related 30-day hospital readmission rate after biopsy from 1.0% in 1996 to 4.1% in 2005.6 The 30-day mortality rate after transrectal prostate biopsy in this study was 0.09% in 1996 and 0.11% in 2005—an increase that was not found to be statistically significant.6 In another large retro spective case-control study, data from 17,472 Medicareinsured patients who underwent prostate biopsy in the USA from 1991 to 2007 were analysed, revealing a 30-day readmission rate of 6.9% and a significant increase in the rate of readmissions caused by infectious complications (from approximately 0.4% in 1991 to 1.1% in 2007).4 A large prospective study of the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer (ERSPC) evaluated 10,474 prostate biopsies per formed from 1993 to 2011 and established a febrile infec tion rate of 4.2% and a hospital readmission rate of 0.8% (81% of which were caused by infection).5
www.nature.com/nrurol © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Key points ■■ Rates of infectious complications after transrectal prostate biopsies have increased over the past 15 years ■■ The predominant risk factor for postbiopsy infection seems to be exposure to faecal fluoroquinolone-resistant bacteria ■■ Every patient should undergo a preoperative assessment, including history of fluoroquinolone intake over the past 6 months, and a urine culture ■■ Choice of empirical antibiotic prophylaxis should take regional resistance rates into consideration, as well as the pharmacokinetic and pharmacodynamic characteristics of each antibiotic ■■ Novel strategies to reduce rates of infectious complications after transrectal prostate biopsies include targeted prophylaxis after sampling of the rectal flora, bowel preparation, and perineal prostate biopsy
The European Section for Infections in Urology (ESIU) of the European Association of Urology (EAU) performed a prospective prevalence study between 2010 and 2011— known as the Global Prevalence Study of Infections in Urology (GPIU)—specifically to assess rates of infectious complications after prostate biopsy.12 Of the 521 men who enrolled, 5.2% had a symptomatic UTI and 3.5% had a febrile UTI after prostate biopsy. The rate of hospital readmissions resulting from infectious complications was 3.1%.12 Overall, the average rate of severe febrile UTI after transrectal prostate biopsy reported in these recent studies is approximately 3%,2,4–6,12 with reported frequen cies of severe sepsis and septic shock ranging from 0% to 0.6%.2,12 In some studies, a mortality rate of 0.1–0.2% has been reported.2,6 In summary, symptomatic and febrile UTI occur in approximately 4% of patients who receive transrectal prostate biopsy, leading to hospital readmis sion in approximately 0.8% to 3.1% of patients, depending on the health-care system. By contrast, infection rates after transperineal prostate biopsy are rarely investigated. A single-centre prospec tive evaluation of 3,000 patients exclusively undergoing transperineal prostate biopsy from 2002 to 2012 showed clinical complications in 40.2%, with 9.1% needing an emergency department visit and 1.2% requiring hospital admission after intervention.16 UTI was the reason for hospital readmission in 0.7% of patients.16 The economic impact of the increasing rate of infectious complications after prostate biopsy has been evaluated for hospitals in England and Wales, where 72,500 prostate biopsies are performed annually and approximately 2,700 patients are readmitted to hospital as a result of infectious compli cations.17 With an estimated cost of £4,260 per patient readmitted, the total readmission cost for patients with infectious complications after prostate biopsy is estimated to be approximately £11.1 million per year.17
Strategies to reduce infection Changes to biopsy indications At present, prostate biopsy is indicated by a repeatedly elevated blood PSA value or suspicious digital rectal examination (DRE) of the prostate. 18 The therapeutic consequences of prostate biopsy should be considered, taking into account age and comorbidities. A single elevated PSA value should not be the only indication leading to prostate biopsy; this should be repeated under
optimal conditions (for example, no prior manipulation of the prostate, no prior ejaculation, and excluded infec tion) using an identical method of testing.18,19 It has been shown that 21–37% of patients without prostate cancer will have at least one abnormal elevated PSA test, depend ing on the cut-off value used, which will return to normal in approximately half of these patients.20 The major uro logical associations, including the AUA, the EAU, and the German Urological Association (DGU), all agree that the ratio of benefit to harm needs to be improved for prostate cancer management.18,19,21 Repeat biopsy within 6 months is indicated when the PSA value rises or is constantly elevated, if the DRE is suspicious, if atypical small acinar proliferation (ASAP) has been described in the initial pathological specimen, or if extensive prostatic intraepithelial neoplasia (PIN) is found (present in at least four biopsy cores). High-grade PIN in a single specimen is no indication for repeat pros tate biopsy.18,19 More than two repeat biopsies should be undertaken with caution.
Preoperative assessment Preoperative assessment includes an evaluation of poten tial risk factors for postbiopsy infection, such as a high comorbidity index,4 inadequately treated diabetes melli tus,3,11,22 an enlarged prostate,5 untreated bacteriuria,8 a history of bacterial prostatitis or recent urogenital infec tion,10 an indwelling urinary catheter,11 and bladder stones.10 The most important risk factor during pre operative evaluation is the presence of fluoroquinolone- resistant pathogens in the rectal flora.2,3,12,23,24 In various studies, fluoroquinolone-resistant bacteria have been detected in 50–90% of patients with symptomatic infections after biopsy.2,3,12,23,25–27 Indeed, Lindstedt et al.8 found that all patients with untreated asymptomatic bacteriuria before biopsy developed febrile UTI or sepsis, suggesting that patients should be routinely tested for asymptomatic bacteriuria and, if present, should be given antibiotics before biopsy. However, in the GPIU study, a urine culture was per formed before biopsy in just 23.4% of patients.12 In this study, 6 of 10 isolates cultured from patients with infec tious complications after biopsy were fluoroquinolone- resistant bacteria.12 3.5% of 521 patients had febrile UTI, and one patient presented with septic shock.12 By com parison, in a study by Steensels and colleagues,24 236 consecutive patients were screened for fluoroquinoloneresistant bacteria by rectal swab before prostate biopsy and all patients received fluoroquinolone prophylaxis according to international guidelines at that time, when fluoroquinolone prophylaxis was recommended.19,21,28,29 In 22% (n = 58) of patients, fluoroquinolone-resistant bacteria were cultured from the rectal flora. Infectious complications were found in 12% of these 58 patients, but in none of the patients with fluoroquinolone-susceptible pathogens (n = 178).24 Of the seven patients with infec tious complications, six patients had sepsis and one had epididymitis.24 Molecular investigations have shown that certain multiresistant clones of Escherichia coli (for example,
NATURE REVIEWS | UROLOGY
VOLUME 11 | FEBRUARY 2014 | 81 © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Prostate biopsy indicated
Assessment of faecal fluoroqinoline-resistant bacteria by: History Microbiology*
Risk of faecal fluoroqinoline-resistant bacteria Yes
No
Alternative antibiotics: Trimethoprim/sulfamethoxazole* Cephalosporines* (Acly)aminopenicillines/ beta-lactamase inhibitors*
Fluoroqinolones
Figure 1 | Algorithm for reducing rates of infectious complications when performing transrectal prostate biopsy. Assessment of faecal fluoroquinolone-resistant bacteria is based on history of fluroquinonlone intake over recent months, travel history to regions of high fluoroquinolone resistance, or microbial sampling of the faecal flora. *Not sufficiently evaluated in clinical studies; multidrug-resistant bacteria might be present.
ST131) are frequently found in patients with bacterae mia after transrectal prostate biopsy.23 These bacterial clones are also commonly resistant to gentamicin, cotrim oxazole, and ciprofloxacin.23 As the risk of harbouring fluroquinolone-resistant bacteria in the faeces is signifi cantly increased in patients who have received fluoro quinolones within the past 6 months or after international travel to countries with high levels of antibiotic resist ance,24–27 antibiotic and travel history should be assessed before biopsy (Figure 1).
Antibiotic prophylaxis Infections following transrectal prostate biopsy are most commonly associated with Gram-negative entero bacteria, especially E. coli.2,3,12,22–27,30 Anaerobic bacteria and enterococci have been identified as causative agents in some, but not many, studies.24 Systematic reviews of studies mainly performed in the 1990s provide evi dence to show that antibiotic prophylaxis significantly reduces rates of symptomatic bacteriuria, UTI, bacterae mia, and hospitalization.1,31,32 Choice of prophylactic antibiotic is guided by the expected bacterial spectrum at the operative site, the expected antibiotic suscepti bility, the pharmacokinetic and pharmacodynamic properties of each antibiotic, and the predicted efficacy (based on clinical trial data) of each antibiotic for the respective indication.33 Antibiotic resistance surveillance No resistance surveillance data are available for the anti biotic resistance of the entire rectal flora microbiome. However, surveillance studies have reported antibiotic resistance for specific specimens in different geographi cal regions. The European Centre for Disease Control (ECDC) regularly publishes data from the European Antimicrobial Resistance Surveillance Network (EARSnet) database regarding the resistance of blood culture 82 | FEBRUARY 2014 | VOLUME 11
isolates taken from patients from most European coun tries. Fluoroquinolone resistance of E. coli bloodstream isolates ranges from 10–40% in Europe, with an average resistance of approximately 20%.34 In North America, ciprof loxacin resistance has reached 12% for E. coli strains that cause bloodstream infections.35 The GPIU of the European Section for Infection in Urology (ESIU) has been collecting prevalence and resistance data for the past 10 years in 1,395 infected urological patients.36 According to this database, approximately 45% of E. coli strains are resistant to fluoroquinolone, although this rate is subject to significant regional variation.36 In some regions, multi resistant bacteria, such as extended-spectrum β-lactamaseproducing enterobacteria, must also be considered,2 with carbapenemase-producing bacteria represent ing the latest threat in certain regions of the world. 37 Different types of carbapenemases, such as the KPC type from the USA, Israel, Greece, Puerto Rico, and Colombia, the VIM type from the southern Mediterranean region, and the NDM-1 type from the Indian subcontinent have reportedly been introduced into European health-care systems via cross‑border patient transport.37 Minimal inhibitory concentration (MIC) testing— defined as the antimicrobial concentration needed to inhibit visible growth—of feacal pathogens in patients undergoing prostate biopsy has shown high-level fluoro quinolone resistance, with MIC values of ≥32 mg/l for 65% of fluoroquinolone-resistant strains.24 This finding is in agreement with a recent study of the molecular changes in fluoroquinolone-resistant pathogens in patients under going transrectal prostate biopsy, which showed that diverse combinations of chromosomal mutations lead to high-level fluoroquinolone resistance.38 Pharmacokinetic and pharmacodynamic properties In general, the pharmacokinetic properties of antibiotics used for prophylaxis, including those used before prostate biopsy and transrectal prostate biopsy, have been insuffi ciently evaluated. A comparative study performed in 1991 showed significantly higher rates of bacteraemia (37% versus 7%) and symptomatic bacteraemia (22% versus 0%) when patients receiving transrectal prostate biopsy were prophylactically treated with gentamicin rather than ciprofloxacin.39 Although, at that time, resistance was not a known problem for either of these antibiotics, the choice of antibiotic clearly had a substantial impact on the prevention of bacteraemia, most probably resulting from the marked pharmacokinetic differences between the two substances. Aminoglycosides generally exhibit a very low volume of distribution (which is an indicator of tissue penetration) compared with fluoroquinolones (which have high volumes of distribution), leading to low concentrations of gentamicin (an aminoglycoside) and relatively high concentrations of ciprofloxacin (a fluoro quinolone) in the prostate.40,41 The results of this study suggest that antibiotic concentration in the prostate might be more important than blood concentration at the time of biopsy. The exact prostate concentration needed for drug efficacy is, however, not yet known and is likely to be dependent on the MIC of the pathogen.
www.nature.com/nrurol © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS The prostate exhibits anatomical barriers to the distri bution of antimicrobial molecules. The prostate gland is perfused by nonfenestrated capillaries and the pros tatic epithelium lacks active transport mechanisms for drug uptake.42 Pharmacokinetic investigations of anti biotics have been performed in prostate tissue, as well as in prostate secretions, although data from studies in prostate tissue should be interpreted with caution, owing to potential contamination with urine or blood.43 Of all the antibiotics, fluoroquinolones (Box 1) and macrolides have been the most thoroughly investigated and exhibit the highest ratios of prostate secretion or tissue concen tration to serum concentration, indicating good prostate penetration;41,43–45 data for other antibiotics are somewhat scarce. β-lactam antibiotics, such as piperacillin, cefpo doxime, meropenem, and doripenem, reach concentra tions in the prostate that are just 10–20% of those in the serum,44–47 whereas the prostatic tissue to serum ratio of amikacin is approximately 50%,44 and, in one study, con centrations of trimethoprim were higher in the prostate than in the serum.48 As such, it is difficult to establish the concentration needed for prophy laxis. The same principles used for treating bacterial prostatitis are likely to apply, although it is worth noting that these are also poorly defined. One study of levofloxacin treatment for chronic bacterial prostatitis demonstrated a significantly higher bacterial eradication rate of 79% when the MIC of the pathogen was
Introduction
Clinic for Urology, Paediatric Urology and Andrology, Justus‑Liebig-University, Rudolf-Buchheim Strasse 7, 35392 Giessen, Germany (F. M. E. Wagenlehner, A. Pilatz, P. Waliszewski, W. Weidner). Urologic Clinic, Oslo University Hospital, P. O. Box 4959 Nydalen, N‑0424 Oslo, Norway (T. E. Bjerklund Johansen). Correspondence to: F. M. E. Wagenlehner [email protected]
Diagnosis of prostate cancer relies on histopathological proof of neoplastic tissue in prostate biopsies or other prostate tissue samples. Although prostate biopsy can be performed transperineally, the transrectal route is more frequently performed.1 Until recently, transrectal prostate biopsy was considered to be a very safe pro cedure, associated with only rare occurrences of infec tious complications, such as UTI, male genital infection, prostatitis, epididymo-orchitis, 2,3 and urosepsis. 1–3 However, reports published within the past few years suggest that the complication rate might be higher than previously thought.4–6 Even if the rate of infectious com plications is regarded as low, the total number of patients developing complications should be considered signifi cant, as millions of men undergo prostate biopsy every year worldwide.4 Patient-specific risk factors for infectious compli cations after transrectal prostate biopsy have not been conclusively evaluated, although factors such as increased comorbidity scores, untreated asymptomatic bacteriuria, indwelling urinary catheters, history of prostatitis or UTI, diabetes mellitus, presence of bladder stones, and prostate size have all been reported to increase infection rates to some extent.4,5,7–10 Procedure-specific risk factors that have been identified in some studies, but not others, Competing interests F. M. E. Wagenlehner declares associations with the following companies: Astellas, AstraZeneca, Bionorica, Calixa, Cerexa, Cernelle, Cubist, Glaxo-Smith Kline, Janssen-Cilag, Johnson & Johnson, Merlion, OM‑Pharma, Lilly Pharma, Pharmacia, Pierre Fabre, Rosen-Pharma, Sanofi-Aventis, Serag Wiessner, Strathmann, and Zambon. See the article online for full details of the relationships. The other authors declare no competing interests.
80 | FEBRUARY 2014 | VOLUME 11
include number of biopsy cores11,12 and number of repeat biopsy procedures.13,14 Although these risk factors are not well established, the introduction of prostatic MRI and the use of MRI–transrectal ultrasonography (MRI–TRUS)fusion-guided 3D targeted biopsies is likely to reduce the number of procedures needed, especially in patients with tumours in the anterior part of the prostate.15 The anterior prostate is difficult to biopsy with the trans rectal approach, but this region can be reached with MRI guidance and transperineal sampling. In this Review, we discuss strategies to reduce severe symptomatic infections and urosepsis after prostate biopsy.
Incidence of postbiopsy infection A retrospective study of 75,190 men in Ontario, Canada, who underwent prostate biopsies between 1996 and 2005 reported a significant increase in the infection-related 30-day hospital readmission rate after biopsy from 1.0% in 1996 to 4.1% in 2005.6 The 30-day mortality rate after transrectal prostate biopsy in this study was 0.09% in 1996 and 0.11% in 2005—an increase that was not found to be statistically significant.6 In another large retro spective case-control study, data from 17,472 Medicareinsured patients who underwent prostate biopsy in the USA from 1991 to 2007 were analysed, revealing a 30-day readmission rate of 6.9% and a significant increase in the rate of readmissions caused by infectious complications (from approximately 0.4% in 1991 to 1.1% in 2007).4 A large prospective study of the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer (ERSPC) evaluated 10,474 prostate biopsies per formed from 1993 to 2011 and established a febrile infec tion rate of 4.2% and a hospital readmission rate of 0.8% (81% of which were caused by infection).5
www.nature.com/nrurol © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Key points ■■ Rates of infectious complications after transrectal prostate biopsies have increased over the past 15 years ■■ The predominant risk factor for postbiopsy infection seems to be exposure to faecal fluoroquinolone-resistant bacteria ■■ Every patient should undergo a preoperative assessment, including history of fluoroquinolone intake over the past 6 months, and a urine culture ■■ Choice of empirical antibiotic prophylaxis should take regional resistance rates into consideration, as well as the pharmacokinetic and pharmacodynamic characteristics of each antibiotic ■■ Novel strategies to reduce rates of infectious complications after transrectal prostate biopsies include targeted prophylaxis after sampling of the rectal flora, bowel preparation, and perineal prostate biopsy
The European Section for Infections in Urology (ESIU) of the European Association of Urology (EAU) performed a prospective prevalence study between 2010 and 2011— known as the Global Prevalence Study of Infections in Urology (GPIU)—specifically to assess rates of infectious complications after prostate biopsy.12 Of the 521 men who enrolled, 5.2% had a symptomatic UTI and 3.5% had a febrile UTI after prostate biopsy. The rate of hospital readmissions resulting from infectious complications was 3.1%.12 Overall, the average rate of severe febrile UTI after transrectal prostate biopsy reported in these recent studies is approximately 3%,2,4–6,12 with reported frequen cies of severe sepsis and septic shock ranging from 0% to 0.6%.2,12 In some studies, a mortality rate of 0.1–0.2% has been reported.2,6 In summary, symptomatic and febrile UTI occur in approximately 4% of patients who receive transrectal prostate biopsy, leading to hospital readmis sion in approximately 0.8% to 3.1% of patients, depending on the health-care system. By contrast, infection rates after transperineal prostate biopsy are rarely investigated. A single-centre prospec tive evaluation of 3,000 patients exclusively undergoing transperineal prostate biopsy from 2002 to 2012 showed clinical complications in 40.2%, with 9.1% needing an emergency department visit and 1.2% requiring hospital admission after intervention.16 UTI was the reason for hospital readmission in 0.7% of patients.16 The economic impact of the increasing rate of infectious complications after prostate biopsy has been evaluated for hospitals in England and Wales, where 72,500 prostate biopsies are performed annually and approximately 2,700 patients are readmitted to hospital as a result of infectious compli cations.17 With an estimated cost of £4,260 per patient readmitted, the total readmission cost for patients with infectious complications after prostate biopsy is estimated to be approximately £11.1 million per year.17
Strategies to reduce infection Changes to biopsy indications At present, prostate biopsy is indicated by a repeatedly elevated blood PSA value or suspicious digital rectal examination (DRE) of the prostate. 18 The therapeutic consequences of prostate biopsy should be considered, taking into account age and comorbidities. A single elevated PSA value should not be the only indication leading to prostate biopsy; this should be repeated under
optimal conditions (for example, no prior manipulation of the prostate, no prior ejaculation, and excluded infec tion) using an identical method of testing.18,19 It has been shown that 21–37% of patients without prostate cancer will have at least one abnormal elevated PSA test, depend ing on the cut-off value used, which will return to normal in approximately half of these patients.20 The major uro logical associations, including the AUA, the EAU, and the German Urological Association (DGU), all agree that the ratio of benefit to harm needs to be improved for prostate cancer management.18,19,21 Repeat biopsy within 6 months is indicated when the PSA value rises or is constantly elevated, if the DRE is suspicious, if atypical small acinar proliferation (ASAP) has been described in the initial pathological specimen, or if extensive prostatic intraepithelial neoplasia (PIN) is found (present in at least four biopsy cores). High-grade PIN in a single specimen is no indication for repeat pros tate biopsy.18,19 More than two repeat biopsies should be undertaken with caution.
Preoperative assessment Preoperative assessment includes an evaluation of poten tial risk factors for postbiopsy infection, such as a high comorbidity index,4 inadequately treated diabetes melli tus,3,11,22 an enlarged prostate,5 untreated bacteriuria,8 a history of bacterial prostatitis or recent urogenital infec tion,10 an indwelling urinary catheter,11 and bladder stones.10 The most important risk factor during pre operative evaluation is the presence of fluoroquinolone- resistant pathogens in the rectal flora.2,3,12,23,24 In various studies, fluoroquinolone-resistant bacteria have been detected in 50–90% of patients with symptomatic infections after biopsy.2,3,12,23,25–27 Indeed, Lindstedt et al.8 found that all patients with untreated asymptomatic bacteriuria before biopsy developed febrile UTI or sepsis, suggesting that patients should be routinely tested for asymptomatic bacteriuria and, if present, should be given antibiotics before biopsy. However, in the GPIU study, a urine culture was per formed before biopsy in just 23.4% of patients.12 In this study, 6 of 10 isolates cultured from patients with infec tious complications after biopsy were fluoroquinolone- resistant bacteria.12 3.5% of 521 patients had febrile UTI, and one patient presented with septic shock.12 By com parison, in a study by Steensels and colleagues,24 236 consecutive patients were screened for fluoroquinoloneresistant bacteria by rectal swab before prostate biopsy and all patients received fluoroquinolone prophylaxis according to international guidelines at that time, when fluoroquinolone prophylaxis was recommended.19,21,28,29 In 22% (n = 58) of patients, fluoroquinolone-resistant bacteria were cultured from the rectal flora. Infectious complications were found in 12% of these 58 patients, but in none of the patients with fluoroquinolone-susceptible pathogens (n = 178).24 Of the seven patients with infec tious complications, six patients had sepsis and one had epididymitis.24 Molecular investigations have shown that certain multiresistant clones of Escherichia coli (for example,
NATURE REVIEWS | UROLOGY
VOLUME 11 | FEBRUARY 2014 | 81 © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Prostate biopsy indicated
Assessment of faecal fluoroqinoline-resistant bacteria by: History Microbiology*
Risk of faecal fluoroqinoline-resistant bacteria Yes
No
Alternative antibiotics: Trimethoprim/sulfamethoxazole* Cephalosporines* (Acly)aminopenicillines/ beta-lactamase inhibitors*
Fluoroqinolones
Figure 1 | Algorithm for reducing rates of infectious complications when performing transrectal prostate biopsy. Assessment of faecal fluoroquinolone-resistant bacteria is based on history of fluroquinonlone intake over recent months, travel history to regions of high fluoroquinolone resistance, or microbial sampling of the faecal flora. *Not sufficiently evaluated in clinical studies; multidrug-resistant bacteria might be present.
ST131) are frequently found in patients with bacterae mia after transrectal prostate biopsy.23 These bacterial clones are also commonly resistant to gentamicin, cotrim oxazole, and ciprofloxacin.23 As the risk of harbouring fluroquinolone-resistant bacteria in the faeces is signifi cantly increased in patients who have received fluoro quinolones within the past 6 months or after international travel to countries with high levels of antibiotic resist ance,24–27 antibiotic and travel history should be assessed before biopsy (Figure 1).
Antibiotic prophylaxis Infections following transrectal prostate biopsy are most commonly associated with Gram-negative entero bacteria, especially E. coli.2,3,12,22–27,30 Anaerobic bacteria and enterococci have been identified as causative agents in some, but not many, studies.24 Systematic reviews of studies mainly performed in the 1990s provide evi dence to show that antibiotic prophylaxis significantly reduces rates of symptomatic bacteriuria, UTI, bacterae mia, and hospitalization.1,31,32 Choice of prophylactic antibiotic is guided by the expected bacterial spectrum at the operative site, the expected antibiotic suscepti bility, the pharmacokinetic and pharmacodynamic properties of each antibiotic, and the predicted efficacy (based on clinical trial data) of each antibiotic for the respective indication.33 Antibiotic resistance surveillance No resistance surveillance data are available for the anti biotic resistance of the entire rectal flora microbiome. However, surveillance studies have reported antibiotic resistance for specific specimens in different geographi cal regions. The European Centre for Disease Control (ECDC) regularly publishes data from the European Antimicrobial Resistance Surveillance Network (EARSnet) database regarding the resistance of blood culture 82 | FEBRUARY 2014 | VOLUME 11
isolates taken from patients from most European coun tries. Fluoroquinolone resistance of E. coli bloodstream isolates ranges from 10–40% in Europe, with an average resistance of approximately 20%.34 In North America, ciprof loxacin resistance has reached 12% for E. coli strains that cause bloodstream infections.35 The GPIU of the European Section for Infection in Urology (ESIU) has been collecting prevalence and resistance data for the past 10 years in 1,395 infected urological patients.36 According to this database, approximately 45% of E. coli strains are resistant to fluoroquinolone, although this rate is subject to significant regional variation.36 In some regions, multi resistant bacteria, such as extended-spectrum β-lactamaseproducing enterobacteria, must also be considered,2 with carbapenemase-producing bacteria represent ing the latest threat in certain regions of the world. 37 Different types of carbapenemases, such as the KPC type from the USA, Israel, Greece, Puerto Rico, and Colombia, the VIM type from the southern Mediterranean region, and the NDM-1 type from the Indian subcontinent have reportedly been introduced into European health-care systems via cross‑border patient transport.37 Minimal inhibitory concentration (MIC) testing— defined as the antimicrobial concentration needed to inhibit visible growth—of feacal pathogens in patients undergoing prostate biopsy has shown high-level fluoro quinolone resistance, with MIC values of ≥32 mg/l for 65% of fluoroquinolone-resistant strains.24 This finding is in agreement with a recent study of the molecular changes in fluoroquinolone-resistant pathogens in patients under going transrectal prostate biopsy, which showed that diverse combinations of chromosomal mutations lead to high-level fluoroquinolone resistance.38 Pharmacokinetic and pharmacodynamic properties In general, the pharmacokinetic properties of antibiotics used for prophylaxis, including those used before prostate biopsy and transrectal prostate biopsy, have been insuffi ciently evaluated. A comparative study performed in 1991 showed significantly higher rates of bacteraemia (37% versus 7%) and symptomatic bacteraemia (22% versus 0%) when patients receiving transrectal prostate biopsy were prophylactically treated with gentamicin rather than ciprofloxacin.39 Although, at that time, resistance was not a known problem for either of these antibiotics, the choice of antibiotic clearly had a substantial impact on the prevention of bacteraemia, most probably resulting from the marked pharmacokinetic differences between the two substances. Aminoglycosides generally exhibit a very low volume of distribution (which is an indicator of tissue penetration) compared with fluoroquinolones (which have high volumes of distribution), leading to low concentrations of gentamicin (an aminoglycoside) and relatively high concentrations of ciprofloxacin (a fluoro quinolone) in the prostate.40,41 The results of this study suggest that antibiotic concentration in the prostate might be more important than blood concentration at the time of biopsy. The exact prostate concentration needed for drug efficacy is, however, not yet known and is likely to be dependent on the MIC of the pathogen.
www.nature.com/nrurol © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS The prostate exhibits anatomical barriers to the distri bution of antimicrobial molecules. The prostate gland is perfused by nonfenestrated capillaries and the pros tatic epithelium lacks active transport mechanisms for drug uptake.42 Pharmacokinetic investigations of anti biotics have been performed in prostate tissue, as well as in prostate secretions, although data from studies in prostate tissue should be interpreted with caution, owing to potential contamination with urine or blood.43 Of all the antibiotics, fluoroquinolones (Box 1) and macrolides have been the most thoroughly investigated and exhibit the highest ratios of prostate secretion or tissue concen tration to serum concentration, indicating good prostate penetration;41,43–45 data for other antibiotics are somewhat scarce. β-lactam antibiotics, such as piperacillin, cefpo doxime, meropenem, and doripenem, reach concentra tions in the prostate that are just 10–20% of those in the serum,44–47 whereas the prostatic tissue to serum ratio of amikacin is approximately 50%,44 and, in one study, con centrations of trimethoprim were higher in the prostate than in the serum.48 As such, it is difficult to establish the concentration needed for prophy laxis. The same principles used for treating bacterial prostatitis are likely to apply, although it is worth noting that these are also poorly defined. One study of levofloxacin treatment for chronic bacterial prostatitis demonstrated a significantly higher bacterial eradication rate of 79% when the MIC of the pathogen was
