Clinical Value of a Routine Urine Culture Prior to Transrectal Prostate Biopsy

Take Home Message We evaluated a clinical routine with urine culture prior to transrectal prostate biopsy in a comparative, population-based register study of almost 6000 procedures and found that it did not lead to fewer infectious complications.


Introduction
In Europe, >1 million prostate biopsies (PBs) are performed per year [1]. Infectious complications after a transrectal PB are an increasing problem. The proportion of patients having a febrile postbiopsy infection despite the use of prophylactic antibiotics varies across studies, from 0.7% to 7% [2][3][4][5][6]. The infection may cause severe sepsis and in rare cases even death [6]. The optimal prophylactic antibiotic regimen is not known. Until a few years ago, a single dose of an oral fluoroquinolone (FQ) was recommended and widely used [7][8][9], despite recommendations to avoid broad-spectrum and favour narrow-spectrum agents [10][11][12]. Owing to the increasing FQ resistance, recent awareness of potential serious side effects [13] and general overuse of broad-spectrum antibiotics, the European Association of Urology guidelines since 2021 instead recommend antibiotic prophylaxis based on a stool culture or rectal swab, augmented prophylaxis, or non-FQ antibiotics (fosfomycin, cephalosporin, or aminoglycoside) [14,29,30].
As asymptomatic bacteriuria is a risk factor for infectious complications after invasive procedures in the urinary tract, a urine culture and preprocedural, targeted antibiotics are recommended before many urological procedures [15]. A site-specific antimicrobial stewardship programme in Region Kronoberg, Sweden, included a routine culture prior to all PBs in the county from 2015 to enable targeted antimicrobial prophylaxis and reduce postbiopsy infections. However, the value of a such a routine is not known. According to a Swedish study, asymptomatic bacteriuria is a risk factor for a postbiopsy infection [16], but two previous small studies, one from France [17] and one from the USA [18], did not support this. Owing to these conflicting results, we designed a large, population-based study to evaluate the clinical benefit of a routine prebiopsy urine culture and targeted antimicrobial prophylaxis to reduce postbiopsy infections.

Study design
Observational, population-based, register study with a historical control period.

Setting and study population
In 2015, as part of a regional review of surgical antibiotic prophylaxis usage and general antimicrobial stewardship, a routine urine culture prior to a scheduled outpatient PB was implemented at all urology units the study who had a registered outpatient visit with the intervention code TKE00 or KEB00 (PB). Inpatient PB procedures were excluded because most patients selected for an inpatient PB have severe comorbidity and therefore received antibiotic treatment rather than prophylaxis. Patients whose urine was sampled from a catheter, urostomy, or nephropyelostomy were also excluded, as they often have resistant bacteria and should be considered for a pre-PB urine culture also in the absence of a clinical routine.

Data collection
Clinical data were obtained from the EMR system. Age at PB and any diagnosis of diabetes mellitus (E10/E11) or prostate cancer (C61) were registered as baseline characteristics. Medical records for all patients admitted to hospital 0-10 d after the PB were reviewed to ensure complete registration of complications and to define patients who received treatment for severe infections at an intensive care unit All urine and blood culture results from October 1, 2009 to January 31, 2020, including bacterial antibiotic susceptibility test results, were retrieved from the culture register at the Department of Clinical Microbiology.
The study was approved by the Swedish Ethical Review Authority (2020-02066).

Definitions
Any urine culture from 6 wk to 1 d before the date of the PB was defined as the routine urine culture (a urine culture sampled at the day of the PB could not guide treatment). Bacterial growth was quantified as 10 6 - changed during study period [19], and Sepsis-3 definition was used for registration. Serious infectious complications were defined as an infectious complication requiring in-hospital care.

Outcome measures
The primary outcome measure was the proportion of PB procedures followed by an infectious complication within 10 d.

Results
A total of 5932 visits with a PB code were identified from January 1, 2010 to December 31, 2019. Of these, 143 (2.4%) were excluded: 75 because the procedure was incorrectly coded, 40 because the PB was done under inpatient care, and 28 because the urine sample was not voided, leaving 5789 PB procedures in 4041 patients for an analysis (Fig. 1). The urine culture period (2015-2019) included 2971 PB procedures, of which 2684 (90%) were preceded by a urine culture. The control period (2010-2014) included 2818 PB procedures, of which 135 (4.8%) were preceded by a urine culture. The age at PB and the proportions of men with diabetes were similar in both study periods, but the proportion of men who had a biopsy as part of active surveillance for a previously diagnosed prostate cancer was higher in the urine culture period (22%) than in the control period (15%). The patient characteristics at the time of PBs are shown separately for the two time periods in Table 1. A post-PB infection was observed in a total of 150 patients (5.0%) undergoing PBs during the urine culture period and in 120 patients (4.3%) during the control period (p = 0.17). Inpatient care for a post-PB infection was more common during the urine culture period (3.5% vs 2.2%, p = 0.002). A positive blood culture was registered after 27 (0.9%) of the biopsy procedures during the urine culture period and after 15 (0.5%) procedures during the control period (p = 0.12). No patient met the sepsis criteria (SOFAscore 2) and none was treated at an ICU.
During the urine culture period, 160 (5.4%) of the pre-PB urine cultures were positive. All but one of these led to an antibiotic prescription: 113 (3.8% of all biopsy procedures) FQ and 46 (1.5%) non-FQ prescriptions. Of the 160 PBs preceded by a positive urine culture, ten (6.3%) were followed by an infectious complication. Five of the ten men received FQ and five non-FQ antibiotics. During the control period, 25 (0.9%) PBs were preceded by a positive urine culture, one of which was followed by an infectious complication. All 25 PBs with a positive urine culture resulted in an antibiotics prescription: 20 (0.7%) with FQ and five (0.2%) with non-FQ prescriptions ( Table 2).
The proportion of FQ-resistant bacteria in the prebiopsy urinary cultures was 8.8% (14/160) in the urine culture period (E. coli n = 10, enterococci n = 4, and S. aureus n = 2) and 8.0% (2/25) in the control period. None of these men with FQ-resistant bacteriuria had an infectious complication.

Discussion
We evaluated a clinical routine with pre-PB urine cultures in a comparative, population-based register study of almost 6000 PB procedures and found that it did not lead to fewer infectious complications. On the contrary, infections leading to inpatient care were even more common after implementing routine pre-PB urine cultures.
Our results agree well with the results from the previous two small single-centre studies [17,18], where positive prebiopsy urine cultures were left untreated without increased rates of infectious complications.
One possible reason for the higher post-PB infection rate in the urine culture period is that the proportion of patients on active surveillance was greater than that in the control period; previous research has shown that the risk of post-PB infections increases with the number of previous PBs [20]. The postbiopsy infection rate during the routine urine culture period in our study was similar to what was recently reported from a neighbouring county where they did not culture urine routinely before PB [21]. It is therefore unlikely that an increasing FQ resistance rate in the community would have caused an even higher post-PB infection rate in the absence of a routine pre-PB urine culture than was observed in our study. Indeed, the overall prevalence of FQ-resistant bacteria in positive urine cultures in Region Kronoberg was stable at around 10% in 2012-2019 [22].
The routine urine culture identified bacteriuria in 5.4% of the men before their planned PB, all of whom received targeted antibiotic treatment. One can assume that as many patients had undetected bacteriuria in the control period, and because post-PB infections were not more common in that period, it cannot be concluded that the targeted antibiotic treatment prevented any post-PB infection.
Of the 160 men with a positive urine culture prior to PB in the urine culture period, 14 had FQ resistance and were adequately treated without an infectious complication. Of the remaining 136 men with asymptomatic bacteriuria, ten (6.9%) suffered an infectious complication despite targeted treatment. It is thus possible that the routine with a pre-PB urine culture increased the risk of infections by leading to more complete FQ treatment courses that increased FQ resistance. The routine may also have harmed some men by delaying their cancer diagnosis.
Our study does not offer any evidence that isolated asymptomatic bacteriuria is a risk factor for post-PB infections. The results rather support the hypothesis that translocation of bacteria from faecal flora to the prostate and periprostatic tissues is the major cause. As the microbial reservoir of FQ-resistant bacteria is in the colon, future research in this area should focus on the clinical value of pre-PB faecal culture [23], rectal disinfection with povidone-iodine [24], and the transperineal biopsy route [25,26].   The main strengths of our study are the populationbased design, large study sample, high adherence to the pre-PB urine culture routine, and complete coverage of the EMRs for all parts of the health care services in the county. Limitations include the ''before-and-after'' comparison that, although reducing the risk of selection bias, made the results susceptible to confounding factors that caused an increased risk of post-PB infections for the patients who had a PB in the urine culture period. Another limitation is that we did not use a strict guideline for targeted antibiotic treatment. Additionally, we did not have any data on other risk factors such as FQ resistance in the faecal flora. Other limitations are the retrospective design and the short follow-up period of 10 d after a PB (some infectious complications may have occurred later). Moreover, the results of our study may be applicable only to populations with low FQ resistance. The prevalence of 10% in our region during 2012-2019 [22] is lower than that reported from other countries [27,28]. Routine urine cultures before a PB may therefore be beneficial in populations with higher FQ resistance, despite the negative results from our study.

Conclusions
In conclusion, the routine with a urine culture prior to a PB did not reduce the postbiopsy infection rate. Future research should focus on other measures to reduce infectious complications after a PB.