Bacterial Urinary Tract Infection after Transrectal Placement of Fiducial Markers prior to Proton Radiotherapy for Prostate Cancer

Purpose: To determine the incidence of a bacterial urinary tract infection (UTI) necessitating hospitalization after transrectal placement of fiducial markers prior to proton radiotherapy (RT) for prostate cancer. Materials and Methods: Six hundred sixty six patients returning for follow up after proton RT consented to participate in this institutional review board (IRB) approved study. Patients were queried whether they required hospitalization within 1 month of transrectal placement of fiducial markers. Patients were treated with proton RT between August 2006 and December 2014. Median International Prostate Symptom Score (IPSS) was 7. Sixty four patients (9.6%) had diabetes, 9 patients (1.4%) had chronic obstructive pulmonary disease, 6 patients (0.9%) had prior bladder surgery, 7 patients (1.1%) had a transurethral prostatectomy within 3 months, and 549 patients (82.4%) 3 Conclusion: The incidence of a bacterial UTI requiring hospitalization after transrectal placement of fiducial markers prior to proton RT was 1.5% and was impacted by taking saw palmetto and year of treatment. Patients treated during 2013 and 2014 had a significantly higher risk of a bacterial UTI requiring hospitalization.


Introduction
Transrectal placement of fiducial markers into the prostate prior to image-guided radiotherapy (RT) is frequently employed to optimize beam alignment [1][2][3][4][5]. A potentially serious complication of this procedure is a bacterial urinary tract infection that can result in sepsis requiring hospitalization and, in a worst case scenario, death.
There are relatively few data relating to the risk of this complication after transrectal placement of fiducial markers so that most of the relevant data pertain to transrectal ultrasound guided prostate biopsy [1]. Herein we report our experience in men treated with proton beam RT for prostate cancer at our institution.

Materials and Methods
Between August 2006 and December 2014, 3556 men were treated with proton beam RT for prostate cancer at the University of Florida Health Proton Therapy Institute. With institutional review board approval, patients returning for follow up after completing treatment were asked whether they would be willing to participate in this study. If they consented, they were asked whether they required hospitalization within one month of transrectal placement of the fiducial markers prior to RT. Six hundred sixty six men (18.7%) were included in the study. Patient characteristics are depicted in Table 1.
Typically, four fiducial markers were placed into the prostate employing a transrectal approach prior to obtaining treatment planning computer tomography and magnetic resonance imaging [3]. Patients received a 3 day course of levofloxacin 500 mg daily beginning the day before the procedure. They were asked to use an enema three hours before the fiducial placement. No patient had a rectal culture prior to the procedure to direct selection of the optimal antibiotic for infection prophylaxis. The median IPSS was seven. The majority of patients had a prostate biopsy and/or a course of antibiotics within 6 months of

Results
Ten (1.5%) of 666 patients developed an apparent bacterial UTI requiring hospitalization after fiducial marker placement. The time interval between fiducial marker placement and hospitalization was 1 day (2 patients), 2 days (5 patients), 3 days (2 patients), and 7 days (1 patient). Urine cultures revealed a bacterial infection in 7 patients, negative in 2 patients, and no data in 1 patient. Blood cultures were positive in 4 patients, negative in 4 patients, and no data was available in 2 patients. The offending organism was Escherichia coli (E. coli) in 7 patients and Klebsiella pneumoniae in 1 patient. Resistance to the following antibiotics was observed: ampicillin, 6 patients; ciprofloxacin, 4 patients; levofloxacin, 3 patients; trimethoprim/ sulfamethoxazole, 2 patients; gentamicin, 1 patient; cefazolin, 1 patient; cefepime, 1 patient; ceftriaxone, 1 patient; and ceftazidime, 1 patient. The results of a univariate analysis of parameters that could impact the likelihood of developing a bacterial UTI is depicted in Table 2. Multivariate analysis of these parameters is shown in Table 3. The incidence of a bacterial UTI was 6 (1.0%) of 604 patients not taking saw palmetto versus 4 (6.5%) of 62 patients who were taking saw palmetto. The incidence of a bacterial UTI was 3 (0.7%) of 440 patients between 2006 and 2012 and increased to 7 (3.1%) of 226 patients during 2013 and 2014. Because some patients may develop a bacterial UTI after returning home after placement of the fiducial markers, we do not have reliable data pertaining to results of blood or urine cultures that may have been obtained or the identity of offending organism in all 10 patients.

Discussion
Although the transrectal placement of fiducial markers allows for more precise beam alignment, and thus more conformal fields when treating patients with prostate cancer with RT, there is a modest risk of a major complication including a bacterial UTI that may result in sepsis. The risk of bacterial urinary tract infection after transrectal fiducial placement as well as after transrectal prostate biopsy is depicted in Table 4 and 5 [6][7][8][9][10][11][12][13][14][15][16][17][18]. The incidence of a bacterial UTI has been rising in recent years likely due to the increasing prevalence of E. coli resistant to flouroquinolones (FQ) that may be detected in rectal cultures in 13% to 22% of patients [15,[19][20][21][22]. Indeed our data indicate that the risk of a bacterial UTI resulting in hospitalization increased in 2013-2014 compared with patients treated from 2006 to 2012. There are data indicating that patients who have had a prior prostate biopsy may be at an increased risk of developing this complication [1]. Thus, essentially all patients undergoing transrectal placement of fiducial markers are at an increased risk because all have had at least one transrectal ultrasound guided prostate biopsy to diagnose prostate cancer. Additionally, a recent course of antibiotics may increase the risk of urosepsis [1]. Because patients undergoing a transrectal prostate biopsy usually receive prophylactic antibiotics, a substantial proportion of patients undergoing placement of fiducials would have received antibiotics within 3 to 6 months of the procedure.
The reasons that saw palmetto might impact the likelihood of developing a bacterial UTI are unclear. One might postulate that patients with more urinary obstructive symptoms might have larger prostates and be more inclined to take this drug and that prostate volume might be related to this complication. However, we observed no difference in the risk of a bacterial UTI and IPSS.
There are several strategies that may be employed to reduce the risk of a severe bacterial UTI after placement of fiducial markers. These include empiric changes in the prophytactic antibiotic regimen, rectal cultures obtained prior to the procedure to direct the antibiotic prophylaxis, and deployment of the fiducuials through the perineum rather than rectal wall [1]. We have chosen to obtain a rectal culture at the time of initial consultation. If the culture reveals FQ resistant E. coli, the sensitivity of the bacteria to various antibiotics is determined and the prophylactic antibiotic regimen is modified accordingly. Patients who do not have FQ resistant E. coli receive a 3 day course of levofloxacin as previously described and intramuscular gentamicin 80 mg on the day of the procedure. We have also begun to employ the hydrogel SpaceOAR in eligible patients to reduce the RT dose to the anterior rectal wall. Eligible patients are those with low-and intermediate-risk prostate cancers and a prostate less than 100 cc. The SpaceOAR is deployed via the transperineal route and, during the procedure, the fiducial markers are placed into the prostate also using the transperineal route. These patients receive ciprofloxacin 500 mg twice daily for two days beginning the day before the procedure and undergo an enema 3 hours prior to deploying SpaceOAR and placement of fiducial markers.
Weaknesses of our study include that it is retrospective and the study population represents a subset of the overall patient population treated during that period. Patients included in the study were those who returned for a follow up evaluation so that those who elected not to return, or were unable to do so because of age or infirmity, would not have been included. These  patients may have a different likelihood of developing urosepsis, possible higher, than the study population. Patients who were hospitalized often had returned to their community and hospitalized there and results of blood and/or urine cultures were not always available.

Conclusion
The incidence of a bacterial UTI necessitating hospitalization after transrectal placement of fiducial markers is low but is likely increasing due to development of multidrug resistant E.coli. Strategies to decrease the risk of developing this complication include empirical changes in the prophylactic antibiotic regimen, rectal culture and directed changes in the prophylactic antibiotics, and changing from transrectal to transperineal placement of fiducial markers. Although the transperineal approach should reduce the risk of a bacterial UTI there are not convincing data that support this assumption. However, a caveat is that most of the relevant data are 10 to 15 years old before the apparent increased risk of harboring FQ resistant E. coli in the rectum.

ADDITIONAL INFORMATION AND DECLARATIONS
Conflicts of interest: The authors have no conflicts of interest to disclose.