Early infectious complications with transponder placement for external beam radiation therapy for prostate cancer

Authors


Ryan K. Berglund, 9500 Euclid Avenue, Q10-1 Cleveland, OH 44195, USA. e-mail: berglur@ccf.org

Abstract

Study Type – Therapy (case series)

Level of Evidence 4

What's known on the subject? and What does the study add?

Little has been published related to transponders per se, but a number of studies relating to prostate biopsy-related infections and the increased incidence of quinolone-resistant Escherichia coli have been published.

The study alerts the practising urologist to the risk of quinolone-resistant E. coli in the setting of transrectally placed transponders. Furthermore, it proposes an antibiotic regimen that should reduce this risk.

OBJECTIVE

  • • To report our series of early infectious complications after placement of Calypso® transponders (Calypso Medical, Seattle, WA, USA) into the prostate.

PATIENTS AND METHODS

  • • Between February 2008 and October 2010, 50 consecutive patients underwent placement of Calypso® transponders into the prostate.
  • • Patients were administered ciprofloxacin 500 mg every 12 h, starting the night before the procedure and for 2 days after the procedure.
  • • Data were collected via chart review, and complications were classified according to the Clavien classification system.

RESULTS

  • • Of the 50 patients undergoing the procedure, five (10%) developed infectious complications, and three (6%) developed a grade II complication with a UTI requiring antibiotic therapy. One patient (2%) developed a grade IIIb complication with an epidural abscess and osteomyelitis of the lumbar vertebrae requiring open debridement and a lumbar fusion. One patient (2%) developed a prostatic abscess with methicillin-resistant Staphylococcus aureus and subsequently died of an unrelated lower GI bleed.
  • • In 4/50 patients (8%), a culture confirmed the responsible bacteria, of which three cases were quinolone-resistant Escherichia coli.

CONCLUSION

  • • As with prostate biopsy, the emergence of quinolone-resistant E. coli remains a challenging infectious complication with transrectal prostate procedures. We propose an alternative strategy of double antibiotic coverage with one dose of oral ciprofloxacin 500 mg and gentamicin 80 mg i.m. before this procedure.
Abbreviation
MRSA

methicillin-resistant Staphylococcus aureus.

INTRODUCTION

The Calypso® 4D Localization System (Calypso Medical, Seattle, WA, USA) is a real-time localization and monitoring system that uses three electromagnetic transponders (beacons) placed transrectally into the prostate to allow precise intrafractional localization of the prostate during externally delivered radiation therapy for prostate cancer. This has allowed superior targeting accuracy compared with skin marking, CT, ultrasonography, and intraprostatic radiopaque fiducials, thereby limiting radiation dose delivered to neighbouring structures such as the rectum and bladder [10,23]. The electromagnetic transponders are placed using hollow 14-gauge needles with a stylet used to advance the transponder under transrectal guidance before starting radiotherapy. This is performed in an office setting under periprostatic anaesthetic block, with periprocedural antibiotic coverage to prevent infection.

For the last 20 years, TRUS-guided biopsy has been the technique of choice for diagnosing prostate cancer [2], and coverage with a quinolone antibiotic, such as ciprofloxacin, has been the standard treatment during this time [9,12]. Widespread use and overuse of quinolone antibiotics have correspondingly led to the emergence of quinolone-resistant strains of Escherichia coli, and these strains now account for ≈10% of all bacteria isolated at the time of diagnosis of UTI and the majority of post-biopsy infection isolates [8,11,21]. These resistant UTIs are also more likely to occur in a patient who has previously been exposed to quinolone antibiotics [16].

Transrectal placement of a permanent indwelling prostatic implant through a large-bore needle probably presents an equally important infectious risk. In the present study, we report our experience with infectious complications after TRUS-guided placement of Calypso® radiotransmitters into the prostate.

MATERIALS AND METHODS

From February 2008 to October 2010, 50 consecutive patients underwent TRUS-guided placement of Calypso® transponders (Calypso Medical) into the prostate at a single centre. Data were reviewed retrospectively via chart review of electronic medical records after Institutional Review Board approval (#11-160). Data collected included age, most recent PSA level before biopsy, clinical stage, Gleason score and prostate volume. Medical comorbidities and anticoagulation status and type were also recorded.

All patients except one underwent antibiotic preparation with ciprofloxacin 500 mg orally twice daily for a total of six doses, starting the evening before the procedure. One patient underwent a single oral dose of ciprofloxacin 500 mg given 1 h before the procedure, and he subsequently developed no complications. No bowel preparation or enema was given to any of the patients. Patients were asked to refrain from taking aspirin or coumadin anticoagulation for 5 days before the biopsy, and were asked to resume anticoagulation after no further bleeding was evident in urine, ejaculate or stool. On the day of the procedure, a dipstick urine analysis was obtained before the procedure to ensure that no signs of UTI were present, including the presence of leukocyte esterase and nitrites in the urine. If there was a suspicious dipstick result, a microscopic examination of the urine was performed and a urine culture was sent for laboratory analysis, with the procedure delayed until resolution of these findings.

Informed consent was obtained from each patient. The procedures were all performed using a side-fire TRUS probe (BK Medical, Peabody, MA, USA). Periprostatic nerve block was administered bilaterally with a total of 10 mL of 1% lidocaine, and prostate measurements were estimated using the ellipsoid formula. Standard placement of the transponder was performed in the left and right bases and apex of the prostate under real-time TRUS. After the biopsy procedure, patients were observed in the supine postion for ≈15 min to ensure no active haemorrhage and/or rigors were present.

Patients were then seen 1–2 weeks after the procedure in the radiation oncology department where history, physical examination, and radiation treatment planning was performed. Follow-up continued throughout the several-week course of radiation treatment, with standard follow-up including history, physical examination and PSA level assessment subsequently performed at 6-month intervals. Urine analysis and urine culture were not routinely performed during treatment in the asymptomatic patient. Patients were considered to have an infectious complication if they had symptoms of urosepsis (fevers, rigors and flank pain) that required medical intervention (usually in the form of antibiotic treatment) or a positive culture from urine, blood or drained body fluid. Complications were graded according to the Clavien classification system for surgical complications [6]. Culture results were also reviewed for antibiotic sensitivities.

RESULTS

Patient characteristics are listed in Table 1. The median (range) patient age was 67 (54–83) years. The median (range) patient PSA level was 6.0 (1.2–30) ng/mL and the median (range) prostate volume (calculated from ultrasonography) was 29 (16–100) cc. Of the 50 patients, 27 (54%) had clinical stage T1c disease, 14 (28%) had clinical stage T2a disease, two (4%) had clinical stage T2b, four (8%) had stage T2c, and two (4%) had stage T3 disease. One patient (2%) had locally advanced stage T4 disease. Gleason score was <7 in 12 patients (24%), = 7 in 26 patients (52%), and >7 in 12 patients (24%). All patients had undergone prostate biopsy with ciprofloxacin antibiotic prophylaxis <6 months before transponder placement.

Table 1.  Patient characteristics
Patient characteristic, N= 50Value
Median (range) age, years67 (54–83)
Median (range) PSA, ng/mL6.0 (1.2–30)
Median (range) prostate volume, mL29 (16–100)
Clinical stage, n (%) 
 T1c27 (54)
 T2a14 (28)
 T2b2 (4)
 T2c4 (8)
 T3c2 (4)
 T41 (2)
Gleason score, n (%) 
 <712 (24)
 7 (3+4)14 (28)
 7 (4+3)12 (24)
 8 (4+4)6 (12)
 9 (4+5)5 (10)
 9 (5+4)1 (2)
Comorbiditites, n (%) 
 Hypertension20 (40)
 Hyperlipidemia18 (36)
 Diabetes mellitus6 (12)
 Anaemia2 (4)
 Thrombocytopenia1 (2)
 End-stage renal disease1 (2)
Anticoagulation, n (%) 
 None26 (52)
 Aspirin only18 (36)
 Coumadin only4 (8)
 Aspirin and coumadin2 (4)

The three most common patient medical comorbidities were hypertension (20/50, 40%), hyperlipidemia (18/50, 36%), and diabetes mellitus (6/50, 12%). Anticoagulants were used in 24/50 patients (48%) with aspirin and/or warfarin before biopsy, none of which were used for a minimum of 5 days before biopsy.

Of the patients who underwent transponder placement, 6/50 (12%) had at least one complication. One patient (2%) had a Clavien I bleeding complication which he reported to his primary care physician 3 days after transponder placement. This patient reported gross haematuria and gross blood per rectum, which resolved without any further intervention the day after presenting to and being evaluated in the radiation oncology clinic. The patient had not been on anticoagulation therapy before biopsy.

Infectious complications occurred in 5/50 patients (10%). One patient (2%) presented with a Clavien II complication with 3 days of rigors and fevers to 38.5 °C, starting 7 days after transponder placement. He was admitted to the emergency department where urine analysis and microscopic analysis showed 11–25 white blood cells per high power field. Urine and blood cultures were drawn and were ultimately interpreted as negative. He was hospitalized and placed on piperacillin/tazobactam 3.375g i.v. every 6 h for 2 days, and then switched to ciprofloxacin 500 mg orally twice daily, which he continued for a total of 28 days for a presumptive diagnosis of acute bacterial prostatitis. He recovered well without other incident and received radiation therapy.

In 4/50 patients (8%) urine, body fluid and/or blood culture showed the responsible pathogen; the results of these cultures are shown in Table 2. In three of these patients (6%) a quinolone-resistant E. coli was responsible for their infection. Two patients (4%) had a Clavien II complication with a UTI caused by quinolone-resistant E. coli. Patient #1 (Table 2) developed rigors, chills, and fevers to 37.8 °C, 3 days after transponder placement. He had discontinued his ciprofloxacin after four of the six total doses believing the ciprofloxacin to be the cause. A urine culture showed E. coli resistant to ciprofloxacin, levofloxacin, and trimethopim/sulfamethoxazole. He was switched to cephalexin 500 mg orally four times daily for a week, and recovered without further incident. He began radiation therapy 24 days after transponder placement, and otherwise tolerated the procedure well.

Table 2.  Bacterial sensitivities
AntibioticPatient #1Patient #2Patient #3Patient#4
E. coli E. coli E. coli MRSA
  1. S, sensitive; I, intermediate; R, resistant; NA, not tested.

AmpicillinSRSNA
Ampicillin/SulbactamSSSNA
CefazolinSSSNA
CeftazidimeSSSNA
CeftriaxoneSSSNA
CiprofloxacinRRRNA
GentamicinSSSS
ImipenemSSSNA
Trimethoprim/SulfamethoxazoleRSSS
NitrofurantoinSSNANA
Piperacillin/TazobactamSSNANA
LevofloxacinRRNANA
ErythromycinNANANAR
Penicillin GNANANAR
TetracyclineNANANAS
VancomycinNANANAS
OxacillinNANANAR
RifampinNANANAS
TobramycinSNANANA

Patient #2 (Table 2) presented to the emergency department 2 days after transponder placement. He had a history of venous thromboembolism, for which he had been on warfarin, but that had been held for 5 days before transponder placement. He presented with urinary retention and rectal and pelvic pain, rigors, and fevers to 38.5 °C. Urine analysis and microscopic examination revealed >25 red blood cells and white blood cells per high power field, leukocyte esterase and nitrite positivity. Foley catheter placement showed ≈1000 mL of cloudy urine in the bladder. Urine and blood culture at that time confirmed quinolone-resistant E. coli with resistances to ampicillin, ciprofloxacin, and levofloxacin. He was started on empiric piperacillin/tazobactam 3.375 grams i.v. every 6 h, and trimethoprim/sulfamethoxazole 160/800 orally twice daily was added on hospital day 2. He was hospitalized for 7 days and treated with i.v. piperacillin/tazobactam for 2 weeks, and oral trimethoprim/sulfamethoxazole for 6 weeks. He did require one readmission the day after discharge for dehydration, hyponatremia and acute renal failure, which resolved with hydration after 1 day in hospital. His foley catheter was removed 28 days after transponder placement, and he began radiation therapy 19 days after that, which was uneventful.

Patient #3 (Table 2) had a Clavien IIIb complication. This patient was a 58-year-old man with a history of diabetes mellitus diagnosed with a prostate cancer, clinical stage T1c, Gleason 7 (3+4) 3/12 cores, PSA level 11.45 ng/mL. Four days after transponder placement, the patient noted fatigue, and 13 days later he was admitted to the emergency department with severe back pain. MRI of the spine 28 days after the procedure showed L2–3 osteomyelitis with a neighbouring fluid collection, which was drained 2 days later and showed quinolone-resistant E. coli (Table 2). Initial treatment with imipenem/cilastatin 500 mg i.v. every 6 h was initially changed to ceftriaxone 1 g i.v. every 24 h. This was changed to aztreonam 2 g i.v. every 8 h after the patient developed a rash to ceftriaxone. The infection progressed to an epidural abscess, and on day 67 after transponder placement, he was taken to the operating room for an L 2-3 discectomy, debridement, L 2-3 laminectomy and spinal fusion. He was subsequently treated with i.v. aztreonam 2 g every 8 h and i.v. vancomycin 1 g every 12 h until day 105 after the procedure, with reported resolution of symptoms. He is currently awaiting definitive treatment for prostate cancer.

Patient #4 (Table 2) had a Clavien IIIb complication. This 60-year-old man had a history of end-stage renal disease on hemodialysis, congestive heart failure, and hypertension, and he was diagnosed with a clinical stage T2c, PSA 27.5, Gleason 7 (4+3) in 12/12 cores prostate cancer with no evidence of metastases on bone scan or CT. He was treated with androgen deprivation therapy for 6 months before transponder placement and radiotherapy and received 5800 cGy of 7800 cGy planned external beam radiation therapy. On day 61 after transponder placement, the patient presented with rectal pain and purulent urinary drainage, along with an unexplained shift in the relative locations of the three transponders. Imaging with non-contrast enhanced pelvic CT showed a prostatic abscess, and the patient underwent suprapubic tube placement with endoscopic TURP to unroof the abscess. The culture obtained from that procedure showed methicillin-resistant Staphylococcus aureus ([MRSA]Table 2), and he was placed on vancomycin 1 g i.v. with dialysis. The patient was later found to have been non-compliant with medications historically, and it remains unknown whether he had actually taken the prescribed oral antibiotics at the time of the transponder placement. After convalescing in a nursing home, the patient was readmitted 91 days after transponder placement with fatigue and anaemia. He was found to have a lower gastrointestinal bleed from his cecum, outside of the irradiated field, with an anatomically separate, non-bleeding anterior rectal ulcer noted on colonoscopy, which was probably caused by acute radiation changes. At 95 days after transponder placement, he was moved to the intensive care unit to undergo multiple transfusions for haemodynamic instability, and he underwent bowel resection of the proximal colon. He was returned to the intensive care unit, where he died 102 days after transponder placement.

DISCUSSION

The widespread overuse of quinolone antibiotics has led to higher rates of resistance in hospitals and worldwide [7,13,14]. Of particular concern is the emergence of quinolone-resistant E. coli, which now accounts for ≈10% of UTIs and the majority of infections after prostate biopsy [8,11,21]. Quinolones have long been the mainstay of antibiotic prophylaxis after prostate biopsy because of their broad antibiotic spectrum and penetration of the genitourinary tract [4], but infection after prostate biopsy is increasingly common (1.7–6.3% of all biopsies), and they are increasingly resistant to commonly used urinary antibiotics such as ciprofloxacin, trimethoprim/sulfamethoxazole, and ampicillin [3,15,17,21].

In the present series, 12% of all patients had a complication, with all but one being an infectious complication, and 3/4 of the infectious complications were found to be caused by quinolone-resistant E. coli. There are several potential causes of this increased rate of infectious complications. Firstly, the use of a large-bore hollow needle could create a wider tract through which rectal pathogens could be introduced into the prostate. The 14-gauge Calypso® transponder introducer needle has a 66% larger diameter than a corresponding 18-gauge standard biopsy needle. In our previously published prostate biopsy series, in which we use an 18-gauge needle to perform standard 12-core biopsies, only 2.7% of patients had an infectious complication [25], Yet, when the bore was increased to 14-gauge for our transponder implants, the rate increased to 10%. Secondly, the injection of an indwelling foreign body implant into the prostate could provide a harbouring surface on which bacteria could grow and thrive, possibly explaining the delayed development of severe infections seen in patient #3 and patient #4. Unfortunately, during the time period of this study, only two cases of gold fiducial implant placement into the prostate were performed at our institution. There were no infections in either procedure, but this limits our ability to compare the effect of placing an implant using a smaller 18-gauge needle. All other equipment and antibiotic preparation were the same when placing either gold fiducials or transponders, so this would have been a reasonable comparison. It is possible that a smaller gauge needle reduces infection rates even when placing a permanent implant. Thirdly, the type of antibiotic and bowel preparation used may have allowed resistant bacterial flora to be introduced into the prostate with transponder placement.

Regarding bowel preparation, we did not use a rectal enema, although many urologists do administer an enema before prostate biopsy [5]. In a previous series of 1438 prostate biopsies looking at the use of an enema, our group has shown that enema bowel preparations did not reduce post-prostate biopsy infection rates [25]. As we point out in that study, the rectum is not a location of storage for stool. In fact, use of an enema often only contributes to liquid stool being present in the rectum during the prostate biopsy, which leads to patient and clinician dissatisfaction. Furthermore, the literature does not support the use of enemas to reduce infection rates during prostate biopsy [4], Although the role of enemas is not directly reported in the setting of fiducial placement, the similarities preclude any logic of using enemas to reduce infections in this setting.

Another problem with the preparation for these patients was the type and duration of antibiotics used. All patients received ciprofloxacin, and none of the patients in this series were quinolone naïve, having received ciprofloxacin within the preceding 6 months for their initial prostate biopsy. Wagenlehner et al.[22] showed that even receiving a single dose of ciprofloxacin quadruples the rate of E. coli resistance within a week to 12% from a background of 3%, and these strains also have high rates of resistance to trimethoprim/sulfamethoxazole and ampicillin [21]. Secondly, giving ciprofloxacin the night before the procedure increases the risk of colonizing the rectum with quinolone-resistant E. coli. Shigehara et al.[19] recommend giving the first dose of antibiotic 1 h before the biopsy to reduce this risk, as is currently done with preoperative antibiotics in the operating room. They further recommend the use of a cephalosporin or carbapenem in the event of infection after biopsy given the high rates of resistance to other urinary antibiotics. Last of all, use of more than 1 day of antibiotics for infectious prophylaxis is not supported in the literature. Several studies support the use of only 24 h of antibiotics for bacterial prophylaxis [18,20], and a randomized trial supports the use of only a single dose of ciprofloxacin in this setting [1]. The AUA Best Practice Statement recommends no more than 24 h antibiotic coverage [24]. We recently published a prostate biopsy series showing similar rates of infection between patients receiving an enema and 3 days of ciprofloxacin vs no enema and a single dose of ciprofloxacin given 1 h before prostate biopsy. Importantly, the single-dose group had a lower rate of resistant pathogens when an infection occurred, particularly with quinolone-resistant E. coli. Furthermore, longer courses of antibiotics increase the risk of opportunistic infection such as Clostridium difficile. This led us to conclude that no enema and a single dose of ciprofloxacin was a preferable prebiopsy preparation, and we have adopted those guidelines as an institution [12]. Furthermore, based on advice from our infectious disease department, we now recommend double coverage with single-dose ciprofloxacin 500 mg orally and gentamicin 80 mg orally in any patients exposed to quinolones within the past 2 years, because of the high rate of resistance and high rate of sensitivity to aminoglycosides in those patients with floroquinolone-resistant infections. An alternative antibiotic to gentamicin is single-dose cephalexin 2000 mg orally in the patient unable to take aminoglycosides, as there is high sensitivity to cephalosporins as well. Finally, parenteral cephalosporins also have similar activity in this population of patients undergoing transponder placement, but on the advice of our infectious disease colleagues we reserve their use for patients with sepsis, or those that cannot take oral cephalasporins.

Given these findings, we currently recommend that all patients undergoing transponder placement be treated with only a single dose of ciprofloxacin 500 mg orally and gentamicin 80 mg i.m., given 1 h before transponder placement. We continue to use ciprofloxacin as one of the two agents, based on the fact that 97% of our biopsy patients from our aforementioned series avoided infectious complications with that single drug, but add the aminoglycoside based on the higher rate of floroquinolone resistance in these patients after biopsy. An acceptable alternative to gentamicin would be the use of a single dose of cephalexin 2000 mg orally in the place of gentamicin. We also continue to advocate not using a rectal enema. We do recommend that all patients presenting with systemic symptoms of infection or pain be treated empirically with a cephalosporin, fourth generation penicillin and beta lactamase inhibitor, or carbapenem. In addition, given the 4% rate of serious infections requiring drainage of a distant site, we also recommend CT or MRI in this patient population.

The weaknesses of the present study include its retrospective nature and lack of comparison with an alternatively treated population. Furthermore, conclusions of prophylaxis and treatment for prostate biopsy may not be transferable to transponder placement. Only a randomized and blinded prospective study can draw definitive conclusions as to which regimen would be appropriate; however, such a study would require large numbers of patients and we believe it unlikely that this would be feasible within the current financial and institutional constraints. The present study is small as a result of its being an early report of concern regarding infection risk in a single centre experience. Clearly, in a small group of patients such as this, the high infection rate may merely be a statistical artifact and not representative of what might be found in a larger prospective multicentre series. We may be drawing conclusions based on just a few random events. The present study does address an observed risk for infection after transrectal transponder placement, particularly from quinolone-resistant E. coli, and uses evidence from prostate biopsy series to support a different preparation regimen.

In conclusion, as with prostate biopsy, the emergence of quinolone-resistant E. coli remains a challenging infectious complication with transrectal prostate procedures. We propose an alternative strategy of double antibiotic coverage with one dose of oral ciprofloxacin 500 mg and gentamicin 80 mg before this procedure. Single-dose cephalexin 2000 mg orally is an acceptable alternative to gentamicin in the presence of drug allergy.

CONFLICT OF INTEREST

None declared.

Ancillary