Randomized trial to assess the efficacy of intraoperative steroid use in decreasing acute urinary retention after transperineal radioactive iodine-125 implantation for prostate cancer
Acute urinary retention is a potential complication of brachytherapy, with the literature estimating that 5% to 22% of patients require catheterization within 48 hours after implantation. In theory, postimplantation edema could be reduced by using intraoperative steroids. A prospective trial was conducted randomizing patients to a single intraoperative dose of dexamethasone versus no steroid use.
In all, 196 evaluable patients who received iodine‒125 (I125) interstitial brachytherapy alone as definitive treatment for low-to-intermediate risk prostate cancer were randomized to receive either dexamethasone at a dose of 6 mg administered intravenously intraoperatively (Arm A) or no steroids (Arm B). All patients completed the International Prostate Symptom Score before treatment. Patients were contacted by telephone 72 to 96 hours after treatment and the need for catheterization was reported.
Between 2003 and 2005, 99 patients received steroids on treatment Arm A and 97 patients were treated according to control Arm B. Treatment arms were balanced with respect to pretreatment characteristics. A total of 3 patients required catheterization (2 in Arm A and 1 in Arm B). The overall rate of catheterization was 1.5%, with no statistically significant difference noted between treatment arms. The 3 patients requiring catheterization had no statistical differences from other patients with respect to pretreatment characteristics, number of seeds/needles used, or postimplantation computed tomography volume of the prostate.
There was no statistically significant difference noted between treatment arms in the current study, leading the authors to conclude that intraoperative dexamethasone did not decrease the rate of catheterization required after brachytherapy. The overall rate of postimplantation catheterization in the current study was 1.5%, which is lower than reported elsewhere in the literature and in a retrospective review from the study institution. Cancer 2008. © 2008 American Cancer Society.
A large number of patients choose brachytherapy as their definitive treatment for early‒stage prostate cancer. Several series in the literature have shown excellent long-term biochemical control rates and progression-free survival.1, 2 The costs, convenience to patients, and morbidities also compare favorably with other local treatment modalities.
Acute urinary morbidity requiring catheterization is a known complication of prostate brachytherapy. The rate of catheterization within 48 hours of brachytherapy has been quoted in multiple series in the literature to be between 5% and 22%. Factors contributing to increased acute urinary obstructive symptoms have also been cited in the literature and include baseline International Prostate Symptom Score (IPSS) score, baseline prostate volume, a history of diabetes, surgeon learning curve, and amount of postimplantation edema as assessed by the ratio of postimplantation computed tomography (CT) to planning ultrasound (US) target volume.3–12
Theoretically, the use of dexamethasone perioperatively could decrease postimplantation edema. A retrospective review of 200 patients at our institution demonstrated a 10% incidence of postoperative need for catheterization in patients between 1992 and 1998 without the use of steroids. After a change in institutional policy, all patients received intraoperative steroids, and a 2% rate of catheterization was noted between 1998 and 2002.13
We conducted a prospective, single-institution, randomized trial in 204 patients to determine the effect of intraoperative dexamethasone on the rate of postimplantation catheterization. The effect of steroids on postimplantation edema and prostate volume was also evaluated through retrospective review of preimplantation US volumes and postimplantation CT scans.
MATERIALS AND METHODS
From 2003 through 2005, 204 patients were treated on protocol with iodine-125 (125I) interstitial, low-dose-rate prostate implantation alone as definitive treatment for low-to-intermediate risk prostate cancer at the University of Cincinnati. All patients had histologically confirmed adenocarcinoma of the prostate, clinical stage T1c-T2a disease, N0, a Gleason score of 6 to 7, and a prostate-specific antigen (PSA) level <20 ng/dL (pretreatment characteristics are listed in Table 1). Patients were excluded for any history of pelvic radiotherapy, radiographic evidence of pelvic lymph node involvement (N1), clinical/radiographic evidence of metastatic disease (M1), or current use of adrenocortical steroid therapy. No patient received preimplantation hormonal therapy.
Table 1. Pretreatment and Treatment Characteristics
|Age, y||68 (52-83)||67 (49-79)||.25|
|T1cN0|| || || |
|PSA||5.7 ± 3.0||6.0 ± 2.7||.22|
|6|| || || |
|US volume, cm3||42 ± 12||44 ± 15||.19|
|IPSS score||9 ± 5||8 ± 4||.34|
|No. of needles||17 ± 4||16 ± 5||.21|
|No. of seeds||61 ± 22||62 ± 20||.22|
After providing informed consent, patients completed a baseline IPSS questionnaire. Pretreatment characteristics were recorded for analysis, including patient age, Karnofsky performance status, clinical stage, Gleason score, PSA level, number of seeds used, number of needles used in implantation, preimplantation prostate volume by US, total activity of implant, and the use of pretreatment α‒blockers. Patients were excluded from definitive brachytherapy for any IPSS >18. Patients were then stratified by IPSS score (1-9 or 10-18) and the use of pretreatment α-blockers, and were then randomized preoperatively to receive dexamethasone at dose of 6 mg administered intravenously in a single intraoperative dose or no steroids.
Implants were performed by 2 radiation oncologists. Preoperative planning target volumes were determined by US according to American Brachytherapy Society recommendations and represented the volume of the prostate only (no margin). A minimal peripheral dose of 14.4 gray (Gy) was prescribed to the periphery of the prostate per TG-43 protocol. Seed activity ranged from 0.357 to 0.68 millicuries (mCi). A modified peripheral needle loading pattern was used. A Foley catheter was placed intraoperatively and removed when distal extremity neurologic function returned to baseline after spinal or general anesthesia (median time of 60 minutes after the completion of procedure). Fluoroscopy was used to confirm needle placement. All patients underwent postimplantation CT scan with 3-mm slice thickness on postoperative Day 0 or 1 for postimplantation dosimetry. Postimplantation prostate volumes were contoured on axial CT slices by 1 of 2 investigators.
All patients received tamsulosin at a dose of 0.4 mg daily on the operative and first postoperative days. Patients were contacted by telephone 48 to 72 hours after implantation and the need for catheterization was reported. All medical records were obtained for patients requiring catheterization outside the treating radiation oncologist's office.
On the basis of our retrospective data, an accrual goal of 200 patients was set to have 85% power in detecting a 10% decrease in the rate of catheterization with the addition of intraoperative steroids. Pooled Student t test and logistic regression were performed for statistical comparison of the group receiving steroids and the control group with regard to pretreatment characteristics. These tests were also used for comparison of patients requiring catheterization to those not catheterized.
In all, 196 of 204 patients were evaluable. Eight patients were excluded because of refusal of brachytherapy after study consent (2 patients), lack of follow-up (1 patient), positive toxicology screen and brachytherapy not performed (1 patient), patient noncompliance with tamsulosin (2 patients), or incorrect administration of steroid dose by anesthesiologist/investigator (2 patients). Ninety-nine patients received intraoperative steroids and 97 patients were treated according to the control arm. Pretreatment characteristics were not found to be statistically different between the 2 groups (Table 1). There was a trend toward a higher Gleason score in the control group (P = .06).
Three patients in total required catheterization for an overall rate of 1.5%. Two patients in the steroid group and 1 patient in the control arm were catheterized. The mean time to catheter placement was 36 hours and the mean duration of catheterization was 24 days. No patient required surgical intervention to relieve urinary obstruction. Each of the 3 patients was placed on α‒blockers until the obstruction resolved, and no patient required long-term catheterization (>45 days).
The pretreatment characteristics of patients requiring catheterization were quite variable and are shown in Table 2. Comparison of the mean pretreatment characteristic values between the patients who were catheterized and those not requiring catheterization yielded no statistically significant difference, as demonstrated in Table 3. There was a trend toward a larger prostate volume in patients requiring catheterization because of a single high value, but we have no statistical reason to expect other patients with similarly large volumes would require catheterization.
Table 2. Characteristics of Patients Requiring Catheterization
Table 3. Comparison of Patient and Treatment Characteristic Means Between Catheterized and Noncatheterized Patients
|Preimplant US volume, cm3||49.8||42.8||.49|
|No. of needles||18.3||16.4||.44|
|No. of seeds||62.3||61.4||.47|
When preimplantation dosimetry and postimplantation CT volumes were analyzed, there was no significant difference noted between patients receiving steroids and the control group, nor between catheterized and noncatheterized patients. The average D90 in the group receiving steroids was 142 Gy and was 140 Gy in the nonsteroid group, with no statistical difference between the groups. The ratio of postimplantation CT volume to preimplantation US volume was analyzed in each patient and the average volume increase on postoperative CT scan was 20%. Again, no significant difference was noted in this ratio between the steroid group and the control arm, concluding that intraoperative steroid use did not decrease short-term postimplantation edema. In addition, there was no significant difference noted with regard to the CT:US ratio between catheterized and noncatheterized patients (Table 4).
Table 4. Comparison of CT:US Ratio Between Patient Groups
|Mean CT:US ratio||1.19||1.23||.22|
| ||Catheterized||Noncatheterized|| |
|Mean CT:US ratio||1.18||1.17||.9|
The overall catheterization rate of 1.5% noted in the current study was lower than expected based on retrospective data from the literature and from our institution. Stock et al5 reported a review of 251 patients in 1998 with a 6% rate of catheterization. The Brigham and Women's group reported a 12% incidence of postimplantation catheterization and Bucci et al reported a 15% catheterization rate in 282 patients in Vancouver, British Columbia.4, 15 Ohio State University reported 15% of patients requiring catheterization in 2000, with a higher IPSS score found to correlate with increased rates of acute urinary obstruction.8 Sacco et al9 reported on the Massachusetts General Hospital retrospective experience of 400 patients, demonstrating an 8% catheterization rate in patients receiving perioperative steroids and 18% in patients treated without steroids. On multivariate analysis, the use of steroids and prostate volume at the time of diagnosis emerged as significant predictors of acute urinary obstruction at their institution. At the University of Cincinnati from 1992 until present, our catheterization rate within 48 hours has decreased from 10% to 1.5%. There are several hypotheses that can be generated from this low catheterization rate.
Our decrease in catheterization rates over time may represent an institutional learning curve, which has also been reported in other institutions. Keyes et al6 noted a decrease in their catheterization rate from 17% to 6.3% between 1998 and 2002. Speight et al3 cited an improvement in postimplantation dosimetry based on changes in needle placement over time at the University of California at San Francisco. All implants were performed by 2 radiation oncologists at our institution between 1992 and the present. This protocol was offered to all patients who were eligible for definitive prostate brachytherapy between 2003 and 2005 within our institution, and very few patients refused. The patients treated on this protocol represent the vast majority of patients who were treated with brachytherapy alone for prostate cancer during that timeframe at our institution. Although it may be helpful to compare these patients with those treated off protocol during the same period, this would not be possible because of low patient numbers for comparison.
The timeframe of acute urinary obstruction and the need for catheterization after 125I brachytherapy points to a traumatic rather than radiation cause for edema, because the dose delivered by 48 hours after implantation is not significant. The relation between a high number of periurethral needle manipulations in prostate brachytherapy implantation has been positively correlated in the literature with acute urinary toxicity by Eapen et al.15 In addition, in the Vancouver experience, the number of needles used per implant emerged as being predictive of acute urinary morbidity on univariate analysis.4 The mean number of needles used per implant in the current study was 17. Other institutions have reported between 15 to 33 needles used per implant.
No patient in the current study received pretreatment hormonal therapy. Within the literature, patients requiring hormonal therapy to decrease prostate size have had high rates of acute urinary morbidity in comparison with patients with a similar prostate volume who did not require hormones.5, 9 Patients in our trial were also prescreened for brachytherapy on the basis of their baseline urinary function. The mean IPSS score before therapy was 9. A high IPSS score has emerged on multivariate analysis in the literature as being predictive of higher rates of catheterization and acute urinary obstructive symptoms.4, 5, 8
In addition, all patients enrolled in this trial received perioperative tamsulosin. The use of second-generation and third-generation α‒blocker medications has been associated with a significant improvement in urinary obstructive symptoms since their availability in the late 1990s. Patients previously treated at our institution and included in our retrospective data who demonstrated a 10% catheterization rate had not routinely been treated with α‒blockers.
Merrick et al16 previously conducted a small trial of 22 patients who were randomized to either preoperative dexamethasone or a control group. They found no difference in the short-term catheterization rate, dosimetric quality, Day 0/Day 28 prostate volume, or clinical outcomes between the 2 groups. We have now conducted a larger randomized trial powered to detect a 10% improvement in the catheterization rate, which confirms the lack of improvement in the catheterization rate noted with the addition of dexamethasone. In addition, our experience did not demonstrate a decrease in postimplantation edema on postoperative CT scan in patients receiving dexamethasone. Therefore, it is our opinion that dexamethasone should not be routinely be prescribed for this purpose.