A pilot study of topical intrarectal application of amifostine for prevention of late radiation rectal injury.

Purpose: Clinical symptomatic late injury to the rectal wall occurs in about one-third of patients with prostate cancer treated with external beam irradiation. Reducing the physical dose to the anterior rectal wall without a similar reduction in the posterior peripheral zone is difficult because of the proximity of the prostate to the anterior rectal wall. On the basis of our previous observations in an animal model that intrarectal application of amifostine resulted in very high concentrations of amifostine and its active metabolite WR-1065 in the rectal wall, a Phase I dose-escalation clinical trial was undertaken.

Methods And Materials: Twenty-nine patients with localized prostate cancer were accrued. Eligibility criteria included histologically confirmed adenocarcinoma, Karnofsky performance status >or=70, and no pelvic lymphadenopathy or distant metastases. The total dose to the prostate was 70.2 Gy in 20 patients and 73.8 Gy in 9 patients. Therapy was delivered using a 4-field technique with three-dimensional conformal planning. Amifostine was administered intrarectally as an aqueous solution 30 min before irradiation on the first 15 days of therapy. Amifostine was escalated in cohorts from 500 to 2500 mg. Proctoscopy was performed before therapy and at 9 months after completion. Most patients underwent repeat proctoscopy at 18 months. On Days 1 and 10 of radiotherapy, serum samples were collected for pharmacokinetic studies. The clinical symptoms (Radiation Therapy Oncology Group scale) and a proctoscopy score were assessed during follow-up.

Results: All patients completed therapy with no amifostine-related toxicity at any dose level. The application was feasible and well tolerated. No substantial systemic absorption occurred. With a median follow-up of 26 months, 9 patients (33%) developed rectal bleeding (8 Grade 1, 1 Grade 2). At 9 months, 16 and 3 patients developed Grade 1 and Grade 2 telangiectasia, respectively. This was mostly confined to the anterior rectal wall. No visible mucosal edema, ulcerations, or strictures were noted. No significant differences were found between the proctoscopy findings at 9 and 18 months. Four patients (14%) developed symptoms suggestive of radiation damage that, on sigmoidoscopy, proved to be secondary to unrelated processes. These included preexisting nonspecific proctitis (n = 1), diverticular disease of the sigmoid colon (n = 1), rectal polyp (n = 1), and ulcerative colitis (n = 1). Symptoms developed significantly more often in patients receiving 500-1000 mg than in patients receiving 1500-2500 mg amifostine (7 [50%] of 14 vs. 2 [15%] of 13, p = 0.0325, one-sided chi-square test).

Conclusion: Intrarectal application of amifostine is feasible and well tolerated. Systemic absorption of amifostine and its metabolites is negligible, and close monitoring of patients is not required with rectal administration. Proctoscopy is superior to symptom score as a method of assessing radiation damage of the rectal wall. The preliminary efficacy data are encouraging, and further clinical studies are warranted.