Phase I/II study of the CAPABLE regimen for patients with poorly differentiated carcinoma of the nasopharynx

PURPOSE: To evaluate the role of aggressive combination chemotherapy in patients with poorly differentiated carcinoma of the nasopharynx (NPC). PATIENTS AND METHODS: In a prospective phase I/II study, 90 chemotherapy-naive patients with NPC (21 with very advanced locoregional disease, 18 with locoregional persistent and/or recurrent disease postradiotherapy, and 51 with metastatic disease) were treated with cyclophosphamide, doxorubicin, cisplatin, methotrexate, and bleomycin (CAPABLE). Two schedules of this regimen were used over a 9-year period, with the second schedule being a modification of the first in an attempt to minimize treatment-related toxicity. RESULTS: Of 21 patients with very advanced local disease, one had a complete response (CR) and 17 had partial responses (PRs) (response rate, 86%). Seventeen of these 21 patients had subsequent radiotherapy. Of 17 patients with measurable locoregional disease either persistent and/or recurrent postradiotherapy, there were four CRs and three PRs (response rate, 41%). Of 44 patients with measurable metastatic disease, there were three CRs and 32 PRs (response rate, 80%). The median survival durations for these three groups of patients were 47, 16, and 14 months, respectively. The two chemotherapy schedules had similar received dose-intensities (RDIs) and produced similar response rates and survival. Toxicity was severe with frequent mucositis and myelosuppression. Overall, 37 patients required at least one hospital admission for management of toxic side effects and there were seven drug-related deaths. Three of the deaths were due to fulminant hepatitis, likely from reactivation of hepatitis B. CONCLUSION: This aggressive regimen provides a high rate of tumor response, but limited palliation for most patients with recurrent or metastatic NPC. The results with chemotherapy followed by radiotherapy for patients with very advanced local disease are encouraging, but proof of benefit would require evaluation in a randomized trial.