Compliance and toxicity profiles of the CARISSA trial of postoperative cisplatin-based chemoradiation ± gefitinib in head and neck cancer

Abstract

e22152 Background: Concurrent cisplatin and postoperative irradiation (PORT) is the standard treatment for resected poor-risk squamous cell carcinoma of the head and neck (HNSCC) . However, a substantial number of patients fail to respond. Anti-EGFR agents, such as gefitinib, may improve outcomes and molecular markers might predict treatment efficacy. This was the basis for the CARISSA trial. There was some concern after hemorrhagic events occurred in a large independent trial of recurrent unresectable HNSCC. This caused a temporary interruption and safety profile analysis of CARISSA, which resumed after security reports showed no excess of hemorrhagic events. Methods: The CARISSA multicenter blinded randomized phase II trial compares PORT and cisplatin alone (arm A) or in combination with the EGFR Inhibitor ZD1839 gefitinib (arm B) in HNSCC and analyses EGFR-binding, gene polymorphisms, Western Blotting and immunohistochemistry. Patients are followed weekly. Toxicities are graded according to the NCI/CTCv.3 scale. Results: Thirty-eight patients have been included. Compliance to cisplatin was 56% in arm A and 44% in arm B (NS). Compliance to gefitinib was 60% with half cases interrupting treatment for toxicity. There were 18 and 20 grade (G) 3/4 non-hematological toxicities, 12 and 6 hematological G3/4 toxicities in arms A and B, respectively. Mucositis was the most common toxicity. G3 mucositis occurred in 4 and 5 cases in arms A and B, dermatitis in three cases in each arm and diarrhea in one case in each arm (NS). There were 7 cancer- related deaths and 3 of other causes. One tumor-bleeding event occurred in the arm without gefitinib. Full-dose radiotherapy was delivered in all but one case in each arm. Mean doses were 63.9 ± 1.9 (arm A) and 65.5 ± 0.9 Gy (arm B) or 64.7 ± 0.9 (arm A) and 65 ± 0.9 Gy (arm B) in groups where RT was uninterrupted or temporarily interrupted, respectively (NS). The two patients who did not receive full dose received 60 Gy. Conclusions: Compliance to PORT was similar with cisplatin alone or cisplatin + gefitinib. While there was no excess of hemorrhagic morbidity postoperatively (no gross disease), accrual continues with particular caution. No significant financial relationships to disclose.