Purpose: To define favorable pretreatment characteristics for overall survival (OS), progression-free survival (PFS), locoregional control, and freedom from distant metastasis for patients with recurrent and second primary head-and-neck cancer treated with concomitant chemotherapy and reirradiation.
Methods and materials: Our study population comprised a subset of 115 previously irradiated patients without overt metastases from 304 poor-prognosis head-and-neck cancer patients treated in seven consecutive phase I-II protocols. Of the 115 patients, 49, who had undergone surgical resection, were treated with a median of four cycles of concurrent chemotherapy and reirradiation and 66, who had not undergone surgical resection, were treated with a median of five cycles. The following regimens were used: 5-fluorouracil and hydroxyurea concurrent with reirradiation (FHX) (n=14), cisplatin plus FHX (n=23), paclitaxel plus FHX (n=42), gemcitabine plus paclitaxel and 5-fluorouracil concurrent with reirradiation (n=26), and irinotecan plus FHX (n=10).
Results: The median lifetime radiation dose was 131 Gy. The median follow-up for surviving patients was 67.4 months (range, 18.5-158.7). The median OS and PFS was 11 and 7 months (range, 0.2-158.7), respectively. The 3-year OS, PFS, locoregional control, and freedom from distant metastasis rate was 22%, 33%, 51%, and 61%, respectively. Multivariate analysis identified reirradiation dose, triple agent (cisplatin-, paclitaxel-, or gemcitabine-containing chemotherapy), and surgery before protocol treatment as independently prognostic for OS, PFS, and locoregional control. Triple-agent chemotherapy was prognostic for freedom from distant metastasis. Nineteen patients died of treatment-related toxicity, five of these of carotid hemorrhage.
Conclusion: For recurrent and second primary head-and-neck cancer, trimodality therapy with surgery, concurrent chemotherapy, and reirradiation for a full second dose offers potential for long-term survival. Owing to the substantial toxicity and lack of an optimal regimen, reirradiation of recurrent head-and-neck cancer should be limited to clinical trials.