Predictors of response and survival after concurrent chemotherapy and radiation for locally advanced squamous cell carcinomas of the head and neck

BACKGROUND The objective of this study was to determine prognostic factors for response and survival on three consecutive institutional trials utilizing concurrent chemotherapy and radiation for locally advanced squamous cell carcinomas of the head and neck (SCCHN). METHODS Since 1985, patients with locally advanced SCCHN at the University of Maryland have been managed with concurrent chemotherapy and radiation therapy (RT). Three consecutive pilot studies have been performed evaluating the utility of weekly chemotherapy with standard fractionated RT. Chemotherapy consisted of carboplatin either alone (28 patients) or in combination with bleomycin (23 patients) or paclitaxel (60 patients). In all three studies, RT was given to 70.2 gray (Gy) at 1.8 Gy/fraction/day to the primary site. All patients had locally advanced SCCHN and were believed to be poor surgical candidates. Sixty-seven percent of patients had T4 disease, and 21% had T3 disease. Seventy-five percent of patients had N2–N3 disease. One hundred eleven patients were examinable for toxicity, response, and survival analysis. Factors including age, race, gender, primary site location, histologic grade, T classification, N classification, and treatment regimen were evaluated to identify predictors of these endpoints. RESULTS The median follow-up for patients treated on study 1 (carboplatin and RT) and study 2 (carboplatin and bleomycin [C + B]/RT) was 98 months, and it was 30 months for study 3 (carboplatin and paclitaxel [C + P]/RT). The complete response rates were 54%, 52%, and 70% respectively (P = 0.01). Multivariate analysis identified length of treatment break ( 1 week) as the only predictor of complete response to therapy. The local control for the entire group was 50%. The local control for C + P/RT was 63%, versus 32% and 36% for C/RT and C + B/RT respectively (P = 0.004). The 2-, 3-, and 5-year disease free and overall survivals for the entire population were 41%, 41%, and 35% and 42%, 36%, and 33%, respectively. The 3-year overall survival rates by treatment regimen were 18% (C/RT), 35% (C + B/RT), and 47% (C + P/RT; P = 0.01). On univariate analysis, age younger than 50 years (P = 0.01), treatment with C + P/RT (P = 0.005), and treatment break of 5 days or fewer (P < 0.05) were also predictive of improved overall survival. On multivariate analysis, only complete response (P < 0.0001) and treatment with C + P/RT (P = 0.02) remained statistically significant. CONCLUSIONS Chemoradiation provides patients with locally advanced SCCHN the opportunity for long term survival. Among the three chemoradiation regimens studied, C + P/RT was associated with the best complete response and survival rates. Complete response to therapy was the single most important predictor of overall survival. These three consecutive concurrent chemotherapy and radiation trials achieved a 5-year survival of greater than 30% for the entire population. These results support the use of this nonoperative approach for this group of patients with a historically poor prognosis. Cancer 2001;91:548–54. © 2001 American Cancer Society.