PATTERN OF FAILURE IN LONG-TERM SURVIVORS AFTER RADIOCHEMOTHERAPY FORINOPERABLE HEAD AND NECK-CANCER

Background: After conventionally fractionated radical radiotherapy for advanced unresectable squamous cell carcinoma of the tongue, the floo r of mouth and the supraglottic larynx, oro- and hypopharynx loco-regi onal tumor control rates rarely exceed 30%. Therefore, cytotoxic chemo therapy has been added in an attempt to improve loco-regional turner c ontrol rates and reduce the incidence of distant spread and ultimately long-term survival. Still, in these patients comorbidity may often co mpromise possible improvements achieved by treatment intensification. Patients and Methods: In a prospective phase II study from 1984 to 198 9, patients with loco-regionally advanced head and neck: carcinoma wer e treated with 3 courses of simultaneous radio-chemotherapy. Radiother apy was administered in two fractions per day 1.8 Gy each. During one course from day 3 to day 11 23.4 Gy were delivered in 13 fractions. Ch emotherapy consisted of cis-Platin 60 mg/m(2), leucovorin 50 mg/m(2) i ntravenous bolus and 5-FU 350 mg/m(2) intravenous bolus on day 2 and l eucovorin 100 mg/m(2)/24-hour continuous infusion and 5-FU 350 mg/m(2) /24 h from day 2 to 5. Treatment was repeated on days 22 and 44, total radiation dose was 70.2 Gy. From day 12 to 21 and 34 to 43 treatment breaks were scheduled. Results: 105 patients are evaluable with a foll ow-up from 36 to 105 months in surviving patients. The crude overall s urvival rate at 5 years is 29%, the probability of progression-free su rvival is 39% (Kaplan-Meier estimations). Loco-regional recurrences we re observed in 52 patients after a median interval of 11 (2-81) months . 16 patients developed distant metastases after a median interval of 21 (6-81) months. Conclusions: Compared with data from the literature the pattern of relapse is very similar to that observed after radical radiotherapy alone. Median interval to distant spread appears to be pr olonged by simultaneous chemotherapy but this may be due to prolonged follow up. Three-year figures of loco-regional control and survival ar e valid parameters to apreciate long-term data. Since nearly a third o f patients are long-term survivors chronic toxicity deserves special a ttention in future trials.