Background: For many years, high-dose radiation therapy was the standard treatment for patients with locally or regionally advanced non-small-cell lung cancer (NSCLC), despite a 5-year survival rate of only 3%-10% following such therapy. From May 1984 through May 1987, the Cancer and Leukemia Group B (CALGB) conducted a randomized trial that showed that induction chemotherapy before radiation therapy improved survival during the first 3 years of follow-up. Purpose: This report provides data for 7 years of follow-up of patients enrolled in the CALGB trial. Methods: The patient population consisted of individuals who had clinical or surgical stage III, histologically documented NSCLC; a CALGB performance status of 0-1; less than 5% loss of body weight in the 3 months preceding diagnosis; and radiographically visible disease. Patients were randomly assigned to receive either 1) cisplatin (100 mg/m(2) body surface area intravenously on days 1 and 29) and vinblastine (5 mg/m(2) body surface area intravenously weekly on days 1, 8, 15, 22, and 29) followed by radiation therapy with 6000 cGy given in 30 fractions beginning on day 50 (CT-RT group) or 2) radiation therapy with 6000 cGy alone beginning on day 1 (RT group) for a maximum duration of 6-7 weeks. Patients were evaluated for tumor regression if they had measurable or evaluable disease and were monitored for toxic effects, disease progression, and date of death. Results: There were 78 eligible patients randomly assigned to the CT-PT group and 77 randomly assigned to the RT group. Both groups were similar in terms of sex, age, histologic cell type, performance status, substage of disease, and whether staging had been clinical or surgical. All patients had measurable or evaluable disease at the time of random assignment to treatment groups. Both groups received a similar quantity and quality of radiation therapy. As previously reported, the rate of tumor response, as determined radiographically, was 56% for the CT-PT group and 43% for the RT group (P = .092). After more than 7 Sears of follow-up, the median survival remains greater for the CT-RT group (13.7 months) than for the RT group (9.6 months) (P = .012) as ascertained by the logrank test (two-sided). The percentages of patients surviving after years 1 through 7 were 54, 26, 24, 19, 17, 13, and 13 for the CT-RT group and 40, 13, 10, 7, 6, 6, and 6 for the RT group. Conclusions: Long-term follow-up confirms that patients with stage III NSCLC who receive 5 weeks of chemotherapy with cisplatin and vinblastine before radiation therapy have a 4.1-month increase in median survival. The use of sequential chemotherapy-radiotherapy increases the projected proportion of 5-year survivors by a factor of 2.8 compared with that of radiotherapy alone. However, inas-much as 80%-85% of such patients still die within 5 years and because treatment failure occurs both in the irradiated field and at distant sites in patients receiving either sequential chemotherapy-radiotherapy or radiotherapy alone, the need for further improvements in both the local and systemic treatment of this disease persists.
