Gemcitabine, cisplatin, and hyperfractionated accelerated radiotherapy for locally advanced non-small cell lung cancer

Background: Due to potent radiosensitization and potential serious or fatal toxicity, concurrent gemcitabine and irradiation should only be applied within clinical trials. We here present experience from a phase I-II clinical trial for patients with locally advanced non-small cell lung cancer (NSCLC) treated with hyperfractionated accelerated radiotherapy and concurrent low-dose gemcitabine. Methods: Eligible patients had locally advanced inoperable NSCLC without pleural effusion, Eastern Cooperative Oncology Group performance status 0-1, were chemotherapy naive and had no previous radiotherapy to the chest, and had adequate hematopoietic, liver, and kidney function. Routine brain computed tomography was not performed, and positron emission tomography/computed tomography was not available. Treatment consisted of three parts: induction chemotherapy with gemcitabine and cisplatin in standard doses, local treatment with concurrent chemotherapy and radiotherapy, and consolidation chemotherapy. Patients were irradiated with opposed AP-PA and oblique fields, using 2.5-D treatment planning. Although corrections for inhomogeneous tissue were made, volume of total lung receiving >= 20 Gy (V20) could not be determined. The trial started as phase 1, aimed to determine the dose-limiting toxicity and maximal tolerated dose (MTD) for concurrent hyperfractionated radiotherapy (1.4 Gy twice daily) and gemcitabine 55 mg/m(2) twice weekly as a radiosensitizer. Phase 11 of the trial then continued at the level of MTD. Results: Twenty-eight patients with NSCLC, nine patients with stage IIIA, 16 patients with 11113, and three patients with an inoperable recurrence after previous surgery, entered the trial. The first 12 patients entered Phase I of the trial at the initial level of 42 Gy in 30 fractions in 3 weeks. Dose-limiting toxicity was acute esophagitis; 47.6 Gy in 34 fractions in 3.5 weeks was the MTD for this regimen of concurrent chemotherapy and radiotherapy. In phase 11 of the trial, this dose was applied to the next 16 patients. Among all 28 patients, 13 had grade 3 or 4 acute toxicity: esophagitis (eight patients), neutropenia (eight patients), thrombocytopenia (four patients), and anemia (two patients). No pulmonary toxicity and no persistent or serious late toxicity were seen. Local and/or regional relapse was documented in nine patients, distant in five and both locoregional and distant in 10 patients. The most common sites of distant spread were the brain and lung in eight and six patients, respectively. At 2 years, progression-free survival was 43% and overall survival was 57%. After 43 to 85 months of follow-up, seven patients are alive, of whom six (21%) are without evidence of disease and may be regarded as long-term survivors. Among the long-term survivors, one was in the group irradiated to 42 Gy and six in the groups irradiated to 47.6 Gy. Conclusion: Judging from current standards, the methods used in diagnostics and in planning of radiotherapy were suboptimal. Using modem radiotherapy planning, a higher MTD, possibly a different profile of toxicity, and better long-term results may be expected. The high incidence of brain relapse emphasizes the need for careful screening for unsuspected brain disease before treatment and the importance of clinical studies on prophylactic cranial irradiation for patients with locally advanced NSCLC. Although the small number of patients in this study precludes any definitive conclusion, it appears that our program of concurrent chemotherapy and radiotherapy offers a chance for disease control at least comparable to previously described programs for inoperable lung cancer.