Purpose: To determine the efficacy and safety of induction chemotherapy followed by concomitant chemoradiotherapy in the treatment of stage III non-small cell lung cancer and whether the response to induction chemotherapy can predict the response to subsequent chemoradiotherapy and survival.
Materials and methods: Between December 1987 and June 1993, 46 patients with previously untreated stage III non-small cell lung cancer received every 21 days induction chemotherapy (ICT) including three cycles of 5-fluorouracil (600 mg/m2/d in short infusion from d1 to d5), cisplatin (15 mg/m2/d from d1 to d5), etoposide (50 mg/m2/d from d1 to d5) and hydroxyurea (1,500 mg/d from d1 to d5). The first 21 patients also received bleomycin (3 mg/m2/d from d1 to d5). All patients received concomitant chemotherapy and had chest radiotherapy (CCRT). Patients received irradiation (65 Gy/33-6 fractions/7 weeks) on d25 after the third cycle of chemotherapy. Concomitant chemotherapy was composed of cisplatin (20 mg/m2) and 5-fluorouracil (500 mg/m2) that were administered each Monday and Thursday during radiotherapy. Maintenance chemotherapy consisted of thiotepa (10 mg/m2) and methotrexate (10 mg/m2) that were administered every 2 weeks for 6 months.
Results: Pulmonary toxicity was observed in four out of 21 patients who had received bleomycin and subsequently developed pulmonary fibrosis, leading to death for two of them. ICT alone produced five complete responses (11%) and 13 partial responses (28%). The combination of chemotherapy and radiotherapy led to 19 complete responses (41%) and 14 partial responses (30%). Eighteen of the 18 responders (100%) to ICT responded to subsequent CCRT, of whom 13 (72%) became complete responders. Fifteen of the 28 non-responders to ICT (53%) responded to CCRT, six of them being complete responders (21%) (P < 0.001). The median overall survival rate was 17 months when considering all patients, 25 months in patients responding to ICT and 13 months in non-responders. The 2-year survival rates were 28, 55 and 11%, respectively (P < 0.05). ICT did not influence the rate of subsequent metastatic events. However, locoregional reprogression was lower in responders to ICT. The number of metastatic events was not significantly related to response to ICT. By contrast, the rate of local failure was higher when there was resistance to ICT (75% versus 39%). Out of the 19 complete responders to CCRT (13 responders to ICT and six non-responders to ICT), four developed secondary locoregional reprogression (21%) and six developed metastatic disease (31%). In complete responders to CCRT, the rate of locoregional failure was 15% in responders to ICT (2/13) and 33% (2/6) in non-responders to ICT. Four out of the 13 responders to CCRT after response to ICT (31%) and two out of the six complete responders to CCRT developed metastatic disease after non-response to ICT.
Conclusion: There is a statistically significant relationship not only between the response to ICT and the response to CCRT, but also between the response to ICT and the local outcome and survival.