Objective: To compare the clinical efficacies of gefitinib versus paclitaxel/carboplatin in patients with advanced pulmonary adenocarcinoma.
Methods: A total of 51 advanced pulmonary adenocarcinoma patients were recruited from Guangdong General Hospital during October 2006 to September 2007. Eligible patients were ≥ 18 years old, either non-smokers or former light smokers and receiving no prior chemotherapy or biological/immunological therapy. According to performance status, smoking status and gender, they were selected with dynamic equilibrium randomized method 1: 1 to receive first-line gefitinib (250 mg/d) in gefitinib arm or carboplatin/paclitaxel (carboplatin, area under the curve 5 mg × ml⁻¹ × min⁻¹, 21-day cycle; paclitaxel, 200 mg/m², 21-day cycle in chemotherapy arm. The primary endpoint was progression free survival (PFS). And the secondary endpoints were objective response rate (ORR) and overall survival (OS).
Results: They were randomized into gefitinib arm (n = 25) and paclitaxel/carboplatin arm (n = 26). The median PFS was 4.2 months in gefitinib arm and 8.3 months in paclitaxel/carboplatin arm (P = 0.422); ORR 36.0% in gefitinib arm and 42.3% in paclitaxel/carboplatin arm (P = 0.645); Median OS 14.4 months in gefitinib arm and 15.0 months in paclitaxel/carboplatin arm (P = 0.290). Multifactorial Cox regression analysis showed that age (P = 0.004), epidermal growth factor receptor (EGFR) gene mutation status (P = 0.012) and subsequent treatments (platinum-based chemotherapy, P = 0.001; EGFR-tyrosine kinase inhibitors (TKI), P = 0.005) were the significant predictors of OS.
Conclusions: No significant differences exist in terms of efficacy and survival between first-line gefitinib and paclitaxel/carboplatin for Chinese patients with pulmonary adenocarcinoma who are non-smokers or former light smokers. And age, EGFR gene mutation status and subsequent treatments are significant predictor of OS. However, first-line gefitinib should not be recommended for advanced non-small cell lung cancer (NSCLC) patients only based on clinical factors, due to a very small sample-size in our study.