Cigarette smoking as a strong prognostic factor after first-line chemotherapy in advanced non-small cell lung cancer.

Abstract

e18099 Background: Sex, disease stage, weight loss, and performance status (PS) are well-established clinical prognostic factors after palliative chemotherapy (CT) in non-small cell lung cancer (NSCLC). In contrast, the correlation between smoking status and benefit to chemotherapy is not routinely reported in clinical trials. Methods: We retrospectively reviewed the medical charts of patients treated with first-line CT between 2000 and 2004 at the Brazilian National Cancer Institute. The primary outcome was overall survival (OS) and prognostic factors were analyzed using Cox model. Results: One hundred seventy patients were treated with carboplatin-etoposide combination during the period. Median age was 60 years (22-85), 73% were male, 56.7% had adenocarcinoma, and 81.5% were diagnosed in stage IV disease. Forty-five percent of patients had more than 5% weigh loss documented at the diagnosis, and most patients had PS 0-1 (83.3%). The median number of cycles was 4 (1-8) and only twenty patients received second-line CT (11.5%). Smoking status data was available for 157 patients, being 83.4% ever-smokers and 16.6% never-smokers. The median OS was 7.1 months (95% CI; 6.5-7.6). Smoking status was strongly correlated with OS among PS 0-1 patients (median 7.2 vs 12.4 months for smokers and non-smokers, respectively; p = 0.016), but not in PS 2 (3.7 vs 1.6 months, respectively; p = 0.19). In the multivariate analysis, smoking status (p = 0.019), PS (p < 0.0001), and weigh loss (p = 0.024) were independent prognostic factors. However, the OS was not worst for patients with ≥ 15 pack-years cigarette smoking history when compared to those with < 15 (p = 0.202). Disease stage (p = 0.73) and sex (p = 0.41) did not predict OS. Conclusions: Past smoking history strongly predicted worst outcome after palliative CT in this cohort. This is a unique population, since all the patients received the same CT combination, and nearly 90% was exclusively treated with first-line schema. Since crossover was not an issue, OS is a very reliable outcome. The 7.1-month OS is inferior to the currently described outcomes in the literature, which is probably related to the absence of second-line CT, as well as the use of a second- generation doublet. No significant financial relationships to disclose.