Adjuvant chemotherapy for non-small cell lung cancer.

BACKGROUND Adjuvant chemotherapy (ACT) is known to improve survival in patients with early-stage non-small cell lung cancer. Herein, we describe chemotherapy regimens used, dose modifications, survival, and treatment-related toxicity in the general population. METHODS All cases of non-small cell lung cancer diagnosed in Ontario in the period 2004-2006 who underwent surgical resection (n = 3354) were identified using the Ontario Cancer Registry in this population-based retrospective cohort study. We linked electronic records of treatment to the registry to identify all cases treated with ACT (n = 1032) and describe drugs, regimens, and dosages delivered. As a proxy measure of ACT-related toxicity, we evaluated deaths and hospitalizations within 16 weeks of starting ACT. Factors associated with dose modification were evaluated by logistic regression. The Cox proportional hazards model was used to describe associations between patient-, disease-, and treatment-related factors and survival. RESULTS ACT regimens were identified for 584 of 1032 ACT cases. Almost all cases included cisplatin- or carboplatin-based regimens (478/584, 82%, and 99/584, 17%, respectively). The most common regimen was a vinroelbine/cisplatin doublet (412/584, 71%); 64% of these cases had a dose reduction or omission. Dose modification was not associated with inferior survival on multivariate analysis. Twelve percent of all ACT cases were admitted to hospital within 16 weeks of starting ACT, and there was a 1.6% death rate potentially attributable to ACT. Survival of all ACT cases was comparable with outcomes reported in clinical trials. CONCLUSIONS ACT regimens used, toxicity, and survival outcomes in the general population are comparable with those reported in clinical trials. Dose modifications used in clinical practice are not associated with inferior survival.