With new agents entering the market, the sequencing of first-line (Tx1), second-line (Tx2), and subsequent chemotherapy/biologics regimens are being examined. We examined how Tx1 regimens impacted the likelihood of receiving Tx2 among metastatic colon cancer (mCC) patients. Surveillance, Epidemiology and End Results (SEER)-Medicare data were used to identify elderly mCC patients between 2003 and 2007. The inverse probability weighting Cox regression method was utilized to study the relationship between receipt of Tx2 and Tx1 regimens, controlling for patient-level factors. Of the 7895 elderly patients identified, 3211 (41%) received Tx1 of which 1440 proceeded to Tx2. The impact of Tx1 on receipt of Tx2 varied by the specific regimens utilized. As compared to 5FU/LV users, IROX (Hazard Ratio [HR] = 0.03; P < 0.01) and IROX + Biologics (HR = 0.20; P < 0.01) users were less likely to receive Tx2; (oxaliplatin) OX + Biologics (HR = 1.26; P < 0.01) users were more likely to receive Tx2. Significant patient-level factors included: Hispanic ethnicity (HR = 0.67; P < 0.01); being married (HR = 0.87; P = 0.01); proxy for poor performance status (HR = 0.82; P = 0.05); each 10-year age increment (HR = 1.14; P < 0.01); and State buy-in status (HR = 1.21; P = 0.01). The specific first-line regimen does impact mCC patients' likelihood of receiving Tx2 in clinical practice. Elderly mCC patients, their health care providers, and policy makers will benefit from new evidence about the impact of sequencing of treatment lines.
Keywords: Chemotherapy/biologics treatment; SEER-Medicare; inverse probability weighting Cox regression; metastatic colon cancer; receipt of second-line treatment; treatment history.
© 2014 The Authors. Cancer Medicine published by John Wiley & Sons Ltd.