Treatment patterns and cost-effectiveness of therapeutic sequences among advanced colorectal cancer patients

ackground: Colorectal cancer (CRC) ranks third among the most common cancers and one in four CRC patients are diagnosed at metastatic stage. Metastatic colorectal cancer (mCRC) has a poor prognosis, with an overall survival of 5% to 13% at five years. Over the last decade multiple chemotherapies and targeted biologics have been approved for mCRC and patients may receive these chemotherapies and targeted biologics in different sequences. Limited evidence exists with regards to the current usage patterns, comparative effectiveness and cost-effectiveness of treatment sequences for elderly mCRC patients. Objective: The first aim was to describe the usage patterns of treatments (chemotherapy and targeted biologics) and treatment sequences administered to mCRC patients and the factors associated with the receipt of common treatment sequences. The common treatment sequences of interest were: 1) first-line oxaliplatin/irinotecan followed by second-line oxaliplatin/irinotecan + bevacizumab (OI-OIB), 2) first-line oxaliplatin/irinotecan + bevacizumab followed by second-line oxaliplatin/irinotecan + bevacizumab (OIB-OIB), 3) OI-OIB followed by a third-line targeted biologic (OI-OIB-TB), and 4) OIB-OIB followed by a third-line targeted biologic (OIB-OIB-TB). The second and third aims of the study were to determine the comparative effectiveness and cost-effectiveness, respectively for the above mentioned sequence. Methods: A retrospective cohort study was conducted in patients diagnosed with mCRC from January 2004 through December 2009 using the Surveillance, Epidemiology and End Results-Medicare linked database. The treatment continuum administered to elderly mCRC patients was empirically identified. Comparative effectiveness of the treatment sequences was evaluated using fixed time and time dependent (immortal time adjusted) Cox-proportional hazard regression for all-cause mortality. A probabilistic discrete event simulation model assuming Weibull distribution was developed to evaluate the cost-effectiveness of common treatment sequences. A probabilistic sensitivity analysis was performed to account for parameter uncertainty. Costs (2014 U.S. dollars) and effectiveness were discounted at an annual rate of 3%. Results: Of 4,418 mCRC patients who received treatment, 1,370 (31%) received first-, second-, and third-line, 1,164 (26%) received first-, and second-line, and 1,884 (43%) received only first-line. The most common first-line of treatment for mCRC patients were 5-fluorouracil/leucovorin + oxaliplatin (FOLFOX) + bevacizumab (23%) and FOLFOX (23%). The most common treatment sequence was first-line oxaliplatin or irinotecan followed by second-line oxaliplatin or irinotecan + bevacizumab followed by a third-line targeted biologic. Forty seven percent of patients who received first-line therapy also received a targeted biologic and the factors associated were age, comorbidity score, cancer site, geographic location and year of diagnosis. Fixed time model Cox-proportional hazard regression showed that as compared to OI-OIB, statistically significantly lower hazard ratios for all-cause mortality were observed for patients receiving treatment sequences OIB-OIB (0.60, 95% CI: 0.46-0.77), OI-OIB-TB (0.53, 95% CI: 0.42-0.67), and OIB-OIB-TB(0.40, 95% CI: 0.31-0.52). In the base case cost-effectiveness analyses, at the willingness-to-pay (WTP) threshold of...