Validation of A Global Economic Model to Evaluate The Cost-Effectiveness Of Targeted Treatments Using Companion Diagnostics In Advanced/Metastatic Cancer Treatment Using Kras Testing For Cetuximab Therapy In Metastatic Colorectal Cancer


      A global economic model to evaluate the cost-effectiveness of targeted treatment using companion diagnostics in advanced/metastatic cancer treatment has been recently developed. Targeted treatment with cetuximab guided with KRAS testing in metastatic colorectal cancer (mCRC) is well implanted and has been subject of different economic evaluations. The objective of this study was to use data on cetuximab guided with KRAS testing in mCRC to validate the global economic model in advanced/metastatic cancer.


      Survival data and incidence of adverse events (infections, rash, and pain) were obtained from the CO. 17 study, which is a pivotal trial on the efficacy of cetuximab used according to the KRAS status vs. best supportive care (BSC). Prevalence of KRAS mutation, specificity/sensitivity and cost of KRAS testing, and the cost of cetuximab were obtained from the literature. Other parameters for the cost-effectiveness analysis were intrinsic components of the global model. Analyses were conducted from a Ministry of Health (MoH) and a societal perspective.


      The incremental cost-utility ratio (ICUR) of cetuximab treatment guided with KRAS testing compared to no testing and BSC to all patients from a MoH perspective was estimated at $192,814/QALY and at $297,198/QALY from a societal perspective. According to the deterministic analysis results, the ICUR of cetuximab with KRAS testing compared to no testing and BSC varied between $149,845/QALY and $387,971/QALY from a MoH perspective and between $232,417/QALY and $492,354/QALY from a societal perspective. Parameters that have the greatest impact on base-case ICURs were the cost of cetuximab and the specificity of KRAS testing.


      Validation using cetuximab with KRAS testing suggests that the global economic model is robust, as results obtained were in the range than those reported in the literature. Although further validations will be performed, the global model appears to produce quick, but accurate estimates.