Decision Analyses of Genomic Strategies for Colorectal Cancer Precision Medicine

Background: The current landscape of genomic medicine is undergoing rapid innovation, and genome-based interventions are being increasingly used in the screening, diagnosis and treatment decisions in clinical practice. In an environment of limited health care resources, evidence regarding the economic and clinical outcomes of genomic tests may be important in facilitating the decisions of policymakers, clinicians, patients and other decision makers. Objectives: The overall objective of the dissertation was to better understand the landscape of health economic challenges and prospects used for screening, or prognosis of a heterogeneous population of colorectal cancer patients. To fulfill the main objective of this dissertation, the following three studies were carried out. The objective of the first study was to better understand and evaluate the quality of health economic evidence in the peer-reviewed literature. In the second study, a cost-effectiveness analysis was conducted comparing four different genomic tests used in a cohort of stage II colorectal cancer patients, and the quality of life of treatment decisions was investigated. The third study involved the development of a novel model in order to better understand and evaluate the pre-screening to post-diagnostic pathways of colorectal cancer disease. Methods: For study 1, a systematic search of the literature between January 2006 and May 2018, for economic studies of colorectal cancer genomics was conducted and the economic studies for colorectal cancer genomics meeting inclusion criteria were evaluated using a validated instrument, the Quality of Health Economic Studies (QHES) scale. For study 2, a decision tree model was developed to evaluate and compare the cost-effectiveness of four different marketed genomic tests (i.e. a 12-gene assay, an 18-gene expression assay, a 482-gene signature and the Immunoscore assay) used for the prognosis of stage II colon cancer patients. For study 3, a new whole disease model was developed to estimate the costs and outcomes of using different genomic tests as compared with non-genomic approaches from pre-screening to post-diagnostic treatment pathways in colorectal cancer patients. Results: In study 1, the average QHES score for economic studies that met eligibility criteria was 82.18. The majority of economic analyses in our study were based on randomized controlled trials (n=35), designed with Markov models (n=20) and provided a clear description of sensitivity analyses (n=27). Twelve independent variables as predictors of the quality were identified. In study 2, the cost of the Immunoscore assay strategy in stage II colorectal cancer patients was estimated to be US$23,564 as compared with the 12-gene assay strategy at US$24,545, the 18-gene assay strategy at US$28,374, and 482-gene signature treatment strategy at US$33,315. The use of strategies based on the Immunoscore assay, the 12-gene assay, the 18-gene assay, and the 482-gene signature resulted in gains of 3.903, 3.623, 3.677, and 3.704 QALYs respectively. In study 3, the whole-disease model demonstrated that patients who received non-genomic interventions for colorectal cancer treatment spent US$13,978 and gained 2.12 QALYs. On the other hand, patients who used genomic tests for colorectal cancer treatment spent US$10,519 and gained 2.32 QALYs. Conclusions: The overall methodological quality of health economic evaluations on colorectal cancer genomics appears to be generally good as evaluated by the QHES. The cost-effectiveness analysis in study 2 suggested that the Immunoscore assay may be a dominant strategy at a threshold willingness-to-pay of $50,000 per QALY. However, this test has only recently begun to be marketed in the US as of January 2019. The results of study 3 indicated that the use of genomic tests (multi-target stool DNA test and the 12-gene assay) might be a more cost-effective strategy than standard approaches at a threshold willingness to pay of $50,000 per QALY