A systematic approach to companion diagnostic development: A case study for omacetaxine (OM) for the treatment of chronic myelogenous leukemia (CML).
100 Background: In the age of genomically informed medicine, therapeutic development carries with it the imperative to employ genomics in patient selection. Physicians expect genomics-based methods to identify treatments likely to be effective, and identify anomalies likely to cause adverse response for a given patient. Companion diagnostics should support such rule-in and rule-out decisions. We demonstrate a systematic approach to companion diagnostics that leverages new methods in translational bioinformatics and clinical economics. CML therapies have been at the forefront of genomically informed medicine. Early TKI inhibitors targeting the BCR-ABL fusion protein are highly effective. With time, however, they induce resistance-creating mutations in many patients. Omacetixine, a translation inhibitor, was expected to help this CML patient subpopulation which has few therapeutic alternatives. We use this well-characterized drug and publicly available data to demonstrate a prospective approach to companion diagnostics. Methods: We used translational bioinformatics, incorporating pathway, cell line, and patient data to identify biologically plausible biomarkers from which alternative companion diagnostic paths were constructed. These alternatives were analyzed using a modified version of the MIT Stratified Medicine Model to assess the clinical economics of each path. Results: From a systematic look at the biology of the disease, the unique mechanism of action of OM and the clinical need, we identified 3 alternative companion diagnostics for OM. Economic analyses quantified the trade-offs of targeting different subpopulations for the indication, clarifying the impact of biomarker selection based on clinical need or biology. Other analyses have shown that eNPV can be halved or doubled based on strategy choice. Conclusions: Combining applied translational bioinformatics and stratified medicine economics provides an effective approach to companion diagnostic selection. This approach can reduce drug-development cost and clinical risk while providing physicians with better genomics-based methods for clinical decision-making.