PM-TOM: A Method for Finding Personalized Polypharmacy Therapies with Minimal Adverse Drug-Drug, Drug-Gene and Drug-Condition Interactions.

Abstract

Polypharmacy therapies, quite frequent in older populations, pose significant health risks for patients due to a high possibility of their cumulative adverse drug reactions. We present Personalized Medicine Therapy Optimization Method (PM-TOM) for the discovery of polypharmacy therapies with minimal drug-drug interactions, drug-gene interactions, and drug-condition interactions. The inputs of the PM-TOM heuristic algorithm are the patient's conditions, genome, and the drug products and therapy considered by a clinician. Its output is a comprehensive report that explains details of the optimal therapies for considered drug products and all drugs that treat the patient's conditions. PM-TOM was developed by using the Electronic Medical Records repository of the Personal Genome Project (PGP), and the public repositories: DrugBank and Comprehensive Toxicogenomics Database (CTD). Testing of PM-TOM showed potential for significant reduction of the cumulative adverse drug interactions in personalized polypharmacy therapies. In the group of patients with 8 to 17 conditions, PM-TOM reduced the average cumulative drug interactions from 22 to 5.83, and the group with 6 to 7 conditions from 11.17 to 3. These results encourage further research and development of clinical decision support tools like PM-TOM.