PII-050 - PHARMACOGENOMIC-AUGMENTED DRUG THERAPY IN MAJOR DEPRESSIVE DISORDER: A SYSTEMATIC REVIEW OF GENES AND A CASE FOR ARTIFICIAL INTELLIGENCE-AIDED GENOME-WIDE PANEL DESIGN.

C. Grant, K. Delaney, L. Jackson, J. Bobo, H. Leslie, S. Scaletty, L. Wang, R. Weinshilboum, P. Croarkin, A. Athreya; Mayo Clinic, Rochester, MN, USA.

Background: Major depressive disorder (MDD) is the leading neuropsychiatric disorder. A systematic evaluation of the genes used in Pharmacogenomic (PGx)-augmented MDD drug therapy is lacking.

Methods: This work systematically reviewed studies of commercial antidepressant (AD) PGx panel genes. Study design (ADs, outcomes, patients, methods [e.g., linear models, artificial intelligence (AI)]) and results were assessed.

Results: Data from 301 MDD antidepressant PGx studies were extracted following a multi-database search. These evaluated 43 genes on 46 panels.
1. Studies assessed 39 drugs/combinations, most commonly pooling data from multiple ADs. Caucasian or Asian individuals comprised 87% of participants. Most studies (N=282) used candidate gene approaches (genome/exome-wide studies [N=19]). Integration of multi-modal data (PGx, biological, clinical) through AI (N=23) more reliably found associations with efficacy endpoints.
2. Outcomes included remission, drug response, change in MDD severity, and side effects. Across studies, these had 30, 22, 16, and 26 definitions, respectively.
3. Top PGx associations by outcome were: ABCB1 with side effects (7/9 studies [78%] with a significant association); SLC6A4 with change in MDD severity (17/22 studies [77%]); and HTR2A with response and remission (14/23 [61%], 12/24 [50%]).

Conclusion: Approaches towards identifying PGx genes for MDD management with ADs are vastly heterogenous. Consequently, current commercial PGx genes lack broad replicability in predicting common AD efficacy endpoints. Panel designs guided by AI and genome-wide PGx data hold promise to optimize efficacy of MDD care through detection of novel PGx genes and gene combinations given multiple drug response phenotypes, AD combinations, and diverse populations.

Oslin, D. W. et al. Effect of Pharmacogenomic Testing for Drug-Gene Interactions on Medication Selection and Remission of Symptoms in Major Depressive Disorder: The PRIME Care Randomized Clinical Trial. JAMA 328, 151-161, doi:10.1001/jama.2022.9805 (2022).