W86. ASSESSMENT OF PHARMACOGENETIC ANALYSIS TECHNOLOGIES APPLIED IN THE G-PAT STUDY: MULTI-MODAL CHARACTERIZATION OF KEY PHARMACOGENES IN THE CONTEXT OF PSYCHOSIS

• Published in: European neuropsychopharmacology 2023-10, Vol.75, p.S150-S151
• Main Authors: Cotic, Marius, Khani, Noushin Saadullah, Varney, Lauren, Abidoph, Rosemary, Manavale, Craig, Panconesi, Daniele, Richards-Belle, Alvin, Ramesh, Soumita, Kuchenbaecker, Karoline, Kristiansen, Mark, Bramon, Elvira
• Format: Article
• Language: English
• ISSN: 0924-977X; 1873-7862
• DOI: 10.1016/j.euroneuro.2023.08.273

As the breadth of pharmacogenetic (PGx) technologies used for research and clinical testing has been developing, so has the question of which methodology is most appropriate for every application. Based on the aim of the Using Genetics to Personalise Antipsychotic Treatment (G-PAT) study to accurately characterize the pharmacogenetic profile of key genes involved in antipsychotic treatment response, we have performed different genotyping assays to try and capture the entire genetic diversity of our study cohort. By furthering our understanding of the strengths and the limitations of each methodology, we aim to develop a holistic and cost-effective approach to the characterization of key pharmacogenes. Prospective study comparing treatment as usual with antipsychotic medications versus treatment with antipsychotics informed by pharmacogenetics testing. Participants: people with psychosis taking an antipsychotic recruited across England (UK) by self-referral or from 11 National Health Service centres. Patients donated a saliva or blood sample for DNA extraction and biobanking at University College London. We tested a multi-gene panel selected from evidence-based pharmacogenetic clinical guidelines for antipsychotics or antidepressants (CYP2D6, CYP1A2, CYP3A4, CYP2C19, CYP2B6, other). We conducted further testing on whole genome genotyping arrays. We measured turnaround times, variant coverage, and costs of different technologies. Ongoing research is aimed at validating a subset of rare variants using long-read sequencing and novel variant-calling software. We tested ∼960 variants from 220 adults with psychosis. Average age 42 years, SD 15, range 18-91 years. Sex: 48% female. Ancestry: 12% identified as Asian/Asian-British, 11% as Black/Black-British/Caribbean/African, 68% White/European and 9% reported multiple/other ancestries. Genotyping was performed with three different technologies (a PCR panel, a multi-gene panel and genome-wide genotyping array with enhanced PGx). We focused on the differences in coverage between the methods regarding CYP2D6, as it is known to be difficult to characterize. The star allele coverage between the two methods varied considerably, with the panel covering 23 alleles, whereas the genotyping arrays covered 135 alleles. This coverage difference was also observed for other relevant pharmacogenes, including CYP2C19. Independent testing with PCR methods provided very consistent results with the multi-gene panel approach including