PII-042 - ANALYSIS OF A CLINICAL PROGRAM TO RETURN CLINICALLY ACTIONABLE PHARMACOGENETIC INFORMATION FROM GERMLINE DNA SEQUENCING DATA GENERATED IN MOLECULAR TUMOR BOARDS: WORKFLOW AND ESTIMATED COSTS

H. Koo^1,2, T. Smith¹, J. Callaghan¹, W. Osei^3,2, E. Tillman¹, M. Tran⁴, C. Fausel⁴, B. Schneider¹, T. Shugg⁵, T. Skaar⁵; ¹Indiana University School of Medicine, Indianapolis, IN, USA, ²Purdue University, West Lafayette, IN, USA, ³Department of Pharmacy Practice, School of Pharmacy, Purdue University, Indianapolis, IN, USA, ⁴Indiana University Health, Indianapolis, IN, USA, ⁵Division of Clinical Pharmacology, Department of Medicine, Indiana University School of Medicine, Indianapolis, IN, US.

Background: Clinically actionable pharmacogenetic (PGx) variant are prevalent in general population with particular implications for oncology patients who often require complex polypharmacy regimens for antineoplastic and supportive care. Germline whole exome sequencing (WES) is increasingly being utilized by molecular tumor boards (MTB). WES includes sequencing of pharmacogenes, but those results are rarely included in clinical reports from sequencing laboratories, in part due to the complexity of accurately extracting PGx genotypes. Recently, we developed a method to accurately extract PGx genotypes from WES data. Thus, our objective was to conduct a comprehensive analysis of the overall time commitment and cost of implementing a clinical program to extract PGx genotypes from WES and to return PGx results for discussion at our institutional MTB.

Methods: Clinical WES was performed as part of standard care for patients with cancer and binary alignment map (BAM) files were ingested into an institutional database for clinical analysis. The workflow began with identifying patients who have sequencing results to be discussed at the weekly MTB. PGx results were extracted from the WES data using Aldy v4.6. Since the informatics pipeline has not been clinically validated, confirmatory PGx testing was recommended for patients with potentially actionable PGx genotypes relevant to the patient’s treatment plans. Each step of the process was timed by an investigator not involved in the analysis of the sequencing data.

Results: The entire hands-on processing time was ~3 hours and 15 minutes and there were 20.7 (± 1.5, 95% CI) patients analyzed per week. Computing time to get the PGx information was approximately one hour. After downloading the BAM files, running Aldy took ~45 minutes. Assessment of the actionability of the PGx variants took ~30 minutes. Participation in the weekly molecular tumor board typically took about one hour. Using the average cost of a pharmacist to do the analysis and reporting, the estimated overall cost per patient for extracting and reporting the PGx information, not including the confirmatory testing, was $14.7 (± 1.1, 95% CI).

Conclusion: These results demonstrate the relatively low cost of a clinical program to extract PGx genotypes from WES data and return them for discussion at a MTB.