A systematic review and quality assessment of the clinical validity and utility of guideline recommendations for the use of germline genetic biomarkers to individualize drug selection or dosing

Abstract

Pharmacogenetics claim to offer the potential to improve health outcomes by identifying individuals who are at greater risk of harm or treatment failure from certain drugs. Routine adoption of pharmacogenetic tests requires evidence of their clinical validity and utility. This study aims to systematically assess the quality of published guidelines on the use of inherited (germline) pharmacogenetic biomarkers to guide drug selection or dose adjustment. A protocol of the study was registered on PROSPERO (registration CRD42016046258). A systematic literature search (EMBASE, MEDLINE, the Pharmacogenomics Knowledge Base, and the International Guideline Library) was performed. Inclusion criteria were: The article must be a guideline, which contains a recommendation for dose or therapy modification for a defined drug based on at least one genetic variant in the patient’s germline genome. The process of the systematic review was supported by self-developed Python scripts for data extraction, synthesizing, filtering and exporting, accessing the web application programming interfaces (APIs) from the literature databases and using RefWorks literature management as a database programmatically. It was possible to extract and individually export data from the Pharmacogenomics Knowledge Base (PharmGKB) to ensure a faster process for the systematic evaluation of this data.

The level of evidence and the strength of recommendation for each genetic marker-drug pair was re-evaluated by applying predefined scales published by the “Evaluation of Genomic Applications in Practice and Prevention” (EGAPP) working group. The focus was on two key questions, namely the clinical validity and the clinical utility. A total of 32 guideline publications were identified, which included recommendations on genotype-stratified prescription for 84 drugs. The guidelines were systematically reviewed on their evidence that particular genetic markers were associated with pharmacokinetics or pharmacodynamics of the drug and how factors such as ethnicity affect these associations. However, for only a few drugs (e.g., abacavir-HLA-B, allopurinol-HLA-B, warfarin-VKORC1/CYP2C9, tacrolimus-CYP3A5) the guidelines identified sufficient evidence supporting benefits outweigh negative effects when genotype-stratified prescribing was applied. For several drugs, guidelines recommended genotype-based dose adjustments despite lack of evidence that pharmacogenetic dose adjustment would improve outcome (e.g. irinotecan dose reductions in UGT1A1 *28/*28 carriers would increase tumor response and survival). Most guideline recommendations are based solely on associations between genetic variants and either blood concentrations of the drug or health outcomes (e.g. adverse drug reactions) of the treatment. Prospective studies investigating the impact of genetic testing on the clinical outcome of patients and data on the number needed to genotype (NNG) or the cost-effectiveness are scarce. Characteristics of clinical utility, however, play an important role in the translation into clinical care.