Point of care genotyping assays, and algorithm for warfarin dosing

Lemieux, Bertrand

Abstract

On August 16, 2007, the FDA updated the label for warfarin to include information in the """"""""precautions"""""""" section to remind physicians that people with variations in the CYP2C9 and VKORC1genes may require a lower initial dose of the drug (see www.fda.gov/cder/drug/infopage/warfarin/default.htm). Unfortunately, the FDA did not provide specific guidance to physicians on how to use this genotyping information. Soon after the announcement for an altered label for warfarin, the FDA cleared a genotyping test for the most common mutations influencing patients' responses to warfarin (i.e., Nanosphere's Verigene warfarin test FDA 510(k) # K070804). Unfortunately, this new diagnostic device did not include instructions on how to use the genetic information to prescribe warfarin. In addition, the cleared warfarin test requires large quantities of highly purified DNA (2 5g), and uses costly instrumentation (such that high test volumes will be required to justify the reagent lease). As FDA has not yet cleared a nucleic acid extraction method for use with the Verigene warfarin test, only large clinics with laboratories capable of complying with the Laboratory Accreditation Program Molecular Pathology Checklist, developed by the College of American Pathologists, will adopt the Verigene warfarin test. We believe the lack of information of how to use the genotyping data, and the complexity of the DNA extraction required to perform the genetic test will limit its utility. We propose to develop a low-cost, instrument free genotyping system that does not require large quantities of extensively purified DNA, and that includes a web-based computational tool that integrates genetic, and physical factors to offer an individualized predictive model for warfarin dose. The genetic test will use BioHelix's proprietary helicase dependent amplification (HDA), as well as a disposable lateral flow device designed to prevent laboratory contamination with amplicons. The HDA reagents will be formulated as a Ready-to-go"""""""" IsoAmp dry reagent formulation with extended stability at room temperature. We have found that crude buccal swabs can be used to perform HDA genotyping tests; therefore no DNA extraction will be required. As lateral flow tests widely used in """"""""CLIA waived"""""""" and """"""""moderate complexity"""""""" assays, we believe coagulation will prefer this detection format to using complex instrumentation. In Phase I, we will develop 3 genotyping assays for the CYP2C9*2 (a.k.a. 3,608 C>T or R149C), the CYP2C9*3 (a.k.a. 42,614 A>C or I359L), and the 1,639 G>AVKORC1 loci using ARMS technology. We will also implement design control for our product development process, and secure ISO 13485 certification for BioHelix. Trimgen (Sparks, MD) will manufacture BuccalQuick"""""""" DNA extraction kits under GMP. BioHelix will manufacture bulk enzymes required for HDA. GE Healthcare will manufacture the Ready-to-go"""""""" IsoAmp warfarin dry reagents under GMP. The amplicon containment, lateral flow device will be manufactured under GMP by Ximedica. Finally, we will validate the 3 warfarin assays at PGX laboratories using 50 samples and compare the performance of the Ready-to-go"""""""" IsoAmp warfarin assays to DNA sequencing and warfarin genotyping assays validated at PGX laboratories. In Phase II, we perform a retrospective, multi-site clinical study aimed at validating the assays using the Verigene warfarin test as a predicate device, and PGXL technology's algorithm with ~500 patients on consistent dosing of Coumadin for 3 months. This study will include a warfarin dosing algorithm. We believe we can offer genetic typing test for warfarin with a """"""""moderate degree of complexity"""""""" with a link to a web-based computational tool for individualized predictive modeling of warfarin dosing for between $15 and $20/locus. Narrative In patients with atrial fibrillation (AF), the uneven and arrhythmic pumping of the heart's two upper chambers results in pools of blood in the atrium that can form clots. Patients can experience strokes when these clots break loose, enter the bloodstream, and travel to the brain to plug an artery. Treatment for AF usually consists of taking an anticoagulant like aspirin (for low risk cases), or warfarin (for the more severe cases). Women with atrial fibrillation are more likely to form dangerous blood clots than men (Fang et al. 2005), and thus are likely to benefit from improved warfarin dosing. Warfarin exerts its anticoagulant effect by inhibiting the vitamin K epoxide reductase (VKOR), which depletes the pool of reduced vitamin K available, and prohibits the activation of the vitamin K- dependent clotting factors and, ultimately, thrombin formation. Genetic polymorphism in the VKORC1, a component of VKOR influence the degree of patient sensitivity to warfarin. Patients typically have either a low-dose haplotype group (A), a high-dose haplotype group (B), or are heterozygous. The hepatic Cytochrome P450 2C9 (CYP2C9) enzyme metabolizes warfarin to clear it from the blood stream, therefore polymorphisms in the CYP2C9 gene that influence enzyme activity will influence the steady state level of warfarin in the blood stream. Factors such as age, gender, height, and weight, also impact the range of possible warfarin dose requirements. Not surprisingly, adverse drug events (ADEs) are common during warfarin therapy. It is estimated that over 50% of patients that could benefit from warfarin are not getting the drug because their physician is weary of the potential for ADE. A recently published study found that the incorporation of genetic testing into warfarin protocols could help patients to avoid 85,000 serious bleeding events and 17,000 strokes every year. Because of the risk of ADEs, warfarin therapy is usually monitored by determining the prothrombin time (PT) at regular intervals using the international normalized ration (INR) as a means of standardizing the PT measurement. However, the antithrombotic effects of warfarin, along with the accompanying interpretable changes in INR, do not become apparent until the pools of Factor II (Prothrombin) and Factor X are depleted over the course of 2-4 days. This results in fluctuations in INR that can result in serious ADEs if the initial dose was miscalculated. The development of an easy-to-use computational tool that integrates all of the relevant genetic, and physical factors into a comprehensive, real-time, individualized predictive model for warfarin dose could translate the results of pharmacogenetic testing into an actionable clinical application. However, FDA regulation requires that such an algorithm be validated with a genetic test in order to be listed in the package insert accompanying the genetic test system. An ideal genetic testing system for typing warfarin dosing polymorphisms should be easy to operate, and should not require extensive instrumentation.

Public Health Relevance

In this proposal, we outline a plan to develop a low-cost instrument free genotyping system that does not require large quantities of extensively purified DNA and that could be performed in the doctor's office. This novel diagnostic system will be validated with an easy-to-use, web-based computational tool for warfarin dosage that integrates all genetic and physical factors to project warfarin levels in the patient's blood and predict changes in INR when a target dose of warfarin is administered to the patient. ? ? ?