Pharmacogenetic Study of Warfarin Dose-Response: a Prospective Trial

Prospective Study Comparing Between the Commonly-Used and Pharmacogenetically-Guided Warfarin Administration Protocols

We propose to develop a personalized pharmacogenetic approach including the major genetic markers of warfarin (coumadin) dosing and patients' age and weight. The known genetic determinants include several functional and common polymorphisms in CYP2C9 and VKORC1 genes, which explain the low-end of warfarin dosing range and mostly occur in patients of Caucasian and Chinese origins. We identified a new VKORC1 polymorphism that is specifically indicative of the high dose requirements and is dominant over the dose-reducing effect of the known CYP2C9 and VKORC1 markers. This marker is significantly over-represented in Jews of Ethiopian origin, but is also common in Ashkenazis, it is also linked to the VKORC1 genetic markers characteristic of the Afro-American population (published in Blood 2007, 109:2477-80). This information prompts the development of a more inclusive and universal diagnostic approach to the individualized warfarin therapy.

The present study aims at evaluation of our novel pharmacogenetic model for predicting warfarin (coumadin) dose response on the basis of patient's genetic markers of warfarin sensitivity and resistance, and other patient specific factors. To this end, we proposes to re-evaluate our previously developed pharmacogenetic model in stabilized warfarin treated patients (N=200) and then to implement it in a prospective study of patients new on warfarin as compared to the "traditionally" treated patients (N=500).

Objective:

To date, most warfarin-related pharmacogenetic studies are cross-sectional or retrospective association studies. There have been no prospective studies designed to evaluate the impact of genotype-guided compared to the "traditional" trial and error approach currently used in clinical practice. The present study is designed to prospectively evaluate the impact of a combined pharmacogenetic warfarin dosing prediction model on the safety and efficacy of warfarin treatment in a community based setting.

Specific Aims:

1. To develop and validate a genotype-based prediction model for warfarin dosing optimization.
2. To prospectively compare anticoagulation quality and clinical outcomes in genotype-guided warfarin dosed patients versus "traditional" dosed patients.

Study Design:

The study will be performed at the Molecular Lab at the Institute of Human Genetics, Sheba Medical Center in collaboration with the Maccabi Health Care facilities, Israel.

Study I: A cross sectional study to develop and validate a genotype-based prediction warfarin dosing model, including two hundred (200) unselected patients at steady-state warfarin anticoagulation defined by stable therapeutic INR's (±10%) achieved by stable warfarin doses (±10%) over 3 months identified from the Maccabi database. Informed consent will be obtained by the treating physicians who will also document demographic and clinical data including age, gender, body weight, ethnic background, indication for warfarin treatment, additional medical problems and concomitant medications on appropriate forms (included in the CRF). A 5cc fresh whole blood sample will be drawn from each patient at the next routine INR monitoring venepuncture and after anonymization, will be delivered to the Genetics laboratory for DNA extraction and genotyping.

Genetic analyses include multiplex analysis of at least 10 polymorphisms associated with warfarin dose response that will be performed using Sequenom Mass Spectrograph technology (CYP2C9*1/*2/*3, VKORC1*1/*2/*3/*4, VKORC1 Asp36Tyr, CALU Arg4Glu and EPHX1 Tyr113His).

Data analysis includes multiple regression analysis that will be employed on the data to create an accurate prediction model for individual therapeutic warfarin steady-state maintenance dose (dependent variable) as a function of the demographic variables and genotype groups (independent variables). The accuracy of the prediction model will be cross-validated with our previous data on warfarin-treated patients at steady state.

Study II: A prospective controlled trial to compare laboratory and clinical anticoagulation outcomes in genotype-guided warfarin dosed patients versus "traditionally" dosed patients, including five hundred (500) patients commencing warfarin therapy for standard indications and confirming informed consent will be recruited by Maccabi physicians (who will receive a training session prior to the study) at the time of their 1st Maccabi laboratory INR monitoring visit. Demographic and clinical data will be recorded as above. Patients will be randomly assigned into one of two dosing groups: (1) genotype-guided warfarin dosing group (N=250); (2) traditional dosing group (N=250).A single 5cc fresh whole blood sample will be drawn in the same venepuncture used for the INR determination, for DNA extraction and anonymized genotyping.

Genetic analyses include an immediate analysis of 4 most significant markers of warfarin dose response (CYP2C9*1/*2/*3, VKORC1*2 haplotype and VKORC1 Asp36Tyr) that will be performed using Real Time PCR or PRONTO methodology.

Controlled trial: For patients in the genotype-dosing group Maccabi physicians will receive dosing recommendations based on the prediction model validated in study I, within 48 hours. Genotype-based dosing recommendations will be subject to changes by the treating physicians based on their own clinical judgment. All dose-recommendation changes will be recorded and analyzed. Patients assigned to the traditional dosing group will be dosed by their Maccabi physician according to accepted clinical practice, (using repeated INR measurements and warfarin dose changes as required by physician judgment). For standardization purposes, warfarin dose adjustments will be carried out according to a predefined algorithm. All patients will be followed up until they achieve stable anticoagulation, defined as 3 scheduled visits to the anticoagulation clinic follow up program within at least 12 weeks. Each follow up visit will include a short interview, INR measurement and warfarin dosing schedule including timing of the next visit. At each visit the following data will be recorded: major active medical problems, INR value and next warfarin daily dose, use of and changes in concurrent medications, dietary intake changes or other possible explanations for inadequate (excessive or under) anticoagulation, changes from scheduled warfarin daily doses and occurrence of minor or major bleeding episodes.

Data analysis: Efficiency and safety of warfarin anticoagulation will be compared between the two study groups applying appropriate statistical methods to the following outcome variables:

1. Mean number of visits from recruitment to stable anticoagulation.
2. Mean number of INR measurements required from recruitment to stable anticoagulation.
3. Proportion of visits with excessive INR values (above the upper limit of the therapeutic INR value).
4. Proportion of visits with INR values below the lower limit of the therapeutic INR value).
5. Number of required changes in warfarin daily doses during the study period.
6. Number of major and minor bleeding episodes.

Patient Benefit:

Participating patients will benefit from early delivery to their treating physician of information relevant to their genetic propensity for requiring high, intermediate or low warfarin doses, possible shortening of their time of treatment titration and possible decreases in their likelihood to experience dose-related complications. There is no envisaged harm to recruited patients.

Organizational Benefit:

Positive results will be implemented as proof of concept, in a project for developing a user-friendly point-of-care pharmacogenetic diagnostic technology.

Study Duration: 3 years

Inclusion Criteria:

• patients starting warfarin therapy

Exclusion Criteria:

• pregnant women