Abstract
Over‐ and undercoagulation with warfarin are associated with hemorrhagic and thromboembolic events, respectively. Genetic and clinical factors affect warfarin response, and the causes of this variability remain unclear. We present descriptive statistics and test for predictors of poor anticoagulation control. The Quebec Warfarin Cohort (QWC) comprises 1059 new warfarin users, with prospective follow‐up using telephone questionnaires every 3 months for 1 year, and using healthcare administrative databases (RAMQ and Med‐Echo) for 5 years prior to cohort entry and up to 10 years following active patient participation. Genetic material was collected, and genotyping of CYP2C9 and VKORC1 genes was conducted. Measured outcomes included the percentage of time patients spent within therapeutic range, anticoagulation control, warfarin dose, bleeding, and thromboembolic events. We report baseline characteristics and outcomes after 1 year of follow‐up. Poor anticoagulation control was defined as time in therapeutic range <60% in the 3‐ to 12‐month interval. Participants had a mean age of 71 years, and 62% were men. The most common indication for warfarin was atrial fibrillation (87%). Mean time in therapeutic range was 56% (±25%) in the 3 months following warfarin initiation, and 70% (±21%) in the 3‐ to 12‐month interval. During follow‐up, the rate of stroke or systemic embolism was 1.8 events per 100 person‐years; for major bleeding events, 3.3 events per 100 person‐years. Independent predictors of poor anticoagulation control were chronic kidney disease, heart failure, dyslipidemia, and age. The QWC represents a good research cohort to investigate clinical and genetic factors in a warfarin‐anticoagulated population.
Keywords: Dose, Thromboembolic and Hemorrhagic Events, TTR, Warfarin
1. INTRODUCTION
Despite the arrival of direct non–vitamin K antagonist oral anticoagulants, warfarin remains a widely prescribed anticoagulant worldwide for the treatment and prevention of thromboembolic diseases.1 Warfarin has a narrow therapeutic index which requires regular monitoring of the anticoagulation response with international normalized ratio (INR) measurements. The clinical importance of monitoring is highlighted by the fact that supra‐ and subtherapeutic INR are associated with increased risks of hemorrhagic and thromboembolic events, respectively.2, 3
Warfarin therapy is further complicated by significant interindividual variability in the dose requirements to achieve optimal anticoagulation.4 Several genetic and nongenetic factors have been associated with variability in the warfarin maintenance dose.4, 5, 6 Indeed, 10% to 20% of the variability in warfarin dose is attributed to the combined effects of height, weight, and age, and concomitant medication is responsible for another 5% to 10%.4 A further 30% to 40% of warfarin dose variability is explained by the genetic variants at VKORC1, the gene encoding the warfarin target (vitamin K epoxide reductase complex subunit 1) and CYP2C9, the gene encoding cytochrome P450 2C9 (CYP2C9), the main enzyme involved in the metabolism of the more potent S‐warfarin enantiomer.4, 7 Yet, a large fraction of the remaining interpatient variation in response to warfarin remains to be elucidated.
The Quebec Warfarin Cohort (QWC) was designed to comprehensively and systematically identify the clinical, lifestyle, and genetic predictors of safety and efficacy of warfarin. Here, we present the descriptive statistics of the study during the first year of follow‐up and report on the predictors of poor anticoagulation control.
2. METHODS
2.1. General objective and design
The QWC is a multicenter, community‐based, prospective cohort study of new warfarin users. Here, we present in details the descriptive statistics of the study during the first year of follow‐up. The primary objective of the present report was to assess the predictors of poor anticoagulation control in the 3‐ to 12‐month follow‐up period defined as a percentage of time within the therapeutic INR range (TTR) <60%. The study was performed under the terms of the Declaration of Helsinki and the study protocol was approved by the local institutional review board or ethics committees. All enrolled patients have given an informed consent.
2.2. Patient recruitment and follow‐up
A total of 1059 patients were recruited consecutively between May 2010 and July 2013 at 19 anticoagulation clinics in the Quebec province of Canada (see Supporting Information, Table S1, in the online version of this article), among which the Montreal Heart Institute (MHI) was the leading and coordinating center. Warfarin new users were identified on daily basis. The patients were recruited by a trained nurse at the time of the initial visit to the warfarin clinic, close to 3 days after warfarin initiation. Eligible patients had an indication for taking warfarin with an expected duration of treatment of ≥12 months including paroxysmal atrial fibrillation (AF), atrial flutter, aortic valve replacement, mitral valve replacement, and mitral stenosis (Table 1).
Table 1.
Inclusion and exclusion criteria for QWC
| Inclusion criteria |
|---|
| Age ≥ 18 y |
| Indication for taking warfarin with an expected duration of treatment of ≥12 mo including acute and paroxysmal AF, atrial flutter, aortic valve replacement, mitral valve replacement, and mitral stenosis. |
| Insurance coverage with the RAMQ or a private insurance plan for ≥2 y before initiation of warfarin |
| Acceptance and ability to sign informed consent |
| Able to be followed in outpatient anticoagulation clinic |
| Exclusion criteria |
| On warfarin therapy at the time of recruitment |
| DVT or PE as primary indication of warfarin therapy |
| Contraindication for warfarin therapy within the past 3 mo including ≥1 major bleeding event (eg, severe GI bleeding or hemorrhagic stroke) |
| Coagulation factor deficiency |
| Cirrhosis and/or chronic hepatitis |
| End‐stage renal disease |
| Clinically recognized mental illness, cognitive disorders (eg, dementia), or any other factors likely to limit ability of patient to follow the study, to provide informed consent, or to adhere to the drug (eg, alcohol abuse, plans to move within the next 12 mo) |
Abbreviations: AF, atrial fibrillation; DVT, deep‐vein thrombosis; GI, gastrointestinal; PE, pulmonary embolism; QWC, Quebec Warfarin Cohort; RAMQ, Régie de l'assurance maladie du Québec.
Participants were followed for 1 year with 5 structured telephone questionnaires: 1 at baseline, 3 follow‐up questionnaires at 3‐month intervals, and 1 questionnaire at the end of the study (Figure 1). Data collection included primary indication of warfarin, treatment start date, target INR, prescribed daily dose, lifestyle habits (including smoking, alcohol intake, diet, and physical activity), predicting stroke and thromboembolism in AF using CHA2DS2‐VASc score, risk of major bleeding using HAS‐BLED score, comorbidities, and over‐the‐counter medications.8, 9 The study's data‐collection plan, as well as a detailed description of variables for which data was collected at recruitment and each follow‐up time point, is provided in the Supporting Information, Tables S2, S3, and S4, in the online version of this article. The collected data from questionnaires and hospital charts were recorded in a structured and standardized case‐report form and transferred into a PostgreSQL database (PostgreSQL Global Development Group) with an Microsoft Access electronic interface (Microsoft, Redmond, WA). The data entries were manually validated by a different team member.
Figure 1.
Flow diagram for patients recruited in the Quebec Warfarin Cohort (QWC) study
The healthcare administrative databases of the Régie de l'assurance maladie du Québec (RAMQ)10 and MED‐ECHO11 were used to capture medication use and hemorrhagic and thromboembolic events for 5 years prior to the recruitment date and extending up to 10 years following active patient participation. Quebec residents are covered by the RAMQ universal public health insurance plan for procedural and diagnostic medical services,12 and the database has physician claims data documenting the date and setting of care (inpatient or ambulatory services), the procedure provided, and the diagnoses.12 We captured filled drug claims by RAMQ drug plan of eligible patients. The MED‐ECHO database is a hospitalization registry that provides dates and length of stays, diagnoses, type of hospital, medical procedures performed, and in‐hospital deaths.12, 13 INR monitoring test results were obtained on a regular basis during the follow‐up period from the computerized hospital databases and from clinical chart reviews. All INRs were measured locally. For each warfarin indication, we considered specific INR targets (eg, 2.0–3.0 for AF and 2.5–3.5 for valve replacement).
2.3. Genotyping
DNA extraction and genotyping were performed at the Beaulieu‐Saucier Pharmacogenomic Centre of the MHI. Genomic DNA was extracted and purified from whole blood using Qiagen reagents on an Autopure LS instrument (Qiagen, Hilden, Germany) and quantified using the QuantiFluor dsDNA System (Promega, Madison, WI). Genotyping was performed using the iPLEX ADME CYP2C9/VKORC1 v1.0 Panel (Agena Inc., San Diego, CA) including 36 variants in CYP2C9 and 9 variants in VKORC1.
2.4. Study outcomes
Primary clinical outcomes included the percentage of time patients spent within the therapeutic INR range (TTR), anticoagulation control defined as TTR ≥60% or < 60%, and the warfarin dose requirements.14, 15 TTR is a reliable measure of quality of anticoagulation control with warfarin. It is a strong predictor of adverse events in warfarin‐treated patients such that lower TTR (<60%) is associated with an increased risk of major bleeding and thromboembolism.14, 15, 16, 17, 18 TTR was calculated using the Rosendaal method as previously described19 based on linear interpolation to infer the proportion of days a patient spent within his or her respective therapeutic range. The median number of days between INR measures was 7 (interquartile range [IQR], 3–21 days). Time periods of >70 days between INR measurements were excluded due to imprecision of the interpolation over long periods.19, 20 Additional study outcomes included the percentage of time patients spent in supra‐ or subtherapeutic INR range, poor anticoagulation (TTR <60%), ever‐occurrence of an INR ≥5, and clinical outcomes including the occurrence of hemorrhagic and thromboembolic events.
Exploratory clinical outcomes included minor and major bleedings and thromboembolic events during the 1 year of follow‐up. Minor bleeding events included self‐reported bruising, nose bleeds, gum bleeding, hematuria, and rectal bleeding requiring no further action. Clinical outcomes including major bleedings, stroke (ischemic/hemorrhagic), and other thromboembolic events were identified during the follow‐up using ICD‐10 codes from the MED‐ECHO administrative database (see Supporting Information, Table S4, in the online version of this article) lists. We used ICD‐9 and/or ICD‐10 codes for the detection of outcomes in the 5‐year prior cohort. These codes performed relatively well in previous validation studies. The positive predictive value ranged from 85% to 95%.21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31
2.5. Statistical considerations
Descriptive data are presented as mean ±SD or median with IQR for continuous variables, or as observation numbers with percentages for categorical variables. CYP2C9 metabolizer status and VKORC1 activity status were determined according to the Clinical Pharmacogenetics Implementation Consortium guideline (CPIC; https://www.pharmgkb.org), using the common CYP2C9 and VKORC1 gene variants (CYP2C9 *2 [rs1799853] and *3 [rs1057910] and VKORC1 *2 [rs9923231], *3 [rs7294], and *4 [rs17708472]). CYP2C9 metabolizer status was defined as extensive metabolizer (*1/*1), intermediate metabolizer (*1/*2 or *1/*3), and poor metabolizer (*2/*2, *2/*3, or *3/*3). VKORC1 activity status was defined as high activity (*1/*3, *1/*4, *3/*3,*3/*4, or *4/*4), intermediate activity/extensive activity (*1/*1, *1/*2, *2/*3, or *2/*4), and poor activity (*2/*2).
The CYP2C9 and VKORC1 phenotypes were included in the final model as 3‐level continuous variables and were coded as follows: poor metabolizer or poor activity, 0; intermediate metabolizer or intermediate activity/extensive activity, 1; and extensive metabolizer or high activity, 2. Stratified analyses for the arbitrary pre‐ and post‐stabilization periods were defined as time from baseline to the 3‐month time point and from 3 months to the end of the follow‐up (12‐month time point), respectively. To explore the factors associated with low anticoagulation control (TTR <60%) from 3 to 12 months after warfarin initiation, we used a univariate logistic regression model. We constructed a multivariable model using stepwise regression including variables with the provided univariate P values <0.10 (see Supporting Information, Table S5, in the online version of this article), with the inclusion of age and body mass index treated as continuous variables, and kept the final variables with P < 0.05. We computed the multivariable logistic regression model's R 2 statistic according to the Nagelkerke equation. All analyses were performed using SAS software, version 9.4 (SAS Institute, Inc., Cary, NC).
Based on the primary outcome of poor anticoagulation control defined as TTR <60% during the 3‐ to 12‐month period, with 243 (29%) poorly controlled patients out of 842, assuming an α of 0.05, we calculated 80% power to detect a significant association for a dichotomous predictive variable of 10% prevalence with effect size odds ratio (OR) ≥1.9, and for a predictive variable of 50% prevalence with effect size OR ≥ 1.5.
3. RESULTS
Of the 1059 patients (Figure 1), 62% were male and the mean age was 70.7 ±11.8 years. The median CHA2DS2‐VASc score was 4.0 and the median HAS‐BLED score was 2 (Table 2). There were 25.8% of participants who reported engaging in physical activity of moderate or greater intensity, 28.8% reported not drinking alcohol, and 78.8% reported regularly eating vitamin K–rich foods. The major clinical indication for warfarin therapy was AF (86.5%; Table 2).
Table 2.
Baseline characteristics of study participants of the QWC
| Full QWC, N = 1059 | Subset With Concomitant Medication Data, n = 807 | Anticoagulation Control From 3 to 12 Months TTR < 60% of the Subset, n = 243 | Anticoagulation Control From 3 to 12 Months TTR ≥ 60% of the Subset, n = 599 | |
|---|---|---|---|---|
| Sociodemographicsa | ||||
| Age, y | 70.7 ± 11.8 | 73.7 ± 10.2 | 69.1 (12.7) | 71.4 (10.9) |
| Male sex | 656 (62) | 475 (59) | 144 (59.3) | 367 (61.3) |
| BMI, kg/m2 | 28.6 ± 6.1 | 28.5 ± 6.2 | 28.7 (6.3) | 28.8 (6.1) |
| Education | ||||
| University or college | 367 (34.6) | 254 (31.5) | 89 (36.6) | 199 (33.2) |
| High school | 282 (26.6) | 199 (24.7) | 71 (29.2) | 166 (27.7) |
| No degree | 410 (38.7) | 353 (43.7) | 83 (34.2) | 233 (38.9) |
| Lifestylea | ||||
| Level of physical activity | ||||
| Inactive | 292 (27.6) | 261 (32.4) | 69 (28.4) | 159 (26.5) |
| Light | 494 (46.6) | 381 (47.2) | 107 (44.0) | 294 (49.1) |
| Moderate | 197 (18.6) | 126 (15.6) | 50 (20.6) | 103 (17.2) |
| Hard | 43 (4.1) | 25 (3.1) | 7 (2.9) | 28 (4.7) |
| Very hard | 33 (3.1) | 14 (1.7) | 10 (4.1) | 15 (2.5) |
| Smoking | ||||
| Nonsmoker or former smoker | 976 (92.2) | 747 (92.6) | 220 (90.5) | 557 (93.0) |
| Current smoker | 83 (7.8) | 60 (7.4) | 23 (9.5) | 42 (7.0) |
| Alcohol intake | ||||
| Heavy drinker | 110 (10.4) | 70 (8.7) | 29 (11.9) | 55 (9.2) |
| Regular drinker | 269 (25.4) | 213 (26.4) | 54 (22.2) | 164 (27.4) |
| Occasional drinker | 375 (35.4) | 272 (33.7) | 93 (38.3) | 217 (36.2) |
| Nondrinker | 305 (28.8) | 252 (31.2) | 67 (27.6) | 163 (27.2) |
| Vitamin K intake | ||||
| Regularly | 835 (78.8) | 630 (78.1) | 179 (73.7) | 488 (81.5) |
| None or occasionally | 224 (21.2) | 177 (21.9) | 64 (26.3) | 111 (18.53) |
| Primary indication of warfarin therapya | ||||
| Paroxysmal AF | 450 (42.5) | 365 (45.2) | 93 (38.3) | 241 (40.2) |
| Chronic AF | 357 (33.7) | 289 (35.8) | 77 (31.7) | 215 (35.9) |
| Atrial flutter | 109 (10.3) | 88 (10.9) | 24 (9.9) | 65 (10.9) |
| Mitral valve replacement | 53 (5.0) | 33 (4.1) | 13 (5.4) | 34 (5.7) |
| Aortic valve replacement | 104 (9.8) | 42 (5.2) | 36 (14.8) | 55 (9.2) |
| Mitral stenosis or other | 19 (1.8) | 12 (1.5) | 7 (2.9) | 10 (1.7) |
| Comorbidities (in 5‐year prior cohort entry)b | 1058c | |||
| HTN | 769 (72.7) | 602 (74.6) | 186 (76.5) | 425 (71.0) |
| CAD | 660 (62.3) | 504 (62.5) | 165 (67.9) | 362 (60.4) |
| Dyslipidemia | 504 (47.6) | 392 (48.6) | 137 (56.4) | 260 (43.4) |
| DM | 386 (36.5) | 312 (38.7) | 98 (40.3) | 214 (35.7) |
| COPD/asthma | 328 (31.0) | 267 (33.1) | 84 (34.6) | 181 (30.2) |
| CHF | 294 (27.8) | 218 (27.0) | 93 (38.3) | 138 (23.0) |
| Chronic renal disease | 193 (18.2) | 162 (20.1) | 70 (28.8) | 83 (13.8) |
| PVD | 206 (19.5) | 150 (18.6) | 57 (23.5) | 105 (17.5) |
| MI | 163 (15.4) | 129 (16.0) | 43 (17.7) | 82 (13.7) |
| Cerebrovascular disease including TIA | 132 (12.5) | 98 (12.1) | 31 (12.8) | 73 (12.2) |
| Procedures during the last 5 yb | 1058c | |||
| Cardiac catheterization/coronary angiography | 193 (18.2) | 193 (23.9) | 68 (28.1) | 97 (16.2) |
| PCI ± stent | 85 (8.0) | 85 (10.5) | 19 (7.9) | 46 (7.7) |
| CABG | 53 (5.0) | 53 (6.6) | 16 (6.6) | 30 (5.0) |
| Healthcare services in 1‐y prior cohort entryb | 1058c | |||
| Specialty visits | 6.4 ± 7.2 | 6.5 ± 7.8 | 7.7 ± 11 | 5.9 ± 5.1 |
| Family physician visits | 1.3 ± 2.4 | 1.4 ± 2.6 | 1.5 ± 2.7 | 1.2 ± 2.3 |
| Emergency visits | 1.8 ± 2.3 | 1.8 ± 2.4 | 2.2 ± 2.9 | 1.6 ± 2.1 |
| Healthcare services in 3‐y prior cohort entryb | 1058c | |||
| All‐cause hospital admission, n (%) | 178 (67.9) | 541 (67.0) | 182 (72.5) | 389 (64.9) |
| All‐cause hospital admission, mean ± SD | 1.4 ± 1.5 | 1.4 ± 1.5 | 1.8 ± 1.7 | 1.3 ± 1.5 |
| Length of hospital stay, d | 6 ± 8 | 6 ± 9 | 6 ± 6 | 6 ± 7 |
| CHA2DS2‐VASc scoreb | 1058c | |||
| CHA2DS2‐VASc score, mean ± SD | 1.7 (1.1) | 1.9 (1.1) | 1.9 (1.2) | 1.7 (1.1) |
| CHA2DS2‐VASc score, median (IQR) | 2 (1–2) | 2 (1–2) | 2 (1–2) | 2 (1–2) |
| HAS‐BLED scored | 1059c | |||
| HAS‐BLED score, mean ± SD | 2.1 ± 1.0 | 2.2 ± 1.0 | 2.2 (1.2) | 2.1 (1.0) |
| HAS‐BLED score, median (IQR) | 2 (1–3) | 2 (2–3) | 2 (1–3) | 2 (2–3) |
| Medications in the 3‐month period prior cohort entryb | N = 186 | N = 464 | ||
| β‐Blockers | — | 375 (46.5) | 86 (45.7) | 212 (45.7) |
| ACEIs/ARBs | — | 220 (27.0) | 50 (26.9) | 125 (26.9) |
| CCBs | — | 292 (36.2) | 62 (33.3) | 174 (37.5) |
| Diuretics | — | 246 (30.5) | 78 (41.9) | 125 (26.9) |
| Antiplatelet agents | — | |||
| Clopidogrel, prasugrel, ticagrelor | — | 90 (11.2) | 22 (11.8) | 54 (11.6) |
| ASA | — | 477 (59.1) | 101 (54.3) | 287 (61.9) |
| Lipid‐lowering agents | — | 506 (62.7) | 117 (62.9) | 293 (63.2) |
| Antidiabetics | — | 209 (25.9) | 52 (28.0) | 119 (25.7) |
| PPIs | — | 291 (36.1) | 74 (39.8) | 165 (35.6) |
| Antidepressants | — | 102 (12.6) | 23 (12.4) | 57 (12.3) |
| Amiodarone | — | 20 (2.5) | 7 (3.8) | 7 (1.51) |
| Genetics | ||||
| Responder status (FDA classificatione) | 1050c | 799c | 241c | 593c |
| Normal responder | 643 (61.2) | 487 (61.0) | 152 (62.6) | 364 (60.8) |
| Sensitive responder | 366 (34.9) | 278 (34.8) | 82 (33.7) | 202 (33.7) |
| Ultrasensitive responder | 41 (3.9) | 34 (4.2) | 7 (2.9) | 27 (4.5) |
| CYP2C9 phenotype (CPIC classificationf) | 1046c | 796c | 240c | 592c |
| Extensive metabolizer | 652 (61.6) | 493 (61.1) | 145 (59.7) | 377 (62.9) |
| Intermediate metabolizer | 350 (33.1) | 271 (33.6) | 85 (35.0) | 190 (31.7) |
| Poor metabolizer | 44 (4.2) | 32 (4.0) | 10 (4.1) | 25 (4.2) |
| VKORC1 phenotype (CPIC classificationf) | 1050c | 799c | 240c | 594c |
| High activity | 393 (37.1) | 298 (36.9) | 91 (37.5) | 222 (37.1) |
| Intermediate/extensive activity | 492 (46.5) | 373 (46.2) | 116 (47.7) | 275 (45.9) |
| Poor activity | 165 (15.6) | 128 (15.9) | 33 (13.6) | 97 (16.2) |
Abbreviations: ACEI, angiotensin‐converting enzyme inhibitor; AF, atrial fibrillation; ARB, angiotensin II receptor blocker; ASA, acetylsalicylic acid (aspirin); BMI, body mass index; CABG, coronary artery bypass grafting; CAD, coronary artery disease; CCB, calcium channel blocker; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; CPIC, Clinical Pharmacogenetics Implementation Consortium; CYP, cytochrome P450; DM, diabetes mellitus; FDA, US Food and Drug Administration; HTN, hypertension; IQR, interquartile range; MI, myocardial infarction; PCI, percutaneous coronary intervention; PPI, proton pump inhibitor; PVD, peripheral vascular disease; QWC, Québec Warfarin Cohort; RAMQ, Régie de l'assurance maladie du Québec; SD, standard deviation; TIA, transient ischemic attack; VKORC1, vitamin K epoxide reductase complex subunit 1.
Data are presented as n (%), mean ± SD, or median (IQR).
Data source: Questionnaire.
Data source: RAMQ dataset.
Number of patients for whom data is available.
The components of the scores are provided in Supporting Information, Table S4, in the online version of this article.
Coumadin (warfarin) FDA drug summary. http://www.accessdata.fda.gov/drugsatfda_docs/label/2010/009218s108lbl.pdf.
PharmGKB, http://www.pharmgkb.org.
Data on prescription drugs were available for 807 (76.2%) patients (Table 2). Major medication categories were lipid‐lowering agents (62.7%), antiplatelet agents (62.7% and 59.1% of low‐dose aspirin), β‐blockers (46.5%), calcium channel blockers (36.2%), and angiotensin‐converting enzyme inhibitors/angiotensin receptor blockers (27.0%). According to genetic variations, there were 44 (4.2%) participants with poor and 350 (33.1%) with intermediate metabolizer status at CYP2C9; 366 (34.9%) were sensitive responders according to the US Food and Drug Administration classification and 41 (3.9%) were ultrasensitive responders. The daily doses of warfarin prescribed were 4.5 (±1.6) mg, 4.7 (±2.2) mg, and 5.0 (±2.5) mg at baseline, the 3‐month time point, and the 12‐month time point, respectively (Table 3).
Table 3.
Daily warfarin dose requirements and TTR at baseline and follow‐up period
| Warfarin dose requirement after initiation of warfarin therapy | No. of Patients With Available Data (% of total) | Mean ±SD or n (%) |
|---|---|---|
| n (%) | mg/d | |
| At baseline | 1059 (100) | 4.5 ±1.6 |
| At 3 mo | 980 (92.5) | 4.7 ±2.2 |
| At 6 mo | 901 (85.1) | 4.9 ±2.3 |
| At 9 mo | 850 (80.3) | 5.0 ±2.4 |
| At 12 mo | 809 (76.2) | 5.0 ±2.5 |
| Poor anticoagulation control: TTR <60% after initiation of warfarin therapy | n (%) | No. of Patients |
| Baseline–3 months | 989 (93.4) | 531 (53.7) |
| 3–12 months | 842 (79.5) | 243 (28.9) |
| Time to therapeutic INR | n (%) | |
| Individuals who reached target INR during follow‐up | 1033 (97.5) | 818 (79.2) |
| Days to target INR for patients who reached target INR | 818 (77.2) | 72.3 ± 78.0 |
| Median and IQR | 41 (17–97) | |
| Days of follow‐up for patients who did not reach target INR | 214 (20.0) | 251.2 ±133.9 |
| Median and IQR | 354 (112–367) | |
| Time in therapeutic INR target range after initiation of warfarin therapy | n (%) | % Time |
| Baseline–3 mo | 989 (93.4) | 56.0 ±24.4 |
| 3–6 mo | 886 (84.0) | 68.3 ±27.3 |
| 6–9 mo | 809 (76.4) | 72.6 ±26.2 |
| 9–12 mo | 749 (70.7) | 70.7 ±28.0 |
| 3–12 mo | 842 (79.5) | 69.7 ±20.7 |
| 0–12 mo | 901 (85.1) | 64.9 ±18.9 |
| Time in supratherapeutic INR target range after initiation of warfarin therapy | n (%) | % Time |
| Baseline–3 mo | 989 (93.4) | 11.6 ±15.5 |
| 3–12 mo | 842 (79.5) | 12.8 ±15.7 |
| Time in subtherapeutic INR target range after initiation of warfarin therapy | n (%) | % Time |
| Baseline–3 mo | 989 (93.4) | 32.4 ±25.4 |
| 3–12 mo | 842 (79.5) | 17.5 ±18.2 |
| Ever‐occurrence of an INR ≥5 after initiation of warfarin therapy | n (%) | No. of Patients |
| Baseline–3 mo | 989 (93.4) | 93 (9.4) |
| 3–12 mo | 842 (79.5) | 48 (5.7) |
Abbreviations: INR, international normalized ratio; IQR, interquartile range; SD, standard deviation; TTR, time in therapeutic INR range.
During the 1‐year follow‐up, 148 (14.0%) patients were switched to another anticoagulant and 32 (3.0%) stopped anticoagulation therapy. In addition, 67 (6.3%) patients were lost during the study because of death, patient preference, or loss to follow‐up. As a result, 807 (76.2%) patients of the original cohort continued to receive warfarin and completed the 1‐year follow‐up (Figure 1).
Mean TTR increased from 56.0% ± 24.4% in the time interval between baseline and the 3‐month time point to 69.7% ± 20.7% between the 3‐ and 12‐month time points, where 71.1% of them had TTR ≥60% (Table 3). In the time interval between baseline and 3 months, participants spent 11.6% ± 15.5% of the time above the target range and 32.4% ± 25.4% below the target range. Among 1033 patients with available data, therapeutic INR was reached by 818 (79.2%) of participants in the 1‐year follow‐up and time to therapeutic INR took a median of 41 days (IQR, 17–97 days). The univariate significant predictors of low anticoagulation control were chronic kidney disease (CKD), prior major bleeding, dyslipidemia, aortic valve replacement, CHA2DS2‐VASc score, congestive heart failure (HF), and coronary artery disease (ORs ranging from 1.40 to 2.53; see Supporting Information, Table S5, in the online version of this article). In a multivariable model, predictors retained were CKD, HF, dyslipidemia, and age (see Supporting Information, Table S6, in the online version of this article), with a total model R 2 = 0.078 (P < 0.0001).
After 1 year of follow‐up, the rate of stroke or systemic embolism was 1.8 per 100 person‐years and 3.3 per 100 person‐years for major bleeding, mostly from major gastrointestinal and urinary tract bleeding (see Supporting Information, Table S7, in the online version of this article).
4. DISCUSSION
Factors that contribute to the variation in warfarin efficacy and complications are manifold and complex. The QWC was designed to systematically and prospectively investigate these multiple factors affecting individual variability to warfarin response and enable the conduct of research from an individualized‐therapy perspective in a large population of warfarin‐naïve patients. The cohort will be used to conduct genomic and candidate gene approaches to assess the influence of these factors on response to warfarin. Also, the potential association between nongenetic factors including sociodemographics, lifestyle habits, as well as co‐medications and comorbidities will be explored in regard to safety and efficacy of warfarin. Using the standardized questionnaires and the administrative registries, we have access to a validated dataset of several measures of quality of anticoagulation with warfarin, including dose‐ and INR‐related variables as well as hemorrhagic and thromboembolic events. Findings from the main objectives will be repeated in preplanned subgroups for primary indication of warfarin therapy, age (≥75 and <75 years), sex, CHA2DS2‐VASc score (<4 and ≥4), history of cerebrovascular disease, prior major bleeding, polypharmacy, and drug interaction, if applicable.32, 33, 34, 35
Mean TTR increased from 56.0% to 69.7% between baseline and the 12‐month time point, which is in accord with published data.2 The predictors of poor anticoagulation control (TTR < 60%) were CKD, prior major bleeding, dyslipidemia, aortic valve replacement, high CHA2DS2‐VASc score, congestive HF, and coronary artery disease. Those distinct patient populations raise concerns regarding optimal drug selection.36 A multivariable model defined an algorithm that included CKD, HF, dyslipidemia, and age. Moreover, the demographic and clinical characteristics observed in the QWC were similar to other population‐based registry studies.30, 31, 37, 38 The rate of thromboembolic events in QWC was 1.8 events per 100 patient‐years, compared with 1.3,31 1.6,30 and 0.9238 events per 100 patient‐years in other studies. And, we noted similar rates of major bleeding as other studies (eg, 3.3 vs 4.4,31 3.7,30 3.0,37 and 3.038 per 100 patient‐years).
4.1. Study limitations
The QWC has the advantage of having collected genetic material and having access to lifestyle habits. It also provides good generalizability for real‐world practice, possibly better than randomized controlled trials, which often have more restrictive populations. Data on co‐medications, however, were available only for a subgroup of patients, thus limiting some drug‐interaction analyses. As all participants were receiving care in the province of Quebec the generalizability of our data from the QWC may have some limitations and would need confirmation for populations with different characteristics. Also, the present findings are limited to new users, so they may not be generalizable to patients previously exposed to warfarin.
5. CONCLUSION
The QWC study benefits from a prospective and consecutive recruitment of incident users, compared with many previous studies investigating anticoagulation control, which have been retrospective. The study benefits from a rigorous strategy to assist identification of adverse events during follow‐up by using administrative databases.39 The QWC will be used to conduct genomic and candidate gene approaches to assess the influence of these factors on response to warfarin and outcomes, considering clinical characteristics and lifestyle, supporting the clinical utility of prescribing decisions to reduce harms of anticoagulation.
ACKNOWLEDGMENTS
The authors thank the RAMQ and Quebec Health Ministry for assistance in handling the data and the Commission d'accès à l'information du Québec for authorizing the study. The authors would like to thank all participants in the QWC.
Conflicts of interest
M.‐P. Dubé has received research support from DalCor, AstraZeneca, and Servier; has received honoraria from DalCor; holds minor equity interest in DalCor; and holds stocks in Xenon. She is mentioned as an author of a pending patent on warfarin therapy. R. Côté received a speaker honoraria 2 years ago from Pfizer and Sanofi. S. de Denus has received compensation from Servier, Pfizer, and Novartis for service as a consultant and was supported through grants from Pfizer, AstraZeneca, Roche Molecular Science, DalCor, and Novartis. A. Diaz has received honoraria from Pfizer, Bristol‐Myers Squibb, AstraZeneca, DalGen, Bayer, Sanofi, and Novartis. L. Lalonde has received partial research support from Boehringer‐Ingelheim Canada Ltd., Sanofi Canada, Bayer Inc., and Pfizer Canada Inc. J.‐C. Tardif holds minor equity interest in DalCor and has received research support from Amarin, AstraZeneca, DalCor, Eli Lilly, Ionis, Merck, Pfizer, Sanofi, and Servier, and has received honoraria from DalCor, Pfizer, Sanofi, and Servier. The authors declare no other potential conflicts of interest.
Supporting information
Appendix S1. Supplementary tables
Perreault S, Shahabi P, Côté R, et al. Rationale, design, and preliminary results of the Quebec Warfarin Cohort Study. Clin Cardiol. 2018;41:576–585. 10.1002/clc.22948
Funding information The study was funded by the Canadian Institutes of Health Research (CIHR) and the Center of Excellence in Personalized Medicine. M.éP. Dubé and S. Perreault received salary awards from the Fonds de Recherche du Québec Santé; P. Shahabi received a postdoctoral fellowship from the Canadian Gene Cure Foundation and CIHR; S. Dumas received a scholarship from CIHR.
REFERENCES
- 1. Johnson JA, Cavallari LH. Warfarin pharmacogenetics. Trends Cardiovasc Med. 2015;25:33–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Merli GJ, Tzanis G. Warfarin: what are the clinical implications of an out‐of‐range‐therapeutic international normalized ratio? J Thromb Thrombolysis. 2009;27:293–299. [DOI] [PubMed] [Google Scholar]
- 3. Carnes CA. What is the role of pharmacogenetics in optimization of warfarin dosing? Trends Cardiovasc Med. 2015;25:42–43. [DOI] [PubMed] [Google Scholar]
- 4. Jonas DE, McLeod HL. Genetic and clinical factors relating to warfarin dosing. Trends Pharmacol Sci. 2009;30:375–386. [DOI] [PubMed] [Google Scholar]
- 5. Ageno W, Gallus AS, Wittkowsky A, et al. Oral anticoagulant therapy. Antithrombotic Therapy and Prevention of Thrombosis. 9th ed: American College of Chest Physicians Evidence‐Based Clinical Practice Guidelines. Chest. 2012;141(2 suppl):e44S–e88S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Kimmel SE. Warfarin pharmacogenomics: current best evidence. J Thromb Haemost. 2015;13(suppl 1):S266–S271. [DOI] [PubMed] [Google Scholar]
- 7. Johnson JA, Caudle KE, Gong L, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for pharmacogenetics‐guided warfarin dosing: 2017 update. Clin Pharmacol Ther. 2017;102:397–404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Lip GY, Nieuwlaat R, Pisters R, et al. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor–based approach: the Euro Heart Survey on atrial fibrillation. Chest. 2010;137:263–272. [DOI] [PubMed] [Google Scholar]
- 9. Pisters R, Lane DA, Nieuwlaat R, et al. A novel user‐friendly score (HAS‐BLED) to assess 1‐year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey. Chest. 2010;138:1093–1100. [DOI] [PubMed] [Google Scholar]
- 10.Régie de l'assurance maladie du Québec. Accueil. 2016. http://www.ramq.gouv.qc.ca. Accessed September 1, 2016.
- 11.Régie de l'assurance maladie du Québec. Banque de données ministérielles MED‐ÉCHO. 2016. http://www.ramq.gouv.qc.ca/fr/donneeset-statistiques/employ%C3%A9-reseau/Pages/tableaux-donneesstatistiques.aspx. Accessed September 1, 2016.
- 12. Monfared AA, Lelorier J. Accuracy and validity of using medical claims data to identify episodes of hospitalizations in patients with COPD. Pharmacoepidemiol Drug Saf. 2006;15:19–29. [DOI] [PubMed] [Google Scholar]
- 13. Levy AR, Tamblyn RM, Fitchett D, et al. Coding accuracy of hospital discharge data for elderly survivors of myocardial infarction. Can J Cardiol. 1999;15:1277–1282. [PubMed] [Google Scholar]
- 14. Connolly SJ, Pogue J, Eikelboom J, et al; ACTIVE W Investigators . Benefit of oral anticoagulant over antiplatelet therapy in atrial fibrillation depends on the quality of international normalized ratio control achieved by centers and countries as measured by time in therapeutic range. Circulation. 2008;118:2029–2037. [DOI] [PubMed] [Google Scholar]
- 15. Razouki Z, Ozonoff A, Zhao S, et al. Pathways to poor anticoagulation control. J Thromb Haemost. 2014;12:628–634. [DOI] [PubMed] [Google Scholar]
- 16. Vestergaard AS, Skjøth F, Larsen TB, et al. The importance of mean time in therapeutic range for complication rates in warfarin therapy of patients with atrial fibrillation: a systematic review and meta‐regression analysis. PLoS One. 2017;12:e0188482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Reynolds MW, Fahrbach K, Hauch O, et al. Warfarin anticoagulation and outcomes in patients with atrial fibrillation: a systematic review and meta‐analysis. Chest. 2004;126:1938–1945. [DOI] [PubMed] [Google Scholar]
- 18. Van Den Ham HA, Klungel OH, Leufkens HG, et al. The patterns of anticoagulation control and the risk of stroke, bleeding and mortality in patients with non‐valvular atrial fibrillation. J Thromb Haemost. 2013;11:107–115. [DOI] [PubMed] [Google Scholar]
- 19. Rosendaal FR, Cannegieter SC, van der Meer FJ, et al. A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost. 1993;69:236–239. [PubMed] [Google Scholar]
- 20. Schmitt L, Speckman J, Ansell J. Quality assessment of anticoagulation dose management: comparative evaluation of measures of time‐in‐therapeutic range. J Thromb Thrombolysis. 2003;15:213–216. [DOI] [PubMed] [Google Scholar]
- 21. Kokotailo RA, Hill MD. Coding of stroke and stroke risk factors using International Classification of Diseases, revisions 9 and 10. Stroke. 2005;36:1776–1781. [DOI] [PubMed] [Google Scholar]
- 22. Thigpen JL, Dillon C, Forster KB, et al. Validity of international classification of disease codes to identify ischemic stroke and intracranial hemorrhage among individuals with associated diagnosis of atrial fibrillation. Circ Cardiovasc Qual Outcomes. 2015;8:8–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Tirschwell DL, Longstreth WT Jr. Validating administrative data in stroke research. Stroke. 2002;33:2465–2470. [DOI] [PubMed] [Google Scholar]
- 24. Cunningham A, Stein CM, Chung CP, et al. An automated database case definition for serious bleeding related to oral anticoagulant use. Pharmacoepidemiol Drug Saf. 2011;20:560–566. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Arnason T, Wells PS, van Walraven C, Forster AJ. Accuracy of coding for possible warfarin complications in hospital discharge abstracts. Thromb Res. 2006;118:253–262. [DOI] [PubMed] [Google Scholar]
- 26. Villines TC, Schnee J, Fraeman K, et al. A comparison of the safety and effectiveness of dabigatran and warfarin in non‐valvular atrial fibrillation patients in a large healthcare system. Thromb Haemost. 2015;114:1290–1298. [DOI] [PubMed] [Google Scholar]
- 27. Yao X, Abraham NS, Alexander GC, et al. Effect of adherence to oral anticoagulants on risk of stroke and major bleeding among patients with atrial fibrillation. J Am Heart Assoc. 2016. 10.1161/JAHA.115.003074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Lauffenburger JC, Farley JF, Gehi AK, et al. Effectiveness and safety of dabigatran and warfarin in real‐world US patients with non‐valvular atrial fibrillation: a retrospective cohort study. J Am Heart Assoc. 2015. 10.1161/JAHA.115.001798. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Larsen TB, Rasmussen LH, Gorst‐Rasmussen A, et al. Dabigatran and warfarin for secondary prevention of stroke in atrial fibrillation patients: a nationwide cohort study. Am J Med. 2014;127:1172.e5–1178.e5. [DOI] [PubMed] [Google Scholar]
- 30. Maura G, Blotière PO, Bouillon K, et al. Comparison of the short‐term risk of bleeding and arterial thromboembolic events in nonvalvular atrial fibrillation patients newly treated with dabigatran or rivaroxaban versus vitamin K antagonists: a French nationwide propensity‐matched cohort study. Circulation. 2015;132:1252–1260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Graham DJ, Reichman ME, Wernecke M, et al. Cardiovascular, bleeding, and mortality risks in elderly Medicare patients treated with dabigatran or warfarin for nonvalvular atrial fibrillation. Circulation. 2015;131:157–164. [DOI] [PubMed] [Google Scholar]
- 32. Kate MP, Shuaib A. Prevention of stroke in atrial fibrillation: cautious optimism. Lancet Neurol. 2012;11:203–205. [DOI] [PubMed] [Google Scholar]
- 33. Hankey GJ, Patel MR, Stevens SR, et al. Rivaroxaban compared with warfarin in patients with atrial fibrillation and previous stroke or transient ischaemic attack: a subgroup analysis of ROCKET AF. Lancet Neurol. 2012;11:315–322. [DOI] [PubMed] [Google Scholar]
- 34. Easton JD, Lopes RD, Bahit MC, et al; ARISTOTLE Committees and Investigators . Apixaban compared with warfarin in patients with atrial fibrillation and previous stroke or transient ischaemic attack: a subgroup analysis of the ARISTOTLE trial [published correction appears in Lancet Neurol 2012;11:1021]. Lancet Neurol. 2012;11:503–511. [DOI] [PubMed] [Google Scholar]
- 35. Ntaios G, Papavasileiou V, Diener HC, et al. Nonvitamin‐K‐antagonist oral anticoagulants in patients with atrial fibrillation and previous stroke or transient ischemic attack: a systematic review and meta‐analysis of randomized controlled trials. Stroke. 2012;43:3298–3304. [DOI] [PubMed] [Google Scholar]
- 36. Ziff OJ, Camm AJ. Individualized approaches to thromboprophylaxis in atrial fibrillation. Am Heart J. 2016;173:143–158. [DOI] [PubMed] [Google Scholar]
- 37. Abraham NS, Singh S, Alexander GC, et al. Comparative risk of gastrointestinal bleeding with dabigatran, rivaroxaban, and warfarin: population‐based cohort study. BMJ. 2015;350:h1857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Yao X, Abraham NS, Sangaralingham LR, et al. Effectiveness and safety of dabigatran, rivaroxaban, and apixaban versus warfarin in nonvalvular atrial fibrillation. J Am Heart Assoc. 2016. 10.1161/JAHA.116.003725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Park YK, Lee MJ, Kim JH, et al. Genetic and non‐genetic factors affecting the quality of anticoagulation control and vascular events in atrial fibrillation. J Stroke Cerebrovasc Dis. 2017;26:1383–1390. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix S1. Supplementary tables