Machine Learning for Prediction of Stable Warfarin Dose in US Latinos and Latin Americans.

Populations used to create warfarin dose prediction algorithms largely lacked participants reporting Hispanic or Latino ethnicity. While previous research suggests nonlinear modeling improves warfarin dose prediction, this research has mainly focused on populations with primarily European ancestry. We compare the accuracy of stable warfarin dose prediction using linear and nonlinear machine learning models in a large cohort enriched for US Latinos and Latin Americans (ULLA). Each model was tested using the same variables as published by the International Warfarin Pharmacogenetics Consortium (IWPC) and using an expanded set of variables including ethnicity and warfarin indication. We utilized a multiple linear regression model and three nonlinear regression models: Bayesian Additive Regression Trees, Multivariate Adaptive Regression Splines, and Support Vector Regression. We compared each model's ability to predict stable warfarin dose within 20% of actual stable dose, confirming trained models in a 30% testing dataset with 100 rounds of resampling. In all patients (n = 7,030), inclusion of additional predictor variables led to a small but significant improvement in prediction of dose relative to the IWPC algorithm (47.8 versus 46.7% in IWPC, p = 1.43 × 10-15). Nonlinear models using IWPC variables did not significantly improve prediction of dose over the linear IWPC algorithm. In ULLA patients alone (n = 1,734), IWPC performed similarly to all other linear and nonlinear pharmacogenetic algorithms. Our results reinforce the validity of IWPC in a large, ethnically diverse population and suggest that additional variables that capture warfarin dose variability may improve warfarin dose prediction algorithms.

Multi-site Investigation of Genetic Determinants of Warfarin Dose Variability in Latinos.

We conducted a multi-site investigation of genetic determinants of warfarin dose variability in Latinos from the U.S. and Brazil. Patients from four institutions in the United States (n = 411) and Brazil (n = 663) were genotyped for VKORC1 c.-1639G> A, common CYP2C9 variants, CYP4F2*3, and NQO1*2. Multiple regression analysis was used in the U.S. cohort to test the association between warfarin dose and genotype, adjusting for clinical factors, with further testing in an independent cohort of Brazilians. In the U.S. cohort, VKORC1 and CYP2C9 variants were associated with lower warfarin dose (ß = -0.29, P < 2.0 × 10-16 ; ß = -0.21, P = 4.7 × 10-7 , respectively) whereas CYP4F2 and NQO1 variants were associated with higher dose (ß = 0.10, P = 2 × 10-4 ; ß = 0.10, P = 0.01, respectively). Associations with VKORC1 (ß = -0.14, P = 2.0 × 10-16 ), CYP2C9 (ß = -0.07, P = 5.6 × 10-10 ), and CYP4F2 (ß = 0.03, P = 3 × 10-3 ), but not NQO1*2 (ß = 0.01, P = 0.30), were replicated in the Brazilians, explaining 43-46% of warfarin dose variability among the cohorts from the U.S. and Brazil, respectively. We identified genetic associations with warfarin dose requirements in the largest cohort of ancestrally diverse, warfarin-treated Latinos from the United States and Brazil to date. We confirmed the association of variants in VKORC1, CYP2C9, and CYP4F2 with warfarin dose in Latinos from the United States and Brazil.

Impact of adherence to warfarin therapy during 12 weeks of pharmaceutical care in patients with poor time in the therapeutic range.

Poor adherence to warfarin treatment is a contributor to poor quality of treatment, which increases the risk of bleeding and thromboembolic events. This study aims to evaluate the impact of adherence to warfarin therapy on anticoagulation quality during 12 weeks of pharmaceutical care and after 1 year of follow-up for patients with atrial fibrillation and with poor TTR. The Arrhythmia Unit of tertiary hospital in Brazil. We included 262 patients with AF and poor quality of anticoagulation therapy with warfarin (TTR < 50%). Pharmacist-driven therapy management was performed for 12 weeks and patients were also evaluated 1 year after the end of the follow-up with a pharmacist. Adherence was classified into high adherence, medium adherence and low adherence. Impact of adherence to warfarin therapy after pharmaceutical care. Of the 262 patients, 160 were high adherence, 71 were medium adherence and 31 were low adherence. No statistically significant difference is found between adherence groups in demographic and clinical variables. The TTR basal means were not different among adherence groups (p = 0.386). However, the means of TTR 12 weeks and TTR 1 year after the end of protocol were statistically different among adherence groups (p < 0.001 and p = 0.002, respectively). When we compared TTR values at different times within the adherence group, we observed that there is a statistical difference between the three TTR means (basal versus 12 weeks versus 1 year after) within the adherence group (p < 0.001). Patients with poor anticoagulation control, who adhered to the treatment with warfarin during the pharmaceutical care had better anticoagulation quality compared to those who did not adhere to the therapy with warfarin.

Evaluation of the Long-Term Impact on Quality After the End of Pharmacist-Driven Warfarin Therapy Management in Patients With Poor Quality of Anticoagulation Therapy.

BACKGROUND: Warfarin is the most common oral anticoagulant drug, especially in low-income and emerging countries, because of the high cost of direct oral anticoagulant (DOACs), or when warfarin is the only proven therapy (mechanical prosthetic valve and kidney dysfunction). The quality of warfarin therapy is directly associated with dose management. Evidence shows that pharmaceutical care achieves a better quality of therapy with warfarin. However, there are no studies showing this intervention in a specific patient group with poor quality of anticoagulation in a long period after the end of the follow-up by a pharmacist. Thus, the aim of this study was to evaluate whether the quality of warfarin therapy driven by a pharmacist remains stable in the long term after the end of follow up with a pharmacist, in AF patients with poor quality of anticoagulation. METHODS: This is a prospective study, which evaluated about 2,620 patients and selected 262 patients with AF and poor quality of anticoagulation therapy with warfarin (TTR<50% - based on the last three values of international normalized ratio). Pharmacist-driven therapy management was performed up to 12 weeks. Data from patients were evaluated 1 year after the end of the follow-up with pharmacist. RESULTS: Comparison between mean TTR after 12 weeks of pharmaceutical care (54.1%) and mean TTR one year after the end of the pharmaceutical care (56.5%; p=0.081) did not achieve statistical difference, demonstrating that the increment of quality due to intervention of 12 weeks was maintained for 1 year after intervention. CONCLUSION: The long-term impact of pharmaceutical care was beneficial for patients with AF and poor quality of warfarin anticoagulation. This design might be an important strategy to treat a subgroup of patients without proven effectiveness of warfarin.

Genetic Factors Influencing Warfarin Dose in Black-African Patients: A Systematic Review and Meta-Analysis.

Warfarin is the most commonly used oral anticoagulant in sub-Saharan Africa. Dosing is challenging due to a narrow therapeutic index and high interindividual variability in dose requirements. To evaluate the genetic factors affecting warfarin dosing in black-Africans, we performed a meta-analysis of 48 studies (2,336 patients). Significant predictors for CYP2C9 and stable dose included rs1799853 (CYP2C9*2), rs1057910 (CYP2C9*3), rs28371686 (CYP2C9*5), rs9332131 (CYP2C9*6), and rs28371685 (CYP2C9*11) reducing dose by 6.8, 12.5, 13.4, 8.1, and 5.3 mg/week, respectively. VKORC1 variants rs9923231 (-1639G>A), rs9934438 (1173C>T), rs2359612 (2255C>T), rs8050894 (1542G>C), and rs2884737 (497T>G) decreased dose by 18.1, 21.6, 17.3, 11.7, and 19.6 mg/week, respectively, whereas rs7294 (3730G>A) increased dose by 6.9 mg/week. Finally, rs12777823 (CYP2C gene cluster) was associated with a dose reduction of 12.7 mg/week. Few studies were conducted in Africa, and patient numbers were small, highlighting the need for further work in black-Africans to evaluate genetic factors determining warfarin response.

Pharmaceutical Care Increases Time in Therapeutic Range of Patients With Poor Quality of Anticoagulation With Warfarin.

Thromboembolic events are associated with high mortality and morbidity indexes. In this context, warfarin is the most widely prescribed oral anticoagulant agent for preventing and treating these events. This medication has a narrow therapeutic range and, consequently, patients usually have difficulty in achieving and maintaining stable target therapeutics. Some studies on the literature about oral anticoagulant management showed that pharmacists could improve the efficiency of anticoagulant therapy. However, the majority of these studies included general patients retrospectively. The aim of this study was to prospectively evaluate a pharmacist's warfarin management in patients with poor quality of anticoagulation therapy (Time in the Therapeutic Range- TTR < 50%). We included 268 patients with atrial fibrillation (AF) and without stable dose of warfarin (TTR < 50%, based on the last three values of International Normalized Ratio-INR). We followed them up for 12 weeks, INR values were evaluated and, when necessary, the dose adjustments were performed. During the first four visits, patient's INR was measured every 7 days. Then, if INR was within the target therapeutic range (INR: 2-3), the patient was asked to return in 30 days. However, if INR was out the therapeutic target, the patient was asked to return in 7 days. Adherence evaluation was measured through questionnaires and by counting the pills taken. Comparison between basal TTR (which was calculated based on the three last INR values before prospective phase) and TTR of 4 weeks (calculated by considering the INR tests from visits 0 to 4, in the prospective phase of the study) and basal TTR and TTR of 12 weeks (calculated based on the INR tests from visits 0 to 12, in the prospective phase of the study) revealed significant statistical differences (0.144 ± 0.010 vs. 0.382 ± 0.016; and 0.144 ± 0.010 vs. 0.543 ± 0.014, p < 0.001, respectively). We also observed that the mean TTR of 1 year before (retrospective phase) was lower than TTR value after 12 weeks of pharmacist-driven treatment (prospective phase) (0.320 ± 0.015; 0.540 ± 0.015, p < 0.001). In conclusion, pharmaceutical care was able to improve TTR values in patients with AF and poor quality of anticoagulation with warfarin.

Impact of incorporating ABCB1 and CYP4F2 polymorphisms in a pharmacogenetics-guided warfarin dosing algorithm for the Brazilian population.

PURPOSE: Interpatient variation of warfarin dose requirements may be explained by genetic variations and general and clinical factors. In this scenario, diverse population-calibrated dosing algorithms, which incorporate the main warfarin dosing influencers, have been widely proposed for predicting supposed warfarin maintenance dose, in order to prevent and reduce adverse events. The aim of the present study was to evaluate the impact of the inclusion of ABCB1 c.3435C>T and CYP4F2 c.1297G>A polymorphisms as additional covariates in a previously developed pharmacogenetic-based warfarin dosing algorithm calibrated for the Brazilian population. METHODS: Two independent cohorts of patients treated with warfarin (n = 832 and n = 133) were included for derivation and replication of the algorithm, respectively. Genotyping of ABCB1 c.3435C>T and CYP4F2 c.1297G>A polymorphisms was performed by polymerase chain reaction followed by melting curve analysis and TaqMan® assay, respectively. A multiple linear regression was performed for the warfarin stable doses as a dependent variable, considering clinical, general, and genetic data as covariates. RESULTS: The inclusion of ABCB1 and CYP4F2 polymorphisms was able to improve the algorithm's coefficient of determination (R2) by 2.6%. In addition, the partial determination coefficients of these variants revealed that they explained 3.6% of the warfarin dose variability. We also observed a marginal improvement of the linear correlation between observed and predicted doses (from 59.7 to 61.4%). CONCLUSION: Although our study indicates that the contribution of the combined ABCB1 and CYP4F2 genotypes in explaining the overall variability in warfarin dose is not very large, we demonstrated that these pharmacogenomic data are statistically significant. However, the clinical relevance and cost-effective impact of incorporating additional variants in warfarin dosing algorithms should be carefully evaluated.

Association Between ABCB1 Polymorphism and Stable Warfarin Dose Requirements in Brazilian Patients.

The ideal dose of the oral anticoagulant warfarin varies widely among patients, mainly due to genetic factors. Genetic variations that impact warfarin pharmacokinetics and the vitamin K cycle are plausible candidates for being associated with warfarin dose requirements. Therefore, the aim of this study was to assess whether polymorphisms in the ABCB1 and CYP4F2 genes were associated with stable warfarin dose requirements in Brazilian patients. This retrospective study included samples from 309 individuals. Genotyping of ABCB1 c.3435C>T and CYP4F2 c.1297G>A were performed by polymerase chain reaction followed by melting curve analysis (HRM-PCR) and TaqMan® genotyping assay, respectively. Stable doses were adjusted in a linear multiple regression model for age, gender, body mass index, self-reported race, use of amiodarone, CYP2C9 (*2 and *3), VKORC1 c.1639G>A, and ABCB1 c.3435C>T or CYP4F2 c.1297G>A. By performing a univariate analysis of variance, we found that the warfarin patients who carry ABCB1 c.3435T variant alleles (CT and TT genotypes) need fewer warfarin stable doses in comparison with the individuals that are CC wild-type: 2.5 (p = 0.003) and 4.3 (p < 0.001) mg/week less, respectively, for the overall group of patients on stable anticoagulation therapeutics (n = 309); and 5.5 (p = 0.006) and 10.2 (p < 0.001) mg/week less, respectively, for the self-declared non-white stable subgroup (n = 76). No statistically significant differences in dose requirements were observed according to CYP4F2 genotypes. In conclusion, our results suggest ABCB1 c.3435C>T variant may influence warfarin dose requirements in Brazilian patients, when associated with other genotypic, demographic and clinical factors.

Genotype-guided warfarin therapy: current status.

Warfarin pharmacogenomics has been an extensively studied field in the last decades as it is focused on personalized therapy to overcome the wide interpatient warfarin response variability and decrease the risk of side effects. In this expert review, besides briefly summarizing the current knowledge about warfarin pharmacogenetics, we also present an overview of recent studies that aimed to assess the efficacy, safety and economic issues related to genotype-based dosing algorithms used to guide warfarin therapy, including randomized and controlled clinical trials, meta-analyses and cost-effectiveness studies. To date, the findings still present disparities, mostly because of standard limitations. Thus, further studies should be encouraged to try to demonstrate the benefits of the application of warfarin pharmacogenomic dosing algorithms in clinical practice.

Evaluation of a pharmacogenetic-based warfarin dosing algorithm in patients with low time in therapeutic range - study protocol for a randomized controlled trial.

BACKGROUND: Time in therapeutic range (TTR) is a measurement of quality of warfarin therapy and lower TTR values (<50%) are associated with greater risk of thromboembolic and bleeding events. Recently, we developed a pharmacogenetic-based warfarin dosing algorithm specifically calibrated for a Brazilian patient sample. The aims of this study are: to evaluate the impact of a genetic-based algorithm, compared to traditional anticoagulation, in the time to achieve the therapeutic target and in TTR percentage; and to assess the cost-effectiveness of genotype-guided warfarin dosing in a specific cohort of patients with low TTR (<50%) from a tertiary cardiovascular hospital. METHODS/DESIGN: This study is a randomized controlled trial in patients (n = 300) with atrial fibrillation with TTR < 50%, based on the last three INR values. At the first consultation, patients will be randomized into two groups: TA group (traditional anticoagulation) and PA group (pharmacogenetic anticoagulation). For the first group, the physician will adjust the dose according to current INR value and, for the second group, a pharmacogenetic algorithm will be used. At the second, third, fourth and fifth consultations (with an interval of 7 days each) INR will be measured and, if necessary, the dose will be adjusted based on guidelines. Afterwards, patients who are INR stable will begin measuring their INR in 30 day intervals; if the patient's INR is not stable, the patient will return in 7 days for a new measurement of the INR. Outcomes measures will include the time to achieve the therapeutic target and the percentage of TTR at 4 and 12 weeks. In addition, as a secondary end-point, pharmacoeconomic analysis will be carried out. Ethical approval was granted by the Ethics Committee for Medical Research on Human Beings of the Clinical Hospital of the University of São Paulo Medical School. DISCUSSION: This randomized study will include patients with low TTR and it will evaluate whether a population-specific genetic algorithm might be more effective than traditional anticoagulation for a selected group of poorly anticoagulated patients. TRIAL REGISTRATION: ClinicalTrials.gov, NCT02592980 . Registered on 29 October 2015.

Age is associated with time in therapeutic range for warfarin therapy in patients with atrial fibrillation.

BACKGROUND: Warfarin is the most prescribed oral anticoagulant used for preventing stroke in patients with atrial fibrillation. Time in the therapeutic range (TTR) has been accepted as the best method to evaluate the quality of warfarin therapy. The main aim of the present study was to evaluate the impact of variables on the time in the therapeutic range for warfarin therapy in patients with atrial fibrillation from a referral cardiovascular hospital. METHODS: This retrospective study included 443 patients were included (190 patients with age < 65 years and 253 patients with age ≥65 years) from 2011 to 2014 and TTR was computed according to Rosendaal's method. RESULTS: Patients with age ≥65 years had higher TTR value (67±22%) compared with patients with < 65 years (60±24%) (p = 0.004). In a linear regression model, only age ≥65 years emerged as a significant predictor of greater TTR values. In multivariate logistic regression model, the variable age ≥65 years was associated with higher OR for having a TTR higher than the median value (OR = 2.17, p < 0.001). CONCLUSION: We suggest that the age influenced TTR through greater drug adherence. Strategies for increasing drug adherence might improve quality of warfarin anticoagulation.

Development of a pharmacogenetic-based warfarin dosing algorithm and its performance in Brazilian patients: highlighting the importance of population-specific calibration.

BACKGROUND: The main aims of the present study were to develop a pharmacogenetic-based warfarin dosing algorithm and to validate it in a highly admixed population. MATERIALS & METHODS: We included two patient cohorts treated with warfarin (first cohort, n = 832; and second cohort, n = 133). RESULTS: Our algorithm achieved a determination coefficient of 40% including the variables age, gender, weight, height, self-declared race, amiodarone use, enzyme inducers use, VKORC1 genotypes and predicted phenotypes according to CYP2C9 polymorphisms. CONCLUSION: Data suggest that our developed algorithm is more accurate than the IWPC algorithm when the application is focused on patients from the Brazilian population. Population-specific derivation and/or calibration of warfarin dosing algorithms may lead to improved performance compared with general use dosing algorithms currently available. Original submitted 26 November 2014; Revision submitted 9 April 2015.