Comparison of the Performance of the Warfarin Pharmacogenetics Algorithms in Patients with Surgery of Heart Valve Replacement and Heart Valvuloplasty

Association between time to therapeutic INR and length of stay following mechanical heart valve surgery

BACKGROUND

Warfarin is the only oral anticoagulant approved for use following mechanical valve surgery (MeVS). Patients may experience prolonged hospital length of stay (LOS) following MeVS awaiting an appropriate warfarin effect. We aimed to determine whether an association exists between time to achieve the first therapeutic international normalized ratio (INR) and LOS following MeVS.

MATERIALS AND METHODS

Retrospective single center cohort study. We included consecutive adult patients undergoing elective MeVS from 2013 to 2018. Landmark analyses and multivariable regression with time-updated INR were used to estimate the association between time to therapeutic INR (TTI) and LOS.

RESULTS

Among 384 patients (median age: 51 years, interquartile range [IQR]: 41-57; 58.3% male), the median TTI was 4 days (IQR: 2-5). Thirty seven percent of patients were discharged with a subtherapeutic INR, many on bridging anticoagulation or with an INR close to target. Those achieving therapeutic INR had an increased rate of hospital discharge (adjusted hazard ratio: 2.17; 95% confidence interval: 1.71-2.76; p < .0001). Attainment of a therapeutic INR anytime between postoperative Days 4 and 13 was significantly associated with a shorter LOS.

CONCLUSIONS

Prolonged time to achieve a therapeutic INR was independently associated with prolonged LOS. Future strategies aimed at improving attainment of therapeutic INR following MeVS may reduce hospital LOS.

An ensemble learning based framework to estimate warfarin maintenance dose with cross-over variables exploration on incomplete data set

MOTIVATION

Warfarin is a widely used oral anticoagulant, but it is challenging to select the optimal maintenance dose due to its narrow therapeutic window and complex individual factor relationships. In recent years, machine learning techniques have been widely applied for warfarin dose prediction. However, the model performance always meets the upper limit due to the ignoration of exploring the variable interactions sufficiently. More importantly, there is no efficient way to resolve missing values when predicting the optimal warfarin maintenance dose.

METHODS

Using an observational cohort from the Xinhua Hospital affiliated to Shanghai Jiaotong University School of Medicine, we propose a novel method for warfarin maintenance dose prediction, which is capable of assessing variable interactions and dealing with missing values naturally. Specifically, we examine single variables by univariate analysis initially, and only statistically significant variables are included. We then propose a novel feature engineering method on them to generate the cross-over variables automatically. Their impacts are evaluated by stepwise regression, and only the significant ones are selected. Lastly, we implement an ensemble learning based approach, LightGBM, to learn from incomplete data directly on the selected single and cross-over variables for dosing prediction.

RESULTS

377 unique patients with eligible and time-independent 1173 warfarin order events are included in this study. Through the comprehensive experimental results in 5-fold cross-validation, our proposed method demonstrates the efficiency of exploring the variable interactions and modeling on incomplete data. The R² can achieve 75.0% on average. Moreover, the subgroup analysis results reveal that our method performs much better than other baseline methods, especially in the medium-dose and high-dose subgroups. Lastly, the IWPC dosing prediction model is used for further comparison, and our approach outperforms it by a significant margin.

CONCLUSION

In summary, our proposed method is capable of exploring the variable interactions and learning from incomplete data directly for warfarin maintenance dose prediction, which has a great premise and is worthy of further research.

Comparison of the prediction performance of different warfarin dosing algorithms based on Chinese patients

Aim: To compare the prediction performance of different warfarin dosing algorithms based on Chinese patients. Materials & methods: A total of 18 algorithms were tested in 325 patients. The predictive efficacy of selected algorithms was evaluated by calculating the percentage of patients whose predicted dose fell within ±20% of their actual stable warfarin dose and the mean absolute error. Results: The percentage within ± 20% and the mean absolute error of the algorithms ranged from 11.9 to 41.2% and -0.20 (-0.29 to -0.11) mg/d to -1.63 (-1.75 to -1.50) mg/d. The algorithms established by Miao et al. and Wei et al. had optimal predictive performance. Conclusion: Algorithms based on geographical populations might be more suitable for the prediction of stable warfarin doses in local patients.

Warfarin Dose and CYP2C Gene Cluster: An African Ancestral-Specific Variant Is a Strong Predictor of Dose in Black South African Patients

Warfarin is a widely prescribed anticoagulant with a narrow therapeutic index. The rs12777823G>A single-nucleotide polymorphism (SNP) in the CYP2C gene cluster has been shown to influence optimal warfarin doses in African Americans. We report here effects of rs12777823G>A SNP on warfarin dose requirements in two South African population groups, black Africans (BA) and mixed ancestry (MA). A total of 425 participants on warfarin treatment were enrolled in the study. The age group of the studied population ranged between 44 and 66 years, with 69% females enrolled. Genetic characterization of the rs12777823G>A was done using the TaqMan SNP genotyping assay. To further compare effects of rs12777823G>A to those of other SNPs, VKORC1 g.-1639G>A and 4 SNPs in CYP2C9 gene (i.e., CYP2C9 c.430C>T, c.1075A>C, c.449G>A, and c.1003C>T) were analyzed. The rs12777823A variant allele frequencies were 0.28 and 0.25 in the BA and MA, respectively. The rs12777823A/A genotype was associated with significantly (p = 0.002) reduced mean warfarin dosage (27 ± 5.3 mg/week) compared with the G/G genotype (45 ± 16.1 mg/week) among BA, but not among the MA. The rs12777823G>A is located in a nongenomic region, suggesting that this SNP might be in linkage disequilibrium with another, likely causal SNP that is present in BA only. Given ongoing worldwide efforts to identify clinically relevant human genetic variation impacting on optimal warfarin dose selection, the African ancestry-specific genetic variant in the CYP2C cluster and others warrant further research and consideration in development of future warfarin dosing algorithms for precision medicine guidelines.

Warfarin maintenance dose Prediction for Patients undergoing heart valve replacement- a hybrid model with genetic algorithm and Back-Propagation neural network

Warfarin is the most recommended anticoagulant drug for patients undergoing heart valve replacement. However, due to the narrow therapeutic window and individual dose, the use of warfarin needs more advanced technology. We used the data collected from a multi-central registered clinical system all over China about the patients who have undergone heart valve replacement, subsequently divided into three groups (training group: 10673 cases; internal validation group: 3558 cases; external validation group: 1463 cases) in order to construct a hybrid model with genetic algorithm and Back-Propagation neural network (BP-GA), For testing the model's prediction accuracy, we used Mean absolute error (MAE), Root mean squared error (RMSE) and the ideal predicted percentage of total and dose subgroups. In results, whether in internal or in external validation group, the total ideal predicted percentage was over 58% while the intermediate dose subgroup manifested the best. Moreover, it showed higher prediction accuracy, lower MAE value and lower RMSE value in the external validation group than that in the internal validation group (p < 0.05). In conclusion, BP-GA model is promising to predict warfarin maintenance dose.

Identification of novel variants associated with warfarin stable dosage by use of a two-stage extreme phenotype strategy

Essentials Required warfarin doses for mechanical heart valves vary greatly. A two-stage extreme phenotype design was used to identify novel warfarin dose associated mutation. We identified a group of variants significantly associated with extreme warfarin dose. Four novel identified mutations account for 2.2% of warfarin dose discrepancies.

SUMMARY

Background The variation among patients in warfarin response complicates the management of warfarin therapy, and an improper therapeutic dose usually results in serious adverse events. Objective To use a two-stage extreme phenotype strategy in order to discover novel warfarin dose-associated mutations in heart valve replacement patients. Patients/method A total of 1617 stable-dose patients were enrolled and divided randomly into two cohorts. Stage I patients were genotyped into three groups on the basis of VKORC1-1639G>A and CYP2C9*3 polymorphisms; only patients with the therapeutic dose at the upper or lower 5% of each genotype group were selected as extreme-dose patients for resequencing of the targeted regions. Evaluation of the accuracy of the sequence data and the potential value of the stage I-identified significant mutations were conducted in a validation cohort of 420 subjects. Results A group of mutations were found to be significantly associated with the extreme warfarin dose. The validation work finally identified four novel mutations, i.e. DNMT3A rs2304429 (24.74%), CYP1A1 rs3826041 (47.35%), STX1B rs72800847 (7.01%), and NQO1 rs10517 (36.11%), which independently and significantly contributed to the overall variability in the warfarin dose. After addition of these four mutations, the estimated regression equation was able to account for 56.2% (R2Adj = 0.562) of the overall variability in the warfarin maintenance dose, with a predictive accuracy of 62.4%. Conclusion Our study provides evidence linking genetic variations in STX1B, DNMT3A and CYP1A1 to warfarin maintenance dose. The newly identified mutations together account for 2.2% of warfarin dose discrepancy.