Low Performance of a Clinical-Genetic Model in the Estimation of Time in Therapeutic Range in Acenocoumarol-Adherent Patients with Nonvalvular Atrial Fibrillation: The Quality of Anticoagulation Challenge

Abstract

Background. Anticoagulation with vitamin K antagonists continues to be a challenging task given the difficulty of achieving a correct time in therapeutic range (TTR). The SAMeTT₂R₂ score has been proposed to identify patients that will be good responders. In this study we aimed to analyse clinical and genetic factors involved in a correct level of anticoagulation in patients with atrial fibrillation and thereby potentially improve the diagnostic performance of SAMeTT₂R₂ score. Methods. We prospectively included 212 consecutive patients with nonvalvular atrial fibrillation under treatment with acenocoumarol for at least 6 months that were attended in a cardiology outpatient clinic and were categorized as adherent to medication. We carried out a multivariate regression analysis to detect the independent predictive factors of good control. In all patients VKORC1, CYP2C92, CYP2C93, and MIR133A2 genotyping was performed. Results. A total of 128 (60.4%) patients presented TTR <70% (average TTR = 63.2). We identified body mass index (OR 0.94, 95%CI 0.89-0.99, p=0.032) and regular vitamin K intake (OR 0.53, 95%CI 0.28-0.99, p= 0.046) as independent predictors of poor anticoagulation control. The discriminatory power of a clinical-genetic model derived from our cohort was significantly better compared to the SAMeTT₂R₂ score (C-statistic 0.658 versus 0.524, p<0.001). Conclusions. In our study the SAMeTT₂R₂ score revealed a poor ability in the prediction of TTR. Besides SAMeTT₂R₂, body mass index and possibly vitamin K intake should be taken into account when deciding the optimal anticoagulation strategy. The information provided by the identified genotypes was marginal.

1. Introduction

The efficacy and safety of treatment with vitamin K antagonists (VKAs) in patients with nonvalvular atrial fibrillation (NVAF) strongly depends on the capability of achieving and maintaining a stable level of correct anticoagulation [1, 2]. Despite the introduction of non-vitamin K antagonist oral anticoagulants (NOAC), the classical VKAs such as acenocoumarol and warfarin remain the most widely used anticoagulant therapy world-wide [3].

Anticoagulation state is monitored by international normalized ratio (INR) whereas quality of anticoagulation is usually assessed with time in therapeutic range (TTR) defined by Rosendaal [4]. It has been proven that the longer the time spent in therapeutic range the lower the risk of emboli or haemorrhage [5]. Although TTR seems to be relatively high in clinical trials, anticoagulation control in the “real-world” is not so strict and TTR results far from optimal [6]. This issue could make NOACs particularly cost-effective when adequate TTR is not achieved with classic anticoagulants [7].

There are very few studies that have appropriately evaluated the quality of anticoagulation in patients with NVAF [8–10]. It is worth noting that assessing quality of anticoagulation is not an easy task since it is a dynamic process with considerable within-patient variation regarding adherence, previous medical status or medical history, pharmacological interactions, and other contributing factors apart from the relatively complex TTR estimation.

Taking all this into consideration, it would be of great interest to have a simple method based on clinical variables that would have the ability to identify those patients under treatment with acenocoumarol that are prone to an inadequate TTR. The SAMeTT₂R₂ score has been proposed with such purpose [11]. Authors initially reported that a score of ≥ 2 predicted an inadequate TTR and showed a good discrimination (C-statistic 0.72) in a derivation cohort. However, the SAMeTT₂R₂ score was elaborated from a group of patients with AF included in a trial, the AFFIRM study [12], and therefore a potential selection bias (lower age, high proportion of non-Caucasians) is plausible. Moreover, there was a concern due to the well-known “statistical overoptimism” [13]. In an initial attempt to further validate this score, Apostolakis [11] reported a C-statistic of 0.70 (95% CI 0.57-0.82) in a small external cohort. However, a more recent study in a bigger sample [14] has reported lower than expected performance of SAMeTT₂R₂ (C-statistic 0.57, 95% CI 0.53-0.60), thus underlying the need to improve the performance of this score.

In this regard, TTR prediction could improve by the determination of genetic polymorphisms involved in the metabolism of acenocoumarol as proved by previous investigations [15]; the most researched ones are CYP2C9, (alleles 2 and 3) and rs9923231 (-1639C>T) in VKORC [16–19]. Along the same line, the progress in genomic technology and bioinformatics has led to the study of micro-RNA that can play an important role in the response to specific treatments. It has been reported in a small exploratory study that variations in MIR133A2 genes can lead to an aberrant expression of VKORC1, thus altering the efficacy of treatment with warfarin [20]. Additionally, as mentioned, the seldomly analysed level of patients’ adherence as well as other clinical factors might have a role in TTR prediction.

The aim of the present study was to further explore clinical and genetic factors that might be involved in a correct level of anticoagulation with acenocoumarol in patients with NVAF, categorized as adherent according to Morisky-Green scale [21], and thus potentially improve the diagnostic performance of SAMeTT₂R₂ score.

2. Materials and Methods

2.1. Study Population and Definitions

From 1st of December 2014 to 30th of June 2016, we prospectively enrolled consecutive patients diagnosed with NVAF, regardless the type, that were attended at the cardiology outpatient clinic of a tertiary hospital and were under treatment with acenocoumarol for at least 6 months prior inclusion in the study. All patients were considered eligible if they were categorized as “adherent” to the medication according to the Morisky-Green scale (four out of four negative answers in the questionnaire) [21]. Exclusion criteria were (1) moderate or severe rheumatic mitral stenosis, (2) biologic or mechanic mitral valve prothesis, (3) unavailability of INR values for the determination of TTR during the treatment period, (4) recent interruptions (<6 months) of anticoagulation treatment, (5) and anticoagulation with warfarin. The study was performed according to the Declaration of Helsinki and approved by the Ethics Committee for Clinical Research at our centre.

The definitions of other included clinical variables were as follows. Chronic kidney disease was defined as estimated glomerular filtration rate by MDRD-4 <60 mL/min/1.73 m²; chronic hepatic disease was defined as persistent elevation of transaminases 3-fold the upper limit of normal, previous history of cirrhosis, hepatitis, or any other chronic liver disorder. The presence of comorbidities was assessed with the Charlson score [22]. Regular vitamin K intake was defined as consumption of green leafy vegetables such as spinach salad, broccoli, and cabbage, from three to seven days a week. Number of active medications included in patients’ standard treatment was quantified as number of tablets taken per day apart from VKA. All abovementioned SAMeTT₂R₂ score items were included: female sex, age<60 years, medical history (≥3 comorbidities among the following: hypertension, diabetes mellitus, ischemic heart disease, peripheral arterial disease, heart failure, stroke, pulmonary disease, and liver or renal disease); treatment (interacting drugs such as amiodarone), active smoking, and non-Caucasian race.

2.2. Quality of Anticoagulation

INR values of the 6 months prior to the study entry were registered and Rosendaal method [4] was used to assess quality of anticoagulation. Adequate anticoagulation control was defined as an estimated TTR ≥70% [22, 23].

2.3. Blood Samples and Laboratory Methods

Peripheral blood samples (5-10 ml) were obtained in EDTA tubes and DNA was extracted using QIAamp DNA minikit and automatic nucleic acid extractor QiaCube (Qiagen, Hilden, Germany). Four SNPs tagging alleles were carefully analysed: rs1799853 (CYP2C92), rs1057910 (CYP2C93), rs9923231 (-1639C>T VKORC), and rs4554 (MIR133A2).

Polymorphisms were determined in 96-well plates on a 7500F real-time thermocycler (Applied Biosystems Foster City, CA, USA) using competitive allele-specific PCR (polymerase chain reaction) KASPar probes which are based on FRET (fluorescent resonance energy transfer) technology and following manufacturer’s instructions.

3. Statistical Analysis

Continuous variables were presented as means (with standard deviations) or medians (with 25th and 75th percentiles). Categorical variables were expressed as frequencies and percentages. The Kolmogorov-Smirnov test and frequency histograms were applied to establish the normality of the included variables. Baseline characteristics were compared between patients with adequate (TTR≥70%) or inadequate (TTR<70%) anticoagulation control. Continuous variables were compared with Student’s t test or Mann-Whitney test as appropriate. Categorical variables were compared with Chi-square or exact Fisher test. We used a binary logistic regression model to predict a TTR<70% (dependent variable) including both clinical and genetic variables. Assumptions of the model were previously tested. Odds ratio and 95% confidence interval (CI) were calculated for each covariate as well as the discrimination of the model by using C-statistic (estimation of the area under the curve (AUC)) and the Hosmer-Lemeshow test to assess calibration. We chose variables with p<0.15 to develop a multivariate regression model using backward method for the clinical variables and enter method for the genetic variables. Likelihood ratio test was used to assess the significance of each variable. To further check the consistency of the multivariable model we performed a 3000-iteration bootstrapped enter method analysis. In all tests, a two-sided p-value 0.05 was considered significant. Software packages SPSS 21.0 (SPSS Inc., Chicago, IL, USA) and STATA 12.0 (StataCorp, USA) were used for the statistical analyses.

4. Results

4.1. Baseline Characteristics

We included two hundred and twelve patients with a mean age of 74 years (standard deviation of 9 years), and 105 (50%) were men. Baseline characteristics of the study population are shown in Table 1. A total of 128 (60.4%, 95% CI 53.7-67.0%) patients presented TTR <70%. The mean TTR was 63.2 (standard deviation 20.3). Mean SAMeTT₂R₂ score was 1.3 ± 1.0 (median 1, p25-p75 1-2, range 4).

4.2. Predictors of Poor Anticoagulation Control

Body mass index (OR 0.93, 95%CI 0.88-0.99, p=0.015), previous heart failure (OR 1.94, 95%CI 0.91-4.15 p= 0.085), regular vitamin K intake (OR 0.52, 95%CI 0.28-0.95, p=0.032), persistent or permanent AF (OR 1.74, 95%CI 0.99-3.05, p=0.055), and number of active medications (OR 1.09, 95%CI 0.99-1.20, p=0.090) were associated (p<0.15) with a TTR <70% (Table 3). In a multivariable setting we found that body mass index (OR 0.94, 95%CI 0.89-0.99, p=0.032) and regular vitamin K intake (OR 0.53, 95%CI 0.28-0.99, p=0.046) were independent predictors of poor anticoagulation control in a correctly calibrated multivariable model (Table 4). In the bootstrapped model, only body mass index remained as an independent predictor (OR 0.94, 95% CI 0.87-0.99, p=0.044).

4.3. Polymorphisms. Frequency and Impact on TTR

The genotype distribution of the VKORC polymorphism was CC 35%, CT 44% and TT 21%. CYP2C92 presented genotype frequencies of CC 65%, CT 33%, TT 2%. Genotype frequency of CYP2C93 was AA 84% and CA 16%. Finally, MIR133A2 genotype was distributed as follows, GG 60%, AA 11%, GA 28%, undetermined 1%.

The frequencies of VKORC, CYP2C92, CYP2C93, and MIR133A2 polymorphisms according to TTR are shown in Table 2. Patients with poor anticoagulation control presented a trend towards a higher prevalence of at least one T allele in the VKORC polymorphism. Notably, all genotype frequencies agreed with the Hardy-Weinberg equilibrium p=0.57 (X² =2.38), p=0.81 (X² =1.31), p=0.92 (X² =1.71), and p=0.75 (X²=14.89), respectively.

4.4. Diagnostic Performance of SAMeTT₂R₂ Score and a Clinical-Genetic Model

Discrimination of SAMeTT₂R₂ as reflected by the C-statistic demonstrated a poor performance in our study population (AUC 0.524 95% CI 0.442-0.606) to detect a TTR≥70%. Moreover, C- statistic for the model that included SAMeTT₂R₂ score plus four analysed polymorphisms was 0.545 (95% CI 0.465-0.626; p=0.269 for comparison with SAMeTT₂R₂). Finally, the AUC corresponding to the model that included two clinical variables (body mass index and regular vitamin K intake) and four polymorphisms was 0.658 (95%CI 0.584-0.732). The increment of the discrimination capacity yielded by the clinical-genetic model compared to SAMeTT₂R₂ score alone was 28.8%, p<0.001. Also, the discrimination capacity of the clinical model and the SAMeTT₂R₂ score above the SAMeTT₂R₂ alone was 20.5% (p=0.034) (Figure 1).

Figure 1 

Diagnostic performance of SAMe-T and SAMe-T plus four polymorphisms/clinical model and clinical-genetic model in the detection of TTR≥70%. AUC, area under the curve; CI, confidence interval. The four included polymorphisms are VKORC, CYP2C9*2, CYP2C9*3, and MIR133A2. The clinical model comprises body mass index (kg/m²) and regular vitamin K intake.

5. Discussion

The first and still the most widely used anticoagulants are the classical VKAs such as acenocoumarol and warfarin [24]. Thus, we believe that it seems reasonable to continue exploring the quality of anticoagulation and the possible factors involved in a poor control of such therapy. In this regard Apostolakis et al. presented the aforementioned SAMeTT₂R₂ score [11]. Apart from the derivation and internal validation cohorts, the author analysed the SAMeTT₂R₂ score performance in a smaller external “real-world” validation cohort. The C-statistic of 0.70 (95%CI 0.57-0.82) reported for this cohort was calculated for discrimination of the 5th percentile (TTR≥64%) of this sample. This notorious result was in contrast with the large nationwide study conducted by Ruiz- Ortiz et al. [14] with a C-statistic of 0.57(95%CI 0.53-0.60) for the prediction of TTR≥65%. Consistently, in our study, SAMeTT₂R₂ showed a C-statistic of 0.549 (95% CI 0.472-0.626) for the prediction of TTR ≥ 65%. Notably, our results are in consonance with the poor predictive ability of SAMeTT₂R₂ score reported in a high-quality setting of a Danish cohort by Jane Skovet al. [25] who applied SAMeTT₂R₂ score to a small cohort of patients with a mean TTR of 76%. SAMeTT₂R₂ score showed a very low prediction of TTR (adjusted R²= 4%) while the use of a model that included age, amiodarone use, alcohol consumption, and perceived stress showed more than double the R² value. Anticoagulation therapy is not exempt from complications as bleeding and drawbacks such as the need of monitoring and interactions with vitamin-K-rich food, aspects that may affect adherence to such therapy. The impact of the adherence to anticoagulants in the performance of the prediction tools for TTR has not been sufficiently analysed. Thus, in an attempt to overcome this pitfall, we used Morisky-Green scale so as to include only “adherent” patients. It has been reported that this scale shows low sensitivity but high specificity and positive predictive value [26]. To the best of our knowledge, these types of medication adherence scales have not been previously used in other studies in a setting similar to ours.

Clearly, decision making in anticoagulation therapy cannot be left to low performance scores, so the search for more precise predictors in guaranteed. In this regard, we found that the combination of only two clinical variables (body mass index and regular vitamin K intake) and four genetic polymorphisms modestly—but significantly—improved diagnostic performance of SAMeTT₂R₂ score (C-statistic = 0.658, 95% CI 0.584-0.732). Along the same line, Abumuaileq et al. carried out a retrospective analysis of a real-world cohort of patients with NVAF [27]. The SAMeTT₂R₂ score C-statistic was evidently poor and mildly improved from 0.56 to 0.60 by adding new factors such as alcohol abuse, low glomerular filtration rate, diabetes mellitus, heart failure, and history of malignancy in accordance with previous reports [28]. However, the study had limitations since authors used the percentage of INR in therapeutic range (PINRR) not completely equivalent to TTR.

Another study worth mentioning is the one published by Lobos-Bejarano et al. [29] with a large sample size based on the PAULA cohort [30]. By applying SAMeTT₂R₂ score in a real-life scenario they confirm its modest prediction capability of INR control (C-index 0.54-0.58) and they identify easy-to-collect factors (seven or more tablets per day, dietary habits, and bleeding history) capable of improving it vaguely. The implication of dietary habits that include vitamin K-rich foods (OR 0.52, 95% CI 0.28-0.95) and number of active medications (OR 1.09, 95% CI 0.99-1.20) are two factors that also have shown to be involved in TTR control in our study.

An interesting finding was recently reported by Bryk et al. [31] and is substantial in countries like Spain where the most widely used VKA is acenocoumarol. By comparing the predictive ability of SAMeTT₂R₂ score in patients with AF treated with warfarin versus acenocoumarol, they detect that it is less effective in predicting unstable anticoagulation with the latter and improves significantly by adding statin use and the presence of COPD (0.66; 95% CI 0.58-0.73 versus 0.56; 0.48-0.64, p = 0.042). This finding could explain at least partially the worse performance of SAMeTT₂R₂ score in countries that use acenocoumarol in opposition to those using warfarin like the population included in the AFFIRM study.

The present study shows that 62.5% of our patients presented TTR <70%. Various study groups over the years have tried to take up the challenge of analysing quality of anticoagulation with varying results. Among Spanish groups, FANTASIIA [8], CALIFA [14], PAULA [30], and ANFGAL [32] studies are retrospective registries that claim that only around 50% of patients achieve a correct TTR. However, the identified factors involved in poor anticoagulation control are diverse and not always consistent.

We identified regular vitamin K intake and BMI as predictors of poor anticoagulation control, being the latter the most consistent as shown by the bootstrapped analysis. During the last several years, multiple studies have attempted to detect the most important clinical, demographic, and even genetic factors capable of predicting anticoagulation control. The abovementioned factors identified in our study are in line with previous findings. In a post hoc analysis of the AMADEUS trial Senoo et al. [33] describes the importance of BMI on the quality of anticoagulation where paradoxically obesity was associated with better anticoagulation control. The reason of such association remains elusive. It has also been suggested that until the completion of further studies, obese patients should be treated with VKA as first option given the weight dependent modifications in drug clearance and the possibility of subtherapeutic drug levels while using NOACs [34].

Regarding the interaction between VKA and vitamin K-rich foods, a systematic review recently published by Violi et al. [35] sustains that the prevailing advice to modify dietary habits in patients with VKAs is not sufficiently supported by up-to-date evidence. In line with other studies [36, 37] they suggest the maintenance of a stable dietary habit or even vitamin K supplementation in order to avoid major changes in the intake of vitamin K. Certainly, vitamin K has already been reported to be an independent predictor of high-quality oral anticoagulation by other authors [38].

Authors have claimed that the clinical prediction tools that are at hand can only explain less than 10% of the variability behind poor anticoagulation control [6], hence genetic variations might play a role in INR stability. Such is the importance of genetic predisposition, that the European Medicines Agency gathers the effect of certain VKORC and CYP2C9 variants on the pharmacokinetics of warfarin and suggests that genotype information may assist dose selection and reduce time to reach target INR [39]. Likewise, the Food and Drug Administration released a communication advising clinicians to consider genetic testing prior initiation of warfarin therapy [40]. Nevertheless, the cost-effectiveness of this approach is debatable. In addition, two validated dosing algorithms (i.e., Gage and IWPC) have been proposed in an updated guideline by the Clinical Pharmacogenetics Implementation Consortium [41].

In this line a very recent meta-analysis of genotype-guided versus standard dosing of VKA [15] reveals that indeed the former can improve TTR in addition to the reduction of risk for bleeding events. However, we must bear in mind that the inclusion of patients that received anticoagulant treatment for at least 6 months might have limited the potential value of genotyping that in fact might be more significant at the start of the therapy. Further, the Chinese group of Liu et al. sought to improve SAMeTT₂R₂ score by replacing “the non-white race” variable with the most representative genotypes, ascending the C-index from 0.60 to 0.67. Therewith, the authors suggest that the modified SAMeTT₂R₂ score could be more appropriate in a racial diverse group of patients [42].

We expanded our genetic analysis to rs4554 MIR133A2 genotyping; unfortunately, we did not find an association between this polymorphism and TTR. In spite of a similar number of analysed patients, the aforementioned exploratory study that examined the effect of variations in MIR133A2 [20] was conducted in a different population to ours (e.g., younger age, absence of concomitant drugs that could alter INR values). In addition, we must take into account the fact that the concept of “one miRNA regulating one gene modifying the efficacy of one drug” might not consider the full extent of miRNA regulation. Nevertheless, future studies in larger samples are warranted. Hence a thorough elaboration of optimized clinical models along with the comprehension of the implication of genetic polymorphisms could play a role for the future of personalized medicine in the field of anticoagulation.

6. Strengths and Limitations

Several limitations of the present study must be noted: first, those regarding the study design such as retrospective character and single centre that limited the evaluation of less common predictors. Another peculiarity of our study is that the predominant VKA in Spain is acenocoumarol, as opposed to other countries, where warfarin is the VKA of choice. Subtle differences in their pharmacokinetics and pharmacodynamics hamper the use of the exact same dosing algorithms. However, our study had strengths. We included a cohort of consecutive patients that were considered adherent to the medication, potentially reducing a risk of bias in this regard. Also, we performed a careful evaluation of previous medical background and explored the usefulness of the MIR133A2 polymorphism, over and above that of VKORC1, CYP2C92, and CYP2C93.

7. Conclusions

In this study, SAMeTT₂R₂ score showed a poor diagnostic performance in the prediction of TTR. We identified body mass index and regular vitamin K intake as factors that could improve SAMeTT₂R₂ score. Finally, although the information provided by the identified genotypes is marginal in our study, the progressive availability of genetic testing could become a promising tool for the future.

Abbreviations and Acronyms

Data Availability

The data that support the findings of this study are available from the corresponding author [Luciano Consuegra-Sánchez] upon reasonable request.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.

Acknowledgments

This research was supported by Fundación para la Formación e Investigación Sanitaria de la Región de Murcia (FFIS) in cooperation with Pfizer.

References

1. S. J. Connolly, J. Pogue, J. Eikelboom et al., “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, vol. 118, no. 20, pp. 2029–2037, 2008. View at: Publisher Site | Google Scholar
2. A. John Camm, G. Y. H. Lip, R. De Caterina et al., “2012 focused update of the ESC Guidelines for the management of atrial fibrillation,” Europace, vol. 15, no. 3, p. 463, 2013. View at: Publisher Site | Google Scholar
3. A. Di Minno, B. Frigerio, G. Spadarella et al., “Old and new oral anticoagulants: Food, herbal medicines and drug interactions,” Blood Reviews, vol. 31, no. 4, pp. 193–203, 2017. View at: Publisher Site | Google Scholar
4. F. R. Rosendaal, S. C. Cannegieter, F. J. M. Van der Meer, and E. Briet, “A method to determine the optimal intensity of oral anticoagulant therapy,” Thrombosis and Haemostasis, vol. 69, no. 3, pp. 236–239, 1993. View at: Publisher Site | Google Scholar
5. A. M. Gallagher, E. Setakis, J. M. Plumb, A. Clemens, and T.-P. van Staa, “Risks of stroke and mortality associated with suboptimal anticoagulation in atrial fibrillation patients,” Thrombosis and Haemostasis, vol. 106, no. 5, pp. 968–977, 2011. View at: Publisher Site | Google Scholar
6. J. R. Schein, C. M. White, W. W. Nelson, J. Kluger, E. S. Mearns, and C. I. Coleman, “Vitamin K antagonist use: Evidence of the difficulty of achieving and maintaining target INR range and subsequent consequences,” Thrombosis Journal, vol. 14, no. 1, 2016. View at: Google Scholar
7. P. Rodríguez Gálvez, L. Valero Verche, S. Pi Ruano, A. Sánchez Martín, and L. Lizán Tudela, “Uso de nuevos anticoagulantes orales (NACO) en la prevención del ictus y la embolia sistémica en pacientes con fibrilación auricular no valvular (FANV),” Àgora de salut, no. 5, pp. 411–420, 2018. View at: Publisher Site | Google Scholar
8. V. Bertomeu-González, M. Anguita, J. Moreno-Arribas et al., “Quality of Anticoagulation with Vitamin K Antagonists,” Clinical Cardiology, vol. 38, no. 6, pp. 357–364, 2015. View at: Publisher Site | Google Scholar
9. F.-E. Cotté, H. Benhaddi, I. Duprat-Lomon et al., “Vitamin K antagonist treatment in patients with atrial fibrillation and time in therapeutic range in four European countries,” Clinical Therapeutics, vol. 36, no. 9, pp. 1160–1168, 2014. View at: Publisher Site | Google Scholar
10. W. L. Baker, D. A. Cios, S. D. Sander, and C. I. Coleman, “Meta-Analysis to Assess the Quality of Warfarin Control in Atrial Fibrillation Patients in the United States,” Journal of Managed Care Pharmacy, vol. 15, no. 3, pp. 244–252, 2009. View at: Publisher Site | Google Scholar
11. S. Apostolakis, R. M. Sullivan, B. Olshansky, and G. Y. H. Lip, “Factors affecting quality of anticoagulation control among patients with atrial fibrillation on warfarin: the SAMe-TT₂R₂ score,” CHEST, vol. 144, no. 5, pp. 1555–1563, 2013. View at: Publisher Site | Google Scholar
12. B. Olshansky, L. Rosenfeld, and A. Warner, “The Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study. Approaches to control rate in atrial fibrillation,” ACC Current Journal Review, vol. 43, no. 7, pp. 1201–1208, 2004. View at: Publisher Site | Google Scholar
13. G. C. S. Smith, S. R. Seaman, A. M. Wood, P. Royston, and I. R. White, “Correcting for optimistic prediction in small data sets,” American Journal of Epidemiology, vol. 180, no. 3, pp. 318–324, 2014. View at: Publisher Site | Google Scholar
14. M. Ruiz-Ortiz, V. Bertomeu, Á. Cequier, F. Marín, and M. Anguita, “Validation of the SAMe-TT2R2 score in a nationwide population of nonvalvular atrial fibrillation patients on vitamin K antagonists,” Thrombosis and Haemostasis, vol. 114, no. 4, pp. 695–701, 2015. View at: Publisher Site | Google Scholar
15. B. Kheiri, A. Abdalla, T. Haykal et al., “Meta-Analysis of Genotype-Guided Versus Standard Dosing of Vitamin K Antagonists,” American Journal of Cardiology, vol. 121, no. 7, pp. 879–887, 2018. View at: Publisher Site | Google Scholar
16. A. I. Anton, J. J. Cerezo-Manchado, J. Padilla et al., “Novel Associations of VKORC1 Variants with Higher Acenocoumarol Requirements,” PLoS ONE, vol. 8, no. 5, pp. 1–7, 2013. View at: Google Scholar
17. J. Wolkanin-Bartnik, H. Pogorzelska, M. Szperl, A. Bartnik, J. Koziarek, and Z. T. Bilinska, “Impact of genetic and clinical factors on dose requirements and quality of anticoagulation therapy in Polish patients receiving acenocoumarol: Dosing calculation algorithm,” Pharmacogenetics and Genomics, vol. 23, no. 11, pp. 611–618, 2013. View at: Publisher Site | Google Scholar
18. R. Nahar, R. Saxena, R. Deb et al., “CYP2C9, VKORC1, CYP4F2, ABCB1 and F5 variants: Influence on quality of long-term anticoagulation,” Pharmacological Reports, vol. 66, no. 2, pp. 243–249, 2014. View at: Publisher Site | Google Scholar
19. E. Jiménez-Varo, M. Cañadas-Garre, C. I. Henriques, A. M. Pinheiro, M. J. Gutiérrez-Pimentel, and M. Á. Calleja-Hernández, “Pharmacogenetics role in the safety of acenocoumarol therapy,” Thrombosis and Haemostasis, vol. 112, no. 3, pp. 522–536, 2014. View at: Publisher Site | Google Scholar
20. C. Ciccacci, S. Rufini, C. Politi, G. Novelli, V. Forte, and P. Borgiani, “Could MicroRNA polymorphisms influence warfarin dosing? A pharmacogenetics study on mir133 genes,” Thrombosis Research, vol. 136, no. 2, pp. 367–370, 2015. View at: Publisher Site | Google Scholar
21. D. E. Morisky, L. W. Green, and D. M. Levine, “Concurrent and predictive validity of a self-reported measure of medication adherence,” Medical Care, vol. 24, no. 1, pp. 67–74, 1986. View at: Publisher Site | Google Scholar
22. M. E. Charlson, P. Pompei, K. L. Ales, and C. R. MacKenzie, “A new method of classifying prognostic comorbidity in longitudinal studies: development and validation,” Journal of Chronic Diseases, vol. 40, no. 5, pp. 373–383, 1987. View at: Publisher Site | Google Scholar
23. P. Kirchhof, S. Benussi, D. Kotecha et al., “2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS,” European Heart Journal, vol. 37, no. 38, pp. 2893–2962, 2016. View at: Google Scholar
24. V. Sjögren, B. Grzymala-Lubanski, H. Renlund et al., “Safety and efficacy of well managed warfarin,” Thrombosis and Haemostasis, vol. 113, no. 06, pp. 1370–1377, 2017. View at: Publisher Site | Google Scholar
25. J. Skov, E.-M. Bladbjerg, M. V. Bor, and J. Gram, “SAMeTT2R2 does not predict time in therapeutic range of the international normalized ratio in patients attending a high-quality anticoagulation clinic,” CHEST, vol. 145, no. 1, pp. 187-188, 2014. View at: Publisher Site | Google Scholar
26. A. Garcia-Perez, F. Leiva-Fernandez, F. Martos-Crespo, A. Garcia-Ruiz, D. Prados- Torres, and F. Alarcon, “¿Cómo diagnosticar el cumplimiento terapéutico en atención primaria?” Med Fam, vol. 1, no. 1, pp. 13–19, 2000. View at: Google Scholar
27. R. R. Abumuaileq, E. Abu-Assi, S. Raposeiras-Roubin et al., “Evaluation of SAMe-TT2R2 risk score for predicting the quality of anticoagulation control in a real-world cohort of patients with non-valvular atrial fibrillation on vitamin-K antagonists,” Europace, vol. 17, no. 5, pp. 711–717, 2015. View at: Publisher Site | Google Scholar
28. A. J. Rose, E. M. Hylek, A. Ozonoff, A. S. Ash, J. I. Reisman, and D. R. Berlowitz, “Patient characteristics associated with oral anticoagulation control: Results of the Veterans AffaiRs Study to Improve Anticoagulation (VARIA),” Journal of Thrombosis and Haemostasis, vol. 8, no. 10, pp. 2182–2191, 2010. View at: Publisher Site | Google Scholar
29. J. M. Lobos-Bejarano, V. Barrios, J. Polo-García et al., “Evaluation of SAMe-TT2R2 score and other clinical factors influencing the quality of anticoagulation therapy in non-valvular atrial fibrillation: A nationwide study in Spain,” Current Medical Research and Opinion, vol. 32, no. 7, pp. 1201–1207, 2016. View at: Publisher Site | Google Scholar
30. V. Barrios, C. Escobar, L. Prieto et al., “Control de la anticoagulación en pacientes con fibrilación auricular no valvular asistidos en atención primaria en España. Estudio PAULA,” Revista Española de Cardiología, vol. 68, no. 9, pp. 769–776, 2015. View at: Publisher Site | Google Scholar
31. A. H. Bryk, K. Plens, and A. Undas, “Prediction of unstable anticoagulation with acenocoumarol versus warfarin in atrial fibrillation,” Cardiology Journal, vol. 24, no. 5, pp. 477–483, 2017. View at: Publisher Site | Google Scholar
32. S. Cinza-Sanjurjo, D. Rey-Aldana, E. Gestal-Pereira, and C. Calvo-Gómez, “Evaluación del grado de anticoagulación de pacientes con fibrilación auricular en el ámbito de atención primaria de Galicia. Estudio ANFAGAL,” Revista Española de Cardiología, vol. 68, no. 9, pp. 753–760, 2015. View at: Publisher Site | Google Scholar
33. K. Senoo and G. Y. H. Lip, “Body Mass Index and Adverse Outcomes in Elderly Patients with Atrial Fibrillation,” Stroke, vol. 47, no. 2, pp. 523–526, 2016. View at: Publisher Site | Google Scholar
34. KK. Graves, K. Edholm, and SA. Johnson, “Use of oral anticoagulants in obese patients,” JSM Atheroscler, vol. 2, no. 4, 2017. View at: Google Scholar
35. F. Violi, G. Y. H. Lip, P. Pignatelli, and D. Pastori, “Interaction between dietary Vitamin K intake and anticoagulation by Vitamin K antagonists: is it really true?: A systematic review,” Medicine (United States), vol. 95, no. 10, pp. 1–7, 2016. View at: Google Scholar
36. E. Sconce, T. Khan, J. Mason, F. Noble, H. Wynne, and F. Kamali, “Patients with unstable control have a poorer dietary intake of vitamin K compared to patients with stable control of anticoagulation,” Thrombosis and Haemostasis, vol. 93, no. 5, pp. 872–875, 2005. View at: Publisher Site | Google Scholar
37. E. K. Rombouts, F. R. Rosendaal, and F. J. M. Van Der Meer, “Daily vitamin K supplementation improves anticoagulant stability,” Journal of Thrombosis and Haemostasis, vol. 5, no. 10, pp. 2043–2048, 2007. View at: Publisher Site | Google Scholar
38. G. L. D. B. Costa, R. M. Lamego, E. A. Colosimo, R. A. Valacio, and M. D. C. V. Moreira, “Identifying Potential Predictors of High-Quality Oral Anticoagulation Assessed by Time in Therapeutic International Normalized Ratio Range: A Prospective, Long-Term, Single-Center, Observational Study,” Clinical Therapeutics, vol. 34, no. 7, pp. 1511–1520, 2012. View at: Publisher Site | Google Scholar
39. EMA/281371/2013, “Guideline on key aspects for the use of pharmacogenomic methodologies in the pharmacovigilance evaluation of medicinal products Guideline on key aspects for the use of pharmacogenomic methodologies in the pharmacovigilance evaluation of medicinal products,” Ema/281371/2013, vol. 44, pp. 1–18, 2014. View at: Google Scholar
40. FDA, “Approves Updated Warfarin FDA (Coumadin) Prescribing information: New genetic information may help providers improve initial dosing estimates of the anticoagulant for individual patient,” 2009, http://www.fda.gov/bbs/topics/NEWS/2007/NEW01684.html. View at: Google Scholar
41. J. Johnson, K. Caudle, L. Gong et al., “Clinical Pharmacogenetics Implementation Consortium (CPIC) Guideline for Pharmacogenetics-Guided Warfarin Dosing: 2017 Update,” Clinical Pharmacology & Therapeutics, vol. 102, no. 3, pp. 397–404, 2017. View at: Publisher Site | Google Scholar
42. J. Liu, N. Wang, L. Qin et al., “Integrating genotypes in the SAMe-TT2R2 score for the prediction of anticoagulation control in Chinese patients with atrial fibrillation on warfarin,” International Journal of Cardiology, vol. 241, pp. 358–363, 2017. View at: Publisher Site | Google Scholar

Copyright

Copyright © 2018 Samantha Wasniewski et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.