Stroke prevention in patients from Latin American countries with non‐valvular atrial fibrillation: Insights from the GARFIELD‐AF registry

Background Atrial fibrillation (AF) is an important preventable cause of stroke. Anticoagulation (AC) therapy can reduce this risk. However, prescribing patterns and outcomes in patients with non‐valvular AF (NVAF) from Latin American countries are poorly described. Methods Using data from the Global Anticoagulant Registry in the FIELD‐AF (GARFIELD‐AF), we examined the stroke prevention strategies and the 1‐year outcomes in patients from four Latin American countries: Argentina, Brazil, Chile, and Mexico. Results A total of 4162 patients (2010‐2014) were included in this analysis. At the time of AF diagnosis, 39.9% of patients were prescribed vitamin K antagonists (VKA) ± antiplatelet (AP) therapy, 21.8% non‐VKA oral anticoagulant (NOAC) ± AP, 24.1% AP only and 14.1% no antithrombotic treatment. The proportion of moderate‐high risk patients receiving no AC therapy at participating centers was highest in Mexico (46.4%) and lowest in Chile (14.3%). During 1‐year follow‐up, the rates of all‐cause mortality, stroke/SE and major bleeding were: 5.77 (95% CI) (5.06‐6.56), 1.58 (1.23‐2.02), and 0.99 (0.72‐1.36) and per 100 person‐years, respectively, which are higher than the global rates across all countries in GARFIELD‐AF. Unadjusted rates of all‐cause mortality were highest in Argentina, 6.95 (5.43‐8.90), and lowest in Chile, 4.01 (2.92‐5.52). Conclusions GARFIELD‐AF results describes the marked variation in the baseline characteristics and patterns of antithrombotic treatments in patients with NVAF in four Latin American countries. Over one‐third of patients with a moderate‐to‐high risk of stroke received no AC therapy, highlighting the need for improved management of patients according to national guideline. Clinical Trial Registration—URL http://www.clinicaltrials.gov. Unique identifier: NCT01090362.

majority of cases. 1 The prevalence of AF is increasing in both developed and developing countries, owing to an aging population, and this increase is predicted to continue over the coming decades. 2 Patients with AF have a 5-fold greater risk of stroke and frequently present in elderly patients. Common risk factors for stroke are hypertension, diabetes, heart failure, smoking, and prior stroke or transient ischemic attack (TIA). 3,4 In addition to an increased risk of stroke, AF patients frequently present with comorbid myocardial infarction, dementia, and chronic kidney disease. 5,6 The Global Anticoagulant Registry in the FIELD-AF (GARFIELD-AF) is an ongoing prospective, international, multicentre registry of adult patients with newly diagnosed NVAF and one or more additional risk factors for stroke. 7 Since December 2009, more than 50 000 patients have been enrolled in the registry from 35 countries and patient follow-up is anticipated to end in the third quarter of 2018.
Major goals of the registry are to identify best practices as well as deficiencies in stroke prevention strategies for AF patients and to describe how patient care has evolved over time. As in all registries, there might be substantial regional and intraregional differences among baseline characteristics and use of antithrombotic therapies in patients with new NVAF. 8 In 2010, the Global Burden of Disease Study estimated that ageadjusted prevalence of AF in Latin America was 737.9 per 100 000 men and 440.3 per 100 000 women, which is higher than the global average estimated to be 596 per 100 000 men and 373 per 100 000 women. 9 It is likely that the true prevalence of AF, in general, may be even higher due to asymptomatic or unrecognized AF, which is estimated to account for up to 27% undiagnosed AF patients. 10 In this paper, we analyzed the baseline characteristics, patterns of antithrombotic therapies and 1-year outcomes in four Latin American countries, Argentina, Brazil, Chile, and Mexico that participated in the GARFIELD-AF.

| Study design and participants
GARFIELD-AF is a non-interventional, observational, worldwide study of NVAF, as described in detail previously. 7 Patients (≥18 years) were diagnosed with AF according to standard local procedures within the previous 6 weeks and had at least 1 additional factor(s) for stroke as judged by the study investigator. Risk factors were not pre-specified in the protocol nor were they limited to the components of existing risk stratification schemes. The study excluded patients with a transient, reversible cause of NVAF, and patients for whom follow-up to 2 years was not envisaged or possible. 7 Consecutive patients were enrolled prospectively into five sequential cohorts (plus one retrospective cohort of 5000 patients). Investigator sites were randomly selected 11 and were representative of the care settings in each country.

| Data collected at baseline
Baseline data from the patients were collected at the time of diagnosis and included the type of AF, patient demographics, medical history, cardiovascular risk factors, care setting speciality and location, antithrombotic therapy regimen in treated patients, and the main reasons for not providing anticoagulant treatment in untreated patients.  12,13 Case report forms (CRFs) were submitted to the registrycoordinating center (Dendrite Clinical Systems Ltd, Henley-on-Thames, UK), and the corresponding data were analyzed by an independent statistician. All CRFs were examined by the coordinating center to ascertain completeness and accuracy, and data queries were sent to participating sites. The data used in the study were extracted from the database on October 18, 2017.
Vascular disease is defined as peripheral artery disease and/or coronary artery disease with a history of acute coronary syndrome (ACS). 14 Moderate-to-severe chronic kidney disease (CKD) includes stage III to stage V according to the National Kidney Foundation's Kidney Disease Outcomes Quality Initiative guidelines. 15

| Statistical analysis
In this analysis, descriptive summaries of patient baseline characteristics were performed for each country, and for all countries taken together. Continuous variables are presented with mean and SDs, and numbers of non-missing observations are included in the tables and figures. Categorical variables are presented using frequencies and percentages. Baseline differences between countries were evaluated using χ 2 tests for categorical variables, and Students t-test for continuous variables. Percentages are rounded to one decimal place.
Occurrences of rate mortality are described using the person-time event rate (per 100 person-years) and 95% CI. We estimated personyear rates using a Poisson model, with the number of events as the dependent variable and the log of time as an offset, that is, a covariate with a known coefficient of 1. A log-rank test was used to evaluate whether at least one country varied in mortality rates compared to the other countries. Stroke/SE and major bleed events were too few for comparison between countries. Due to the large sample size, small differences in results can be statistically significant, thus, clinically important differences are also considered. Data analysis was performed with SAS statistical software, release 9.4 (SAS Institute Inc., Cary, North Carolina).

| INRs and time in therapeutic range
Patients receiving VKA therapy at enrolment with ≥3 INR readings and for whom time in therapeutic range (TTR) could be calculated were included in the analysis. 16

| Antithrombotic treatment for stroke prevention
Upon diagnosis of AF, 39.9% of patients from all four countries were prescribed VKAs ± AP therapy, 21.8% NOACs ± AP therapy and 24.1% AP alone. 14.1% of patients received no antithrombotic treatment (Table 2).
Notable differences in patterns of antithrombotic treatment were found between the countries. The use of NOAC ± AP was higher in Mexico (28.8%) and Brazil (25.8%) than in Argentina (19.3%) and Chile (12.3%). The choice of NOAC prescribed also varied between the countries ( Table 2). The proportion of patients who did not receive any antithrombotic treatment was highest in Brazil (19.2%) and Argentina (17.8%) followed by Mexico (12.0%) and Chile (7.6%).

| Reasons for anticoagulant therapy was not used
The main reason for not giving an AC to patients at moderate-to-high risk of stroke varied between countries but was most frequently the physician's choice (53.8%) (Table S1, Supporting Information).

| DISCUSSION
In this paper, we describe the baseline characteristics, antithrombotic treatment patterns, quality of VKA control and event rates for major clinical outcomes in AF patients from four countries in Latin America, who were enrolled in the GARFIELD-AF registry.
Even though patients were from the same region, the baseline characteristics and comorbidities were remarkably different for patients from each country. Patients from Chile, for example, were typically older than other in this countries analysis, with a higher incidence of obesity, although these patients had a lower incidence of vascular disease. Interestingly, a history of stroke and TIA was more than twice as frequent in Mexico (16.5%) compared with patients from Argentina (7.4%).
At the time of AF diagnosis, there were also major differences in It has been demonstrated that the benefit of AC therapy significantly outweighs the risk of bleeding for AF patients with a CHADS 2 or CHA 2 DS 2 -VASc score of ≥2. [21][22][23] In GARFIELD-AF, over four-fifths    increase in coming years, given their favorable safety and efficacy profile, as guidelines are updated. 33 The new Brazilian Guideline from 2016 34 and Mexican Guidelines, 23 for example, indicates the use of NOACs in a similar manner to the European guidelines. This has also been stressed by different groups of Latin American investigators. 35,36 All-cause mortality was the most frequent major clinical outcome, nearly 4-fold higher than the rate of stroke/SE, and 6-fold higher than the rate of major bleeding. Mortality rates from the combined experience: Argentina, Brazil, Chile, and Mexico are higher to those reported in the entire GARFIELD-AF registry. 37 However, these numbers are consistent with other reports from Latin America, derived from clinical trials, such as ENGAGE-AF and ROCKET-AF. 27,38 In addition to the poor management of antithrombotic therapies, it is possible that the reasons for the high mortality rates of Latin American patients with AF are related to a higher rate of comorbidities or sociocultural prob-

| LIMITATIONS
This registry is limited to patients with newly diagnosed AF and the study was mainly conducted by the cardiologists. As with all registries, there may be a bias in the selection of patients and medical centers and so the results may not reflect the experience in all centers in these countries.

| CONCLUSIONS
This paper describes the baseline characteristics and patterns of antithrombotic treatment in patients from four Latin American countries, Argentina, Brazil, Chile, and Mexico. Over one-third of patients with a moderate-to-high risk of stroke received no AC therapy, highlighting the need for better adherence to evidence-based guidelines on stroke prevention in AF.