Journal of the American College of Cardiology Vol. 44, No. 8, by the American College of Cardiology Foundation ISSN /04/$30.

Transcription

1 Journal of the American College of Cardiology Vol. 44, No. 8, by the American College of Cardiology Foundation ISSN /04/$30.00 Published by Elsevier Inc. doi: /j.jacc CLINICAL RESEARCH Comparative Effects of Antiplatelet, Anticoagulant, or Combined Therapy in Patients With Valvular and Nonvalvular Atrial Fibrillation A Randomized Multicenter Study Francisco Pérez-Gómez, PHD, FESC,* Eduardo Alegría, MD, PHD, Jesus Berjón, MD, PHD, Jose A. Iriarte, MD, PHD, Javier Zumalde, MD, Antonio Salvador, MD, PHD, Luis Mataix, MD,# for the NASPEAF Investigators Madrid, Pamplona, Bilbao, and Valencia, Spain Clinical Trials OBJECTIVES BACKGROUND METHODS RESULTS CONCLUSIONS This trial evaluated the efficacy and safety of the combination of antiplatelet and moderateintensity anticoagulation therapy in patients with atrial fibrillation associated with recognized risk factors or mitral stenosis. Warfarin was more effective than aspirin in preventing stroke in these patients; combined therapy with low anticoagulant intensity was ineffective. Mitral stenosis patients were not investigated. We performed a multicenter randomized trial in 1,209 patients at risk. The intermediate-risk group included patients with risk factors or age 60 years: 242 received the cyclooxygenase inhibitor triflusal, 237 received acenocumarol, and 235 received a combination of both. The high-risk group included patients with prior embolism or mitral stenosis: 259 received anticoagulants and 236 received the combined therapy. Median follow-up was 2.76 years. Primary outcome was a composite of vascular death and nonfatal stroke or systemic embolism. Primary outcome was lower in the combined therapy than in the anticoagulant arm in both the intermediate- (hazard ratio [HR] 0.33 [95% confidence interval (CI) 0.12 to 0.91]; p 0.02) and the high-risk group (HR 0.51 [95% CI 0.27 to 0.96]; p 0.03). Primary outcome plus severe bleeding was lower with combined therapy in the intermediate-risk group. Nonvalvular and mitral stenosis patients had similar embolic event rates during anticoagulant therapy. The combined antiplatelet plus moderate-intensity anticoagulation therapy significantly decreased the vascular events compared with anticoagulation alone and proved to be safe in atrial fibrillation patients. (J Am Coll Cardiol 2004;44: ) 2004 by the American College of Cardiology Foundation Atrial fibrillation (AF) is the most prevalent cardiac arrhythmia and is also a strong independent risk factor for both systemic embolism and stroke (1,2). Since 1989, several randomized studies on stroke prevention (3) have tested two antithrombotic agents: warfarin and aspirin. See page 1567 Some meta-analyses (4 6) showed that warfarin decreases stroke by 62% compared with placebo. In studies where a statistical benefit with aspirin has been shown, this benefit is much less than seen with warfarin, with fewer bleeding complications and easier monitoring (4). In order to reduce From the *Hospital Clínico San Carlos, Madrid, Spain; Clínica Universitaria de Navarra, Pamplona, Spain; Hospital de Navarra, Pamplona, Spain; Hospital de Basurto, Bilbao, Spain; Hospital Galdakao, Bilbao, Spain; Hospital Dr. Peset, Valencia, Spain; and #Centro Especialidades Avenida de Portugal, Madrid, Spain. The secretariat, data bank, statistical analysis, and expenses for working meetings were supported by grants from the Spanish Society of Cardiology (Madrid) and Uriach Foundation (Barcelona). Manuscript received February 18, 2004; revised manuscript received May 3, 2004, accepted May 11, both thromboembolic and bleeding events, the combination of low-dose warfarin and aspirin was investigated in three trials (7 9). All concluded that this combination therapy with an international normalized ratio (INR) below 1.5 was ineffective. Theoretically, the combination of therapeutic doses of anticoagulant and antiplatelet agents (i.e., combined therapy) should have additional benefits and, recently, the Warfarin Aspirin Re-Infarction Study (WARIS) II found it superior to aspirin alone in the prevention of vascular events after myocardial infarction (MI) (10). Our trial has evaluated, for the first time, the efficacy and safety of this combination compared with one antiplatelet or anticoagulation alone in nonvalvular AF and in patients with mitral stenosis. PATIENTS AND METHODS Eligible patients. Patients with chronic or documented paroxysmal AF were eligible for the study. Patients at low risk according to Stroke Prevention in Atrial Fibrillation (SPAF) III stratification (7) or younger than 60 years of age

2 1558 Pérez-Gómez et al. JACC Vol. 44, No. 8, 2004 Antithrombotic Therapy in Atrial Fibrillation October 19, 2004: Abbreviations and Acronyms AF atrial fibrillation AFASAK Copenhagen Study on Atrial Fibrillation CI confidence interval EAFT European Atrial Fibrillation Trial HR hazard ratio INR international normalized ratio MI myocardial infarction NASPEAF National Study for Prevention of Embolism in Atrial Fibrillation SPAF Stroke Prevention in Atrial Fibrillation TIA transient ischemic attack WARIS Warfarin Aspirin Re-Infarction Study were not included. Eligible patients were divided in two groups: the high-risk group included nonvalvular plus prior embolism and patients with mitral stenosis with and without prior embolism. All others were included in the intermediate-risk group. Exclusion criteria were mechanical valve prosthesis, stroke in the previous six months, serum creatinine over 3 mg/dl, alcoholism or drug addiction, severe uncontrolled hypertension, diffuse arteriosclerosis, and indication for non-steroidal anti-inflammatory drugs or indication/contraindication for antiplatelet or anticoagulant therapy. Study design and organization. The National Study for Prevention of Embolism in Atrial Fibrillation (NASPEAF) is a prospective, multicenter, randomized open-label study. Thirteen Spanish hospitals participated according to the principles of the Helsinki Declaration. The ethical committee of each institution approved the protocol and the patient s written informed consent was signed. Two blind events and safety committees, composed of members not involved in the running of the trial, validated all events and monitored the safety of treatments. The core center for data collection and statistical analysis was San Carlos University Hospital in Madrid. The Working Group on Thrombosis of the Spanish Cardiac Society had access to the database throughout the study. The antiplatelet agent used was triflusal (Grupo Uriach, Spain), a drug structurally related to acetylsalicylic acid (11), of which clinical trials showed that 600 mg/day has similar biologic effects and clinical efficacy to aspirin 300 mg/day with fewer bleeding complications (12,13). The anticoagulant used was acenocoumarol (Novartis Farmaceutica, Spain), the coumarin derivative most used in several European countries. Randomization was balanced, computer-generated, and administered centrally. It was also balanced by study center and by the three high-risk subgroups. The randomization sequence could not be previewed. Patients in the intermediate-risk group were randomized to one of the three arms: oral anticoagulation to a target INR of 2 to 3, triflusal 600 mg daily, or a combination of both with a target INR of 1.25 to 2. In the high-risk group, the triflusal-only arm was omitted and subjects were assigned to anticoagulation with a target INR of 2 to 3 or the combination therapy with a target INR of 1.4 to 2.4. Demographic data, risk factors, concomitant heart disease, blood pressure, clinical examination, electrocardiogram, and echocardiogram were recorded at baseline. Clinical follow-up was scheduled every six months for a maximum of four years. In between, the patients were under the care of their general practitioners. Follow-up was interrupted after a primary outcome or a prosthetic valve implantation. Any possible way (hospital records, phone calls) was used to detect any new event in cases lost to follow-up. This information was registered in a final form. Anticoagulation was controlled at specialized units by recording mean treatment dose, mean INR, time within the preset range, number of INR controls out of the preset range, number of INR controls 3.5 and 2, and the intercontrol intervals. Outcome events. Primary outcome was a composite of vascular death, transient ischemic attack (TIA), and nonfatal stroke or systemic embolism, whichever came first. We also prospectively registered as secondary events severe bleeding, MI, nonvascular death, and non-severe bleeding. These outcomes alone or in combination were also analyzed. The composite of primary outcomes and severe bleeding was jointly analyzed to evaluate the benefit-torisk ratio. Stroke and TIA were defined as focal neurologic deficits lasting 24hor 24 h, respectively. Neuroimaging defined the ischemic or intracranial hemorrhagic etiology. Systemic embolism was diagnosed after an abrupt vascular insufficiency without previous clinical symptoms. Vascular death included either sudden or any other death occurring within 30 days after a vascular event or progressive heart failure. Bleeding was considered severe when requiring hospital admission, blood transfusion, or surgery (14). Statistics. For sample size calculation in the intermediaterisk group, we used the aspirin and anticoagulation arms events rate of the SPAF II younger group (15) and Copenhagen Study on Atrial Fibrillation (AFASAK) I (3) trials. The composite of nonfatal stroke, TIA, systemic embolism, and vascular death in the aspirin arm was 4.72 per 100 person-years in the SPAF II study and 4.8 in the AFASAK I trial. The rates in the anticoagulation arm were 3.53 and 1.2, respectively. The anticoagulation arm event rate in the high-risk group was based on the European Atrial Fibrillation Trial (EAFT) (16), where the primary event rate was 8.0. There were no published trials to assume event rate in the combined arms, because the anticoagulation intensity used resulted subtherapeutic, but Turpie et al. (17), using high anticoagulation intensity associated with aspirin 100 mg/day for embolic prevention in patients with prosthetic heart valves, showed a benefit of 61%. We hypothesized that, using a moderate level of anticoagulation in our combined arms, a 40% reduction in the event rate should be clinically relevant. Assuming these figures, in a four-year follow-up period, two-tailed alpha-error of 0.05, and a

3 JACC Vol. 44, No. 8, 2004 October 19, 2004: Pérez-Gómez et al. Antithrombotic Therapy in Atrial Fibrillation 1559 Figure 1. Flow diagram of the NASPEAF. Lost to follow-up corresponds to those patients about whom no data about efficacy and safety is available. INR international normalized ratio; ITT intention to treat; IQR interquartile range. power of 80%, the calculated cohort was 813 for each treatment arm of the intermediate- and 197 for the highrisk group. The safety committee decided to review outcomes after the SPAF III study reported in September 1996 (7) the failure of the combined therapy arm. This therapy in our trial presented no greater number of events than the other arms, and the median INR in the combined arms of both intermediate- and high-risk groups was 1.93 and 2.08, respectively. A similar level of INR was maintained afterward. In February 2000, when the calculated sample for the high-risk group was complete, the events committee analyzed the outcomes, which were significantly lower in the combined arm. The committee then decided to also analyze the intermediate-risk group, and the combined therapy also showed a significant benefit compared with other arms. Therefore, it was decided to finish recruitment in this group and continue the follow-up for one more year of cases that had not completed the scheduled follow-up period. Statistical analyses were performed according to the intention-to-treat and on-treatment principles. Baseline comparisons were performed using the chi-square test for categorical data and analysis of variance test for continuous data. A logistic multivariate analysis was made to adjust baseline comparisons. The incidences were expressed as rates (events per 100 person-years). We used the Kaplan- Meier method to calculate the four-year cumulative rate of outcome events until the first event occurred. The representation of these charts was cut at the median follow-up. We used the hazard ratio (HR), 95% confidence interval (CI) of a Cox regression model, and the log-rank statistics to express the efficacy and safety of the different treatments. All tests were carried out at a 0.05 level of significance. Statistical analysis was performed using SPSS (Version 8.0, SPSS Inc., Chicago, Illinois). RESULTS Recruitment started in June 1995 and stopped in February The study was closed in June The number of patients recruited was 1,209. Figure 1 shows their distribution: 714 entered the intermediate-risk group (median follow-up 965 days) and 495 entered the high-risk group (median follow-up 1,075 days). The following reasons for treatment withdrawals were registered: adverse events (9%), without significant difference among groups; and general practitioner s or patient s decision (9.3%). These withdrawals resulted in a change from the combined treatment to the anticoagulant alone in 56 patients. The treatment arms were balanced with respect to baseline characteristics (Table 1). Patients with prior embolism at baseline presented a significantly higher rate of female gender, hypertension, diabetes, ischemic heart disease, hyperlipidemia, and peripheral vascular disease; had a larger left atrium; and were older. However, multivariate analysis showed significant differences only for the last four of these factors. Table 2 shows the anticoagulation control data. The combined therapy had a lower median INR than the corresponding anticoagulant arms (1.93 vs and 2.17 vs.

4 1560 Pérez-Gómez et al. JACC Vol. 44, No. 8, 2004 Antithrombotic Therapy in Atrial Fibrillation October 19, 2004: Table 1. Baseline Characteristics of the Study Participants Intermediate-Risk Group High-Risk Group Triflusal (n 242) Anticoagulant (n 237) Combined (n 235) Anticoagulant (n 259) Combined (n 236) Age, yrs (SD) 69.9 (8) 69.6 (7) 69.8 (7) 66.6 (9) 67.3 (10) No risk factors, age yrs Men Current smokers Diabetes Hyperlipidaemia Family history of thrombotic disease 20.4* Hypertension Mean SBP, mm Hg (SD) 138 (17) 135 (18) 136 (17) 136 (19) 134 (18) Mean DBP, mm Hg (SD) 80 (11) 79 (11) 79 (10) 77 (10) 77 (10) Ischemic heart disease Peripheral vascular disease History of heart failure * NYHA functional class II to IV LV ejection fraction (SD) 61.0 (11) 60.0 (12) 60.0 (10) 63.7 (10) 62.1 (10) LA diameter, mm (SD) 46.2 (8) 47.4 (8) 47.0 (8) 52.6 (9) 51.7 (10) Persistent/permanent atrial fibrillation Bioprosthesis Nonvalvular AF prior embolism (n) Mitral stenosis prior embolism (n) Mitral stenosis no prior embolism (n) Data expressed in percentage, unless otherwise indicated. *p 0.05 compared with the other two groups. Bioprosthesis normal functioning are included in intermediate-risk group; if mitral valve area is narrower than 2 cm 2 they are included in the high-risk group. DBP diastolic blood pressure; LA left atrium; LV left ventricle; NYHA New York Heart Association; SBP systolic blood pressure in the intermediate- and high-risk groups, respectively). The combined arms also had a higher number of INR controls 2 in either risk group (67% vs. 21% and 39% vs. 18%) and a lower number of INR controls 3.5 (1.96% vs. 6.30% and 2.40% vs. 7.90%) (p for all). Outcomes. Primary events in the intermediate-risk group were recorded in 42 cases (Table 3). Figure 2A shows the event-free survival curves and Figure 3 shows the HR. The combined therapy arm presented fewer primary events than the antiplatelet (HR 0.24 [95% CI 0.09 to 0.64]; p 0.001) or the anticoagulant-alone group (HR 0.33 [95% CI 0.12 to 0.91]; p 0.02). No significant difference in primary events was observed in the anticoagulant compared with the antiplatelet arm (HR 0.72 [95% CI 0.37 to 1.39], p 0.32). The combined therapy arm had a lower rate of embolismstroke-tia than the antiplatelet arm (HR 0.21 [95% CI 0.06 to 0.74]; p 0.01) and lower vascular death rate than the anticoagulant arm (0.37 vs. 1.98, p 0.01). It also showed a 61% relative reduction in outcome plus severe bleeding compared with either the antiplatelet (HR 0.39 [95% CI 0.17 to 0.87], p 0.02) or the anticoagulant (HR 0.38 [95% CI 0.17 to 0.87]; p 0.02) arms. The ontreatment primary outcome rates were 2.95, 2.34, and 0.55 in the antiplatelet, anticoagulant, and combined therapy arms, respectively, with a significant benefit in the combined therapy when compared with either the antiplatelet (p 0.002) or the anticoagulant arms (p 0.005). There were 43 primary events in the high-risk group. Table 3 shows their distribution and Figures 2B and 3 show the primary event-free survival curves and the HR. The Table 2. Anticoagulation Values in Different Therapeutic Groups Intermediate-Risk Group High-Risk Group Anticoagulant Combined Anticoagulant Combined Number of INR controls 6,813 5,478 7,421 5,833 Preset INR range Mean anticoagulant dose in mg (SD) 2.04 (0.80) 1.61 (0.70)* 2.04 (0.80) 1.63 (0.70)* Median INR * * (IQR: P 25 P 75 ) ( ) ( ) ( ) ( ) Time within range, % (SD) 65 (22) 66 (25) 67 (22) 73 (22) Time over range, % (SD) 16 (17) 31 (25) 15 (16) 23 (21) Time below range, % (SD) 19 (18) 3 (9) 18 (17) 4 (9) Mean inter-controls interval, days Number INRs 2, % 21 67* 18 39* Number INRs 3.5, % * * *p compared with the anticoagulant arm. INR international normalized ratio.

5 JACC Vol. 44, No. 8, 2004 October 19, 2004: Pérez-Gómez et al. Antithrombotic Therapy in Atrial Fibrillation 1561 Table 3. Number of Events in Each Treatment Group at Follow-Up Intermediate-Risk Group High-Risk Group Triflusal (n 235) Anticoagulant (n 232) Combined (n 222) Anticoagulant (n 247) Combined (n 223) Patient-yrs Primary outcome 22 (3.82) 15 (2.70) 5 (0.92)* 29 (4.76) 14 (2.44) * Primary outcome ( OT ) Vascular death 8 (1.39) 11 (1.98) 2 (0.37) 17 (2.79) 6 (1.05)* Systemic embolism Stroke Bleeding Sudden Heart failure HF due to avascular event Myocardial infarction Nonfatal stroke TIA Nonfatal systemic embolism Nonfatal MI Embolism stroke and TIA 15 (2.60) 7 (1.26) 3 (0.55) 20 (3.28) 12 (2.09) Fatal embolism-stroke-tia 1 (0.17) 3 (0.54) 0 (0) 8 (1.31) 4 (0.70) Embolism, stroke, AMI, and death from 16 (2.78) 8 (1.44) 4 (0.74) 25 (4.10) 13 (2.27) vascular events Severe bleeding 2 (0.35) 10 (1.80) 5 (0.92) 13 (2.13) 12 (2.09) Intracranial (intracerebral) 2 (2) 4 (1) 1 (1) 5 (3) 2 (0) Gastric Other Outcome severe bleeding 22 (3.82) 21 (3.78) 8 (1.48)* 34 (5.58) 22 (3.84) Nonvascular death Nonsevere bleeding p values of the combined therapy compared with the other arms: *p 0.05; p 0.01; p Heart failure due to repeated ischemic events. Two deaths from heart failure due to repeated gastrointestinal bleeding. AMI acute myocardial infarction; HF heart failure; MI myocardial infarction; OT on treatment; TIA transient ischemic attack; ( ) rate per 100 person-years anticoagulant arm presented a primary outcome rate of 4.76 and the combined therapy arm decreased this rate by 49% (HR 0.51 [95% CI 0.27 to 0.96]; p 0.03). The ontreatment primary outcome rate difference was higher (1.1 vs. 4.38) (p 0.001); eight additional primary or severe bleeding events occurred after withdrawal of the combined therapy: four during anticoagulant and the other four during antiplatelet therapy alone. The combined therapy, as compared with the anticoagulant therapy, decreased the aggregate embolism, stroke, and TIA rate by 56% in both the intermediate- and high-risk groups (Fig. 4A). Figure 4B shows the embolism-stroke-tia event-free survival curves throughout anticoagulant therapy in the three subgroups of the high-risk group and in nonvalvular AF patients without prior embolism. The survival curves were similar in nonvalvular and valvular patients either with prior embolism at baseline (rates of 4.94 vs. 4.81, p 0.98) or without embolism (1.43 vs. 1.26, p 0.57). However, patients with mitral stenosis had a significantly lower incidence of several other risk factors: age (p 0.001), diabetes (p 0.04), hypertension (p 0.01), and ischemic heart disease (p 0.001). Patients with prior embolism had a significantly higher event rate than those without embolism at baseline (p 0.006). Adverse events. We recorded a total of 42 severe bleeding episodes (Table 3): the antiplatelet arm in the intermediaterisk group showed significantly fewer events than the anticoagulant and the combined therapy arms. No difference was found between the anticoagulant and the combined therapy arms. Fourteen bleeding events were intracranial (two in the antiplatelet arm, nine in the anticoagulant, and three in the combined arms). Fifteen hemorrhages were gastric, most of them in the combined therapy arms. By endoscopy, seven patients in the combined therapy group showed superficial mucosal erosions and three patients in the anticoagulant arm showed recent ulcers. DISCUSSION The NASPEAF allows us to present, for the first time, three new aspects of antithrombotic treatment in AF. These are the effect of anticoagulation alone versus combined therapy with moderate anticoagulation, the event rate in mitral stenosis versus nonvalvular patients, and the event rate in patients with prior embolism versus no embolism at baseline. In the intermediate-risk group, primary outcome was lower with combined therapy than with the anticoagulant therapy despite the anticoagulant maintaining a lower me-

6 1562 Pérez-Gómez et al. JACC Vol. 44, No. 8, 2004 Antithrombotic Therapy in Atrial Fibrillation October 19, 2004: Figure 2. Primary outcome Kaplan-Meier survival curves. (A) Intermediate-risk group (nonvalvular atrial fibrillation without embolism at baseline). (B) High-risk group (mitral stenosis with/without embolism and nonvalvular atrial fibrillation with embolism at baseline). N number at risk; SE standard error. dian INR and three times as many INR controls 2. Primary outcome plus severe bleeding was also lower with the combined therapy, making this therapy the more efficacious and safer treatment. In the high-risk group, the combined therapy showed a significantly lower incidence of primary events when compared with the anticoagulant therapy. This benefit was obtained with a lower median INR than in the anticoagulant arm and with 39% of INR controls 2. Therefore, an INR of 2, considered by the EAFT (16) and Hylek et al. (18) as the lower limit of safety during anticoagulant therapy, is not applicable to patients receiving combined therapy. We planned a primary outcome, similar to that of the EAFT (16), composed of vascular death, systemic embolism, and stroke. Primary outcome in other trials (7 9) was the composite of systemic embolism and stroke. When we analyzed this outcome in the NASPEAF, the combined

7 JACC Vol. 44, No. 8, 2004 October 19, 2004: Pérez-Gómez et al. Antithrombotic Therapy in Atrial Fibrillation 1563 Figure 3. Relative effect on vascular events of combined versus anticoagulant therapy. Log hazard ratio (HR) (95% confidence interval [CI]) logarithm hazard ratio whose 95% CI (error bars) excludes the vertical line are statistically significant at the 5% level (or significant with p 0.05). AMI acute myocardial infarction. therapy compared with the anticoagulant therapy decreased the event rate of this aggregate by 56% in both the intermediate- and high-risk groups. Table 4 shows the characteristics and the embolism-stroke events rate in the anticoagulant-alone and the combined therapy arms in the NASPEAF and the three previous similar trials, but Edvardsson et al. (9) did not include an anticoagulation therapy arm. Our intermediate-group risk profile appears to be similar to that of the AFASAK II cohort (8) and our high-risk group appears to be similar to the SPAF III study (7). Although event rates with standard anticoagulation were similar, those with combined therapy in the intermediate- and high-risk groups in the NASPEAF showed significant benefit compared with the AFASAK II and SPAF III study, respectively. The possible explanation for this difference may be the higher anticoagulation intensity achieved in the NASPEAF (median INR 1.93 to 2.17 in the intermediate- and high-risk groups) compared with the other trials (median INR 1.3). The NASPEAF also showed that efficacious combined therapy needs a moderate anticoagulation intensity and that laboratory control is mandatory. Thus, INR levels 1.8 to 1.9 may not offer any significant benefit and levels 2.5 are not needed. Patients with prior embolism, investigated in published trials, presented very high event rates with standard anticoagulation. The composite of embolism-stroke-tia was 3.9 in the EAFT (16), 4.6 in the SPAF III study high risk (7), and 4.8 in the NASPEAF. The corresponding rates of vascular death were 4.1, 4.9, and 4.3 respectively, and severe bleeding was about 2 per 100 patients/year in the three trials. The NASPEAF showed that combined antiplatelet and moderate-intensity anticoagulation therapy could safely reduce the vascular event rate in this high-risk group of patients. Our trial also investigated antithrombotic treatment in patients with AF plus mitral stenosis and in nonvalvular patients. The embolic event rate, during equal anticoagulant intensity, was similar in either the valvular or nonvalvular groups, without embolism (1.26 vs. 1.43) or with embolism at baseline (4.81 vs. 4.94). The existence of valve stenosis and a predominant female gender in patients with mitral stenosis were probably counterbalanced by an older age and a higher incidence of several other risk factors in nonvalvular patients. We have included mitral stenosis patients without prior embolism in the high-risk group, but these patients presented a lower event rate than we expected and should probably be included in the intermediate-risk group. Patients assigned to antiplatelet therapy presented a relatively low event rate, consistent with other series of patients with relative low intrinsic risk (19). In the WARIS II (10), anticoagulation or combined therapy were both superior to antiplatelet therapy alone in post-mi patients, but were not better or worse than each other. Here, combined therapy appeared to be superior to either anticoagulation or antiplatelet therapy alone in patients with AF; combined therapy could be indicated in patients with both pathologies. Anticoagulant therapy was controlled at anticoagulation units. The percentage of INR tests within the preset ranges was 65 to 73, indicating a good control level (20). The median INR in the anticoagulant arms (target INR of 2 to 3) was 2.47 in the intermediate-risk group and 2.50 in the high-risk group. The preset range in the combined arm of

8 1564 Pérez-Gómez et al. JACC Vol. 44, No. 8, 2004 Antithrombotic Therapy in Atrial Fibrillation October 19, 2004: Figure 4. Kaplan-Meier survival curves of the composite embolism-stroke-transient ischemic attack (TIA) in different groups of patients. (A) Survival curves in the combined therapy and anticoagulant arms of the intermediate- and high-risk groups. (B) Survival curves in groups of patients receiving anticoagulant therapy for INR (2 to 3). ACO anticoagulant; Emb/embol embolism; TIA transient ischemic attack; other abbreviations as in Figure 2. the intermediate-risk group was 1.25 to 2 and the median INR achieved was The range in the combined arm of the high-risk group was 1.4 to 2.4 and the median achieved was 2.17; the achieved median INR in the combined arms was greater than planned because our hematologists did not want low anticoagulation intensity and tended to maintain the level of anticoagulation near the upper limit of the target.

9 JACC Vol. 44, No. 8, 2004 October 19, 2004: Pérez-Gómez et al. Antithrombotic Therapy in Atrial Fibrillation 1565 Table 4. Ischemic and Hemorrhagic Stroke and Systemic Embolism Anticoagulant Therapy Events Combined Antiplatelet-Anticoagulant Therapy Events n INR n Rate (Fatal) Pt/yrs n INR Anti-platelet n Rate (Fatal) Pt/yrs SPAF III (7) (1.7) ASA (5.6) 558 AFASAK II (8) (0) ASA (0.3) 377 Edvardsson et al. (9) placebo 6.5 (0.4) low ASA (0.8) NASPEAF: risk groups Intermediate (0.7) Tri (0) 541 High (1.3) Tri (0.7) 610 In Edvardsson et al. (9), the control group was placebo and the level of INR in the combined therapy arm was not regularly controlled. The intermediate-risk group in NASPEAF has a risk profile similar to AFASAK II and the high-risk group is similar to SPAF III. AFASAK Copenhagen Study on Atrial Fibrillation; ASA aspirin; Tri triflusal; NASPEAF National Study for Prevention of Embolism in Atrial Fibrillation; Pt/yrs patient-years; SPAF Stroke Prevention in Atrial Fibrillation. There was a tendency to preferential intracranial bleeding location with standard anticoagulation, consisting with a higher anticoagulation intensity (21). On the contrary, gastric bleeding was more common in the combined therapy arms. The acid component of triflusal could be responsible for the superficial erosions found by endoscopy, which lead to bleeding when a certain level of anticoagulation is added. Different trials (17,22 25) investigated combined therapy in mechanic valve prosthetic patients. The incidence of severe bleeding seems to be dose-related to either antiplatelet or anticoagulant agent. Accordingly, there is a tendency to use a low antiplatelet dose in the combined therapy. The triflusal dose in our combined therapy arms was 600 mg/day, which corresponded to aspirin 300 mg/day, and the risk of nongastric severe bleeding was low. Two blind published trials, Triflusal in Myocardial Infarction (TIM) (12) in post-mi patients and Triflusal versus Aspirin in Cerebral Infarction Prevention (TACIP) (13) in postcerebral infarction patients, compared triflusal 600 mg/day to aspirin 300 to 325 mg/day. In both studies, no difference in efficacy was found, but there was significantly higher major hemorrhage or central nervous system hemorrhage with aspirin. Also, in the WARIS II (10) the combination of aspirin 80 mg/day plus anticoagulation with an INR of 2.2 offered higher risk of bleeding than anticoagulation alone, for an INR of 2.8. Thus, although one can extrapolate from the NASPEAF that the combination of anticoagulation and aspirin is effective, a safe dose of aspirin cannot be determined, but the guidelines for valvular prosthesis (26) propose 75 to 100 mg/day in combined therapy. The emerging benefit of direct antithrombin agents in patients with AF (27), similar to anticoagulation for INR of 2to3,mayinfluence future antithrombotic therapy in these patients. However, an open question remains about the efficacy of antithrombin agents in patients with previous embolism and those with mechanical valvular prosthesis, in whom anticoagulation for an INR of 2 to 3 was not sufficiently effective. We probably need new trials to evaluate the combination of antiplatelet and direct antithrombin agents in these patients. In conclusion, the addition of antiplatelet therapy to reduced-intensity anticoagulation in AF patients stratified for risk of stroke significantly reduces subsequent events compared with patients receiving standard anticoagulation, and does so without increasing bleeding risk. Acknowledgment We are grateful to Dr. Daniel Singer (Massachusetts General Hospital) for helpful comments on early drafts of the manuscript. Reprint requests and correspondence: Dr. Francisco Pérez- Gómez, Servicio de Cardiología, Hospital Clínico San Carlos, Prof. Martín Lagos s/n, Madrid, Spain. fperezg.hcsc@salud.madrid.org. REFERENCES 1. Wolf PA, Abbot RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke 1991;22: Fuster V, Ryden LE, Asinger RW, et al. ACC/AHA/ESC guidelines for the management of patients with atrial fibrillation. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and Policy Conferences (committee to develop guidelines for the management of patients with atrial fibrillation) developed in collaboration with the North American Society of Pacing and Electrophysiology. Eur Heart J 2001;22: Petersen P, Boysen G, Godtfredsen J, et al. Placebo-controlled, randomised trial of warfarin and aspirin for prevention of thromboembolic complications in chronic atrial fibrillation: the Copenhagen AFASAK study. Lancet 1989;1: Hart RG, Benavente O, McBride R, et al. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med 1999;131: Cairns JA, Connolly SJ. Nonrheumatic atrial fibrillation. Risk of stroke and role of antithrombotic therapy. Circulation 1991;84: Atrial Fibrillation Investigators. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation: analysis of pooled data from five randomised controlled trials. Arch Intern Med 1994;154: Stroke Prevention in Atrial Fibrillation III Investigators. Adjusteddose warfarin versus low-intensity, fixed-dose warfarin plus aspirin for high-risk patients with atrial fibrillation: Stroke Prevention in Atrial Fibrillation III randomised clinical trial. Lancet 1996;348: Gullov AL, Koefoed BG, Petersen P, et al. Fixed minidose warfarin and aspirin alone and in combination vs. adjusted-dose warfarin for stroke prevention in atrial fibrillation: Second Copenhagen Atrial Fibrillation, Aspirin, and Anticoagulation Study. Arch Intern Med 1998;158: Edvardson N, Juul-Möller S, Omblus R, et al. Effects of low dose warfarin and aspirin versus no treatment on stroke in a medium-risk

10 1566 Pérez-Gómez et al. JACC Vol. 44, No. 8, 2004 Antithrombotic Therapy in Atrial Fibrillation October 19, 2004: patient population with atrial fibrillation. J Intern Med 2003; 254: Hurlen M, Abdelnoor M, Smith P, et al. Warfarin, aspirin, or both after myocardial infarction. N Engl J Med 2002;347: McNeely W, Goa KL. Triflusal. Drugs 1998;55: Cruz-Fernandez JM, Lopez-Bescos L, Garcia-Dorado D, et al. Randomised comparative trial of triflusal and aspirin following acute myocardial infarction. Eur Heart J 2000;21: Matías-Guiu J, Ferro JM, Alvarez-Sabín J, et al. Comparison of triflusal and aspirin for prevention of vascular events in patients after cerebral infarction. The TACIP study: a randomised, double blind, multicenter trial. Stroke 2003;34: Landefeld CS, Anderson PA, Goodnough LT, et al. The bleeding severity index: validation and comparison to other methods for classifying bleeding complications of medical therapy. J Clin Epidemiol 1989;42: Stroke Prevention in Atrial Fibrillation Investigators. Warfarin versus aspirin for prevention of thromboembolism in atrial fibrillation. Stroke Prevention in Atrial Fibrillation II Study. Lancet 1994;343: EAFT (European Atrial Fibrillation Trial) Study Group. Secondary prevention in non-rheumatic atrial fibrillation after transient ischemic attack or minor stroke. Lancet 1993;342: Turpie AGG, Gent M, Laupacis A, et al. A comparison of aspirin with placebo in patients treated with warfarin after heart-valve replacement. N Engl J Med 1993;329: Hylek EM, Skates SJ, Sheehan MA, et al. An analysis of the lowest effective intensity of prophylactic anticoagulation for patients with nonrheumatic atrial fibrillation. N Engl J Med 1996;335: The SPAF IIII Writing Committee for the Stroke Prevention in Atrial Fibrillation Investigators. Patients with non-valvular atrial fibrillation at low risk of stroke during treatment with aspirin: Stroke Prevention in Atrial Fibrillation III Study. JAMA 1998;279: Italian Federation of Anticoagulation Clinics. A guide to oral anticoagulation treatment: recommendations of the Italian Federation of Anticoagulation Clinics (FCSA). Haematologica/J Hematol 2003;88 Suppl 2: van Walraven C, Hart RG, Singer DE, et al. Oral anticoagulants vs. aspirin in nonvalvular atrial fibrillation. JAMA 2002;288: Altman R, Boullon F, Rouvier J, et al. Aspirin and prophylaxis of thromboembolic complications in patients with substitute heart valves. J Thorac Cardiovasc Surg 1976;72: Hurlen M, Erikssen J, Smith P, et al. Comparison of bleeding complications of warfarin and warfarin plus acetylsalicylic acid: a study in 3166 outpatients. J Intern Med 1994;236: Chesebro JH, Fuster V, Elveback LR, et al. Trial of combined warfarin plus dipyridamole or aspirin therapy in prosthetic heart valve replacement: danger of aspirin compared to dipyridamole. Am J Cardiol 1983;51: Verheugt FW. Bleeding risk of combined oral anticoagulation and antiplatelet therapy in cardiovascular disease. J Cardiovasc Risk 1996; 3: Stein PD, Alpert JS, Dalen JE, et al. Antithrombotic therapy in patients with mechanical and biological prosthetic heart valves. Chest 1998;114:602 10S. 27. Executive Steering Committee on behalf of the SPORTIF III Investigators. Stroke prevention with oral direct thrombin inhibitor ximegalatran compared with warfarin in patients with non-valvular atrial fibrillation (SPORTIF III): randomised controlled trial. Lancet 2003; 362: APPENDIX For a list of the Events Committee, Safety Committee and Data Monitoring, Central Coordination and Data Management, Statistical Analysis, and Centers and Principal Investigators, please see the October 19, 2004, issue of JACC at