Veterans Affairs (VA) Cooperative Study #362.

Abstract

The Veterans Affairs (VA) Cooperative Study #362, reported by Willard C. Johnson and colleagues (J Vasc Surg 2002;35:413-21), did not demonstrate any difference in patency rate for warfarin sodium (target range, 1.4-2.8 international normalized ratio [INR]) plus aspirin (WASA) versus aspirin (ASA) in 458 patients after peripheral venous bypass grafting. In 373 patients with above-knee prosthetic bypass grafts, patency was significantly better in the WASA group (risk ratio, 0.62). These results conflict with our findings in the Dutch Bypass Oral Anticoagulants or Aspirin (BOA) Study,1The Dutch Bypass Oral Anticoagulants or Aspirin (BOA) Study GroupEfficacy of oral anticoagulants compared with aspirin after infrainguinal bypass surgery (The Dutch Bypass Oral Anticoagulants or Aspirin Study) a randomised trial.Lancet. 2000; 555: 346-351Google Scholar favoring oral anticoagulant treatment after venous bypass grafting and ASA in patients with prosthetic grafts. Therefore we would like to highlight the two most important differences in the design of the two trials and to discuss some differences in results and the implications for daily practice. First, the degree of anticoagulation was much higher in the Dutch BOA study: target range, 3.0 to 4.5 INR. To achieve optimal anticoagulation the target should be within this higher range.2Tangelder M.J.D. Algra A. Lawson J.A. Hennekes S. Eikelboom B.C. on behalf of the Dutch BOA Study GroupOptimal oral anticoagulant intensity to prevent secondary ischemic and hemorrhagic events in patients after infrainguinal bypass surgery.J Vasc Surg. 2001; 33: 522-527Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar The dosage of warfarin in the VA study was monitored once a month, compared with twice a month, on average, in the BOA study, which may have resulted in a higher proportion of time when degree of coagulation was within the target range. There was also a striking difference in percentage of patients who discontinued anticoagulant treatment: 40% in the VA trial versus only 14% in the Dutch BOA study. These differences in dosage and compliance contribute considerably to the difference in antithrombotic efficacy. The difference in ASA dose (325 mg/d in VA, 80 mg/d in BOA) probably does not explain any difference in results.3Antiplatelet Trialists’ CollaborationCollaborative overview of randomised trials of antiplatelet therapy. II Maintenance of vascular graft or arterial patency by antiplatelet therapy.BMJ. 1994; 308: 159-168Crossref PubMed Scopus (19) Google Scholar Second, the trials differed greatly in number of patients and duration of follow-up. The sample size in the BOA study (n = 2690) was based on expected occlusion rate after average follow-up of 2 years. The large number of patient-years (4560) in the BOA study allowed for the predefined subgroup analyses according to type of bypass procedure and graft material.4Dutch Bypass Oral Anticoagulants or Aspirin Study GroupInterpretation of Dutch BOA trial.Lancet. 2000; 355: 1186-1187PubMed Google Scholar The VA study comprised fewer patients (831), with 2638 patient-years of follow-up. Sample size was based on expected 6-year patency rate. Inasmuch as most occlusions occur in the first postoperative year, the number of patients with 1-year follow-up in the VA Study was probably too low to demonstrate a difference between the two treatment groups. It was surprising that the only statistically significant difference between the two treatment groups occurred in patients with prosthetic above-knee bypass grafts, and favored WASA. This difference is mainly due to a higher number of occlusions in the ASA group in the last 3 years of the study. Given the low compliance with warfarin therapy, especially over the long term, and the small number of patients (207) in this subgroup, this effect was probably not caused by allocated treatment but by chance or other unknown factors. What are the implications of the VA and BOA trials for daily practice? The common feature of both trials is the pragmatic design, which allows generalization of the findings to daily practice. Because of a well-organized system of Dutch Anticoagulation Clinics, anticoagulation therapy might be more effective in The Netherlands than in the United States. This could imply that addition of low-dose warfarin therapy to ASA treatment in the United States has little or no effect, whereas in the Dutch health care setting oral anticoagulant agents are more effective than ASA for prevention of venous graft occlusion. For patients with prosthetic bypass grafts, ASA remains the best antithrombotic treatment, worldwide. The authors of the VA Study have improved knowledge of antithrombotic therapy and discussed their findings clearly. We hope to have added further clarification with our expertise and this contribution to the discussion. Regarding “Regarding ‘Veterans Affairs (VA) Cooperative Study #362’”Journal of Vascular SurgeryVol. 39Issue 3Preview Full-Text PDF Open Archive