Thromboembolism and bleeding during therapeutic anticoagulation are the major chronic risks for patients after mechanical prosthetic heart valve replacement. Prophylaxis for thromboembolism based on pathogenesis and risk and the involvement of both fibrin and platelets has led to permanent long-term oral anticoagulant therapy for consistent protection.1 2 3 The incidence of thromboembolic events is related to the type of valve and its location; the incidence after aortic valve replacement is lower than after mitral valve replacement.2 3 The peak incidence of thromboembolism is during the first 3 months after surgery, probably reflecting the lack of endothelialization of the newly implanted prosthetic materials and delay in achieving therapeutic anticoagulation in the early days after operation.4 5 The Dacron sewing ring is common to all prosthetic heart valves and forms a substrate for platelet deposition as soon as blood flows across the valve.2 Current recommendations are levels of oral anticoagulation that prolong the prothrombin time ratio to an international normalized ratio (INR) of 2.5 to 3.5 for tilting-disk and bileaflet valves and probably for caged-ball valves.3
The study reported in this issue of Circulation by Acar et al and the AREVA Group6 is well designed, well conducted, and focused on lower-risk patients. It addresses the optimal target level for anticoagulation. They compared a target INR of 2.0 to 3.0 with 3.0 to 4.5 in a randomized trial after a single mechanical prosthetic valve replacement with either of two bileaflet mechanical prostheses (St Jude valve in 81% and Omnicarbon valve in 19% of the 433 patients) placed in the aortic position in 96% of patients and the mitral position in the remainder. This distribution was based on the definition of low risk, which follows.
Patients were a mean of 59 years old (upper age limit, 75 years) and had to be in normal sinus rhythm with no more than modest left atrial enlargement (left atrial diameter, ≤50 mm). In addition, 20% of patients had an increased risk, namely, associated aortocoronary bypass graft surgery. Thromboembolic events were assessed by both a CT brain scan at 1 year after operation and transient or permanent neurological deficits with a mean follow-up of 2.2 years (range, 1 to 4 years). Although the target ranges for INRs were broad, there was a clustering of INRs among all patients in the study; the median INR for each patient was between 2.5 and 3.5 in 72% of all patients. For the INR 2.0 to 3.0 group, the median INR was 2.7 (median INR, ≥2.5 in 65%), and in the INR 3.0 to 4.5 group, the median INR was 3.2 (median INR, ≤3.5 in 84%). Thus, there appeared to be a normal tendency of the treating physician to stay between the recommended range of 2.5 to 3.5.
This is the first study group to systematically use CT brain scans to measure the incidence of asymptomatic cerebral thromboembolism in patients with prosthetic heart valves. This increases the accuracy and precision of determining the incidence of thromboembolism and is a useful addition for objectively assessing this end point.
The incidence of all thromboembolic events (symptomatic and asymptomatic) was similar between the INR 2.0 to 3.0 and INR 3.0 to 4.5 groups: 3.1 and 2.4 events per 100 patient-years, respectively; the incidence of symptomatic thromboembolic events was also similar: 1.9 and 1.7 events per 100 patient-years, respectively. The incidence of all hemorrhagic events was significantly lower: 11.2 versus 20.5 events per 100 patient-years in the INR 2.0 to 3.0 versus 3.0 to 4.5 groups, respectively, with a trend toward a lower incidence of major hemorrhagic events, 4.0 versus 5.6 events per 100 patient-years.
How low can the INR go and still provide antithrombotic protection? Three studies reported at the National Scientific Sessions of the American College of Cardiology in Orlando, Fla, March 25-27, 1996, showed that oral anticoagulation to INRs of ≤1.5 is ineffective. For prevention of stroke and systemic embolism in nonvalvular atrial fibrillation, very-low-dose warfarin combined with aspirin was significantly less effective than oral anticoagulation alone to an INR of 2.0 to 3.0; this necessitated that the trial be stopped by the Data and Safety Monitoring Board.7 For aortocoronary saphenous vein graft occlusion after coronary bypass graft operation and for prevention of death, recurrent myocardial infarction, and stroke after myocardial infarction, very-low-dose warfarin plus aspirin was no better than aspirin alone (Post CABG Trial: effect of cholesterol lowering and low-intensity oral anticoagulation on late saphenous vein graft status, presented by M.I. Domanski, D.B. Hunninghake, and L. Campeau; and the Coumadin Aspirin Reinfarction Study [CARS], presented by V. Fuster). In addition, the AFTER Trial assessing death and recurrent myocardial infarction after thrombolytic therapy with acylated streptokinase for acute myocardial infarction compared oral anticoagulation to an INR of 2.0 to 2.5 with aspirin 75 mg twice daily and found an equivalent incidence of the combined end point of death or myocardial infarction.6 These data and the combined data of the five trials of stroke prevention in atrial fibrillation8 (showing that oral anticoagulation to an INR of 2.0 to 3.0 significantly reduces stroke compared with placebo) suggest that the target INR should not be <2.0.
Thus, the AREVA Study shows that the low-risk patient with a St Jude or Omnicarbon mechanical prosthetic aortic valve replacement in normal sinus rhythm and a left atrial diameter ≤50 mm is the special case for lower-dose oral anticoagulation to an INR of 2.0 to 3.0. Since too few patients qualified for this low-risk group with mitral valve replacement in this study, there are insufficient data to change the current recommendation of oral anticoagulation, which is an INR of 2.5 to 3.5.
Patients with associated vascular disease are at higher risk. Aspirin added to oral anticoagulation provides additional protection but with some increased risk of bleeding. Turpie et al8 showed that the addition of aspirin (100 mg/d) to oral anticoagulation (median INR, 3.0) in higher-risk patients with prosthetic heart valves reduced mortality from vascular causes and major systemic thromboembolism. In this high-risk group, the bleeding risk of combined treatment was more than offset by the considerable reduction in major systemic embolism or death from any cause. The combined end point of major systemic embolism, nonfatal intracranial hemorrhage, or death of hemorrhage or vascular causes occurred at a rate of 9.9% versus 3.9% per year, respectively.8 This same group is completing a trial of warfarin therapy to an INR of 2.0 to 2.5 plus aspirin 100 mg/d versus warfarin alone to an INR of 3.0 to 3.5 in patients with mechanical heart valves or high-risk patients with prosthetic tissue valves.
A possibility for decreasing the risk of gastrointestinal bleeding might be considered. The current pathogenesis of bleeding from a spontaneous peptic ulcer9 provides a possible rationale. Patients with a prior history of upper gastrointestinal bleeding related to peptic ulcer disease or recurrent peptic ulcer disease who are being considered for chronic oral anticoagulation may be assessed for antibodies to Helicobacter pylori, and, if positive, may be treated with antibiotics to eradicate the H pylori. This requires further study.
The incidence of thromboemboli in the early days and weeks after operation, while patients are being started on oral anticoagulation, and for the first 3 months is high.2 5 Thus, the incidence of thromboembolism in patients on heparin started 6 hours after operation, as administered in the AREVA Study, and the incidence during the first 3 months in this group would be of interest for a future report, because the practice in the United States is not to use heparin and to start oral anticoagulation the evening after (or the day after) operation. Studies by Perier et al previously reviewed2 suggest that early treatment with heparin starting 6 hours after operation result in a low incidence of thromboemboli early and later after operation.
- Copyright © 1996 by American Heart Association
Stein B, Fuster V, Halperin JL, Chesebro JH. Antithrombotic therapy in cardiac disease: an emerging approach based on pathogenesis and risk. Circulation.. 1989;80:1501-1513.
Chesebro JH, Fuster V. Valvular heart disease and prosthetic heart valves. In: Fuster V, Verstraete M, eds. Thrombosis in Cardiovascular Disorders. Philadelphia, Pa: WB Saunders Co; 1992:191-214.
Stein PD, Alpert JS, Copeland J, Dalen JE, Goldman S, Turpie AGG. Antithrombotic therapy in patients with mechanical and biologic prosthetic heart valves. Chest. 1995;108(suppl):371S-379S.
Acar J, Boissel JP, Iung B, Samama MM, Michel PL, Teppe JP, Pony JC, Le Breton H, Thomas D, Isnard R, De Gevigney G, Viguier E, Sfihi A, Hanania G, Ghannem M, Mirode A, for the AREVA Group. AREVA: multicenter randomized comparison of low-dose versus standard-dose anticoagulation in patients with mechanical prosthetic heart valves. Circulation.. 1996;94:2107-2112.
Stroke Prevention in Atrial Fibrillation Investigators (SPAF). Adjusted-dose 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;348:633-638.
Massuda HK, Boyd EJS. Who should undergo testing for Helicobacter pylori? Am J Gastroenterol.. 1996;91:1070-1071.