The Challenge of Stroke Prevention

Stroke prevention represents a critical health challenge. The demographics of aging and the association of stroke with aging bring together factors that compel a concerted effort to limit the individual and societal risks of a worldwide epidemic of stroke.¹ As a disease, stroke functions on 2 biological frontiers: aging and brain vascular function. Society has recognized the epidemic of atherosclerotic vascular disease for the last 4 decades and has begun to deal with the cardiac consequences of atherosclerosis. These efforts are apparent in the public domain and in the physician's office. Multiple acute interventions and prevention programs are available to limit acute disease and delay or reverse coronary artery narrowing and help prevent progression to subsequent myocardial infarctions.

Unfortunately, such is not the case with stroke. Embolic stroke and carotid stenosis represent treatable subsets of stroke etiologies. Aspirin can provide partial prevention, but the majority of stroke survivors still have significant risk of subsequent stroke, and well-tested interventions to eliminate stroke recurrence are not available. A reasonable extrapolation from studies of secondary prevention of coronary artery disease is that lowering plasma homocysteine by use of high-dose folate vitamin therapy might help reduce stroke risk by limiting progressive vascular disease.² The Vitamin Intervention for Stroke Prevention (VISP) trial reported by Toole et al³ in this issue of THE JOURNAL was designed to test this hypothesis.

The VISP trial demonstrates the difficult challenge in attempting to limit recurrent stroke and increase the number of independent years of survival in an aging population. The investigators of this carefully conducted trial report the results of 4.5 years of stroke prevention efforts that combined "best available medical and surgical management" with a vitamin regimen that included either a high-dose or low-dose formulation of folate, vitamin B₆, and vitamin B₁₂ as an organized effort to limit the frequency of secondary stroke and vascular events. The trial randomized 3680 patients with stroke in the previous 120 days and a fasting plasma total homocysteine level at or above the 25th percentile of the North American stroke population. As the population level of total homocysteine declined over the course of the trial, the total homocysteine study eligibility level decreased from 10.5 µmol/L or more in the beginning of the study to 9.5 µmol/L or more for men and 8.5 µmol/L or more for women by the end of the enrollment period.

The type of intervention and the magnitude of planned homocysteine reduction were similar to that used in the study of secondary prevention following coronary interventions.² A representative sample of North American stroke survivors was enrolled. Patients with embolic stroke and recent carotid surgery (forms of stroke requiring established alternative therapeutic interventions) were appropriately excluded. The selection of the vitamin intervention was designed to produce an elevation of plasma folate levels and a reduction of total homocysteine levels in the high-dose group. Careful procedures to assess drug administration compliance included a run-in period, pill counts, and assessment of plasma folate and vitamin B₁₂ levels to ensure a vigorous test of the possible benefits of total homocysteine lowering. Overall, approximately 8% of the participants experienced recurrent stroke during the 2-year study period, with 5% occurring in the first year and 3% in the second year. No difference in the rate of stroke occurred between the high- and low-dose vitamin groups. The study demonstrates that lowering of total homocysteine from 13.4 to 11.0 µmol/L does not reduce stroke recurrence.

How should clinicians integrate this information into daily practice? The results are clearly negative from the intention-to-treat basis, and, thus, most clinical trialists would reject the idea of a causal role for total homocysteine in the etiology of cerebrovascular disease and stroke. However, a closer look at the data suggests that the link may be plausible. One interpretation is that the effect of elevated total homocysteine on cerebral vessels is different than on coronary arteries, where much larger treatment effects have been demonstrated.² This seems possible, given the Kaplan-Meier curves for stroke, coronary events, and death (Figure 3) in the article by Toole et al. Here, no indication of differential stroke rates existed for the 2 treatment groups, but some effect of the high-dose regimen for preventing coronary events and death is evident.

Other interpretations of the data include the possibility that a different trial design would produce beneficial stroke prevention. The VISP authors calculate that a much larger study, estimated at 20 000 participants, would provide a sufficiently powered test of the beneficial trends they observed. Another possibility is a smaller trial treating patients with higher levels of total homocysteine. Such a trial would address the interpretation that either the amount of the total homocysteine reduction was insufficient or the inclusion of stroke patients with total homocysteine levels too close to "normal" limited the beneficial effects. The demonstration of significantly different risk profiles for all 3 sentinel vascular outcomes (stroke, coronary events, and death) among the upper, middle, and lower tertiles of total homocysteine suggests continued plausibility for stroke reduction associated with total homocysteine lowering.³ Further investigation should lead to a better understanding of which patients to treat and how vigorously to treat them. Like other well-designed studies, the VISP trial has produced a solid set of findings that should reformulate the total homocysteine hypothesis in a manner that can lead to a more complete understanding of the role of this important putative risk factor for stroke.

However, the VISP trial also demonstrates that with "best available medical and surgical management," recurrent strokes occur at a cumulative rate of 4% per year for a group with an average age of 66 years. Moreover, the VISP study provides some interesting data about the vigor of stroke risk reduction practiced in the last 4 years. Unfortunately, the picture is not good. For instance, among stroke survivors, about 13% continued to use tobacco, 3% started smoking during the trial, and only 3% discontinued smoking. Neither body mass index nor blood pressure declined, only small reductions in total cholesterol levels occurred (ie, from 200 mg/dL to 193 mg/dL), and almost 20% of patients experienced a decline of at least 2 points on the Mini-Mental State Examination.

Smoking cessation, exercise, and blood pressure reduction are all appropriate and important stroke risk reduction interventions.⁴ Even more evidence is now available that aggressive treatment of these individual risk factors leads to stroke reduction.^{5 - 8} Despite knowledge of benefit, clinicians apparently have not yet accepted the need to aggressively optimize the medical and behavioral factors that lead to risk reduction. One research avenue worth strong consideration is to find novel means to address established, unameliorated stroke risk factors with validated treatment regimens. Such an approach would have to address multiple factors simultaneously but should be explored to improve the period of independent living for the aging population at risk of stroke.

In a second article in this issue of THE JOURNAL, the Antiphospholipid Antibodies and Stroke Study (APASS) investigators⁹ examined whether the presence of antiphospholipid antibodies (aPL), which have been associated with hypercoagulability, was useful for predicting subsequent thromboembolic events, including stroke, in patients with ischemic stroke. The results of this report parallel the VISP findings in several interesting ways. The primary analysis of the APASS data did not show a benefit of stroke prevention for aspirin or warfarin for stroke patients who had a single baseline (ie, prior to study treatment) elevated anticardiolipin (aCL) or lupus anticoagulant (LA) antibody test result. However a distinctly higher risk of stroke and vascular events was demonstrated in the subgroup of patients who demonstrated elevated levels of both aCL and LA antibodies at enrollment. This finding and consensus information from the ninth aPL conference¹⁰ suggest that persistence of elevated antibodies to these factors represents an even greater risk for stroke. Persistence of antibody titers was not an entrance criteria for APASS. Alternatively, as in VISP, the concentration of aPL may be critical for determining the degree of risk of vascular events and stroke. Thus, a different quantitative threshold may exist for stroke risk than the one used in the current study. It is an important possibility that among a higher-risk population, patients with positive aPL titers could benefit from more aggressive risk amelioration. Alternatively, given the increased risk that patients with aPL continue to demonstrate, novel therapies not involving aspirin or warfarin may need to be investigated.

Taken together, the results of the VISP study and the APASS argue for a greater emphasis on research in stroke prevention and better ways to identify patients at high risk of recurrent stroke. The benefits of amelioration of additional risk factors must be explored and treatment models should be developed that allow for more complete implementation of the risk reduction methods already demonstrated to be valid.¹¹ Without such an effort, aging will be complicated by more rapid than necessary physical and mental impairment. Hopefully, government and industry funding will support dynamic stroke risk reduction research efforts. Clearly, the task will not be easy, but the rewards for patients will be great.