Perspective on the efficacy analysis of the Vitamin Intervention for Stroke Prevention trial

The pathogenesis of stroke is very different from that of myocardial infarction; most strokes are embolic, and as elevated plasma total homocysteine (tHcy) increases thrombosis, it increases the risk of cardioembolic strokes, particularly in atrial fibrillation, as well as the risk of deep vein thrombosis leading to paradoxical embolism, and elevated tHcy also increases strokes from cortical vein thrombosis.

The main study results of the Vitamin Intervention for Stroke Prevention (VISP) trial showed no benefit of vitamin therapy for homocysteine lowering. However, there were a number of problems with the trial including: 1. We did not use a placebo control; we used a multiple vitamin containing low doses of folate and B6, and the recommended daily intake of B12; 2. Patients with low serum B12 levels in both arms of the trial received injections of B12; and 3. Folate fortification of grain products in North America coincided with the initiation of the trial. We therefore conducted a hypothesis-driven analysis of a subgroup of VISP from which patients who could not have responded to the study treatment were excluded, i.e., those with low or very high serum B12, and those with renal failure.

In the remaining 2155 patients, there was a significant reduction of stroke, coronary events and death (p=0.049). When the participants were stratified according to the median entry level of B12 (313 pmol/L), it became apparent that serum B12 status was the key determinant of response, those with high serum B12 at entry who received high-dose vitamin had a 33% reduction of events compared to those with low B12 who received low-dose vitamins; the other two groups were intermediate (p=0.02 for survival free of stroke, coronary and death in the Kaplan-Meier analysis).

In this efficacy analysis, we found a significant reduction of vascular risk with vitamin therapy. Similarly, the Heart Outcomes Prevention Evaluation 2 trial showed a 25% significant reduction of stroke with vitamin therapy (p=0.03). The authors thought this result was a chance finding, because they did not think that stroke and myocardial infarction would be biologically different; however, stroke and myocardial infarction have very different pathogenesis, so that conclusion is unwarranted. It seems likely that even if vitamin therapy does not significantly reduce coronary events, it will reduce the risk of stroke.