Secondary Prevention with Antioxidants—The SPACE Stories—News from the Antioxidant World: Real or Chance?

Abstract & Commentary

Synopsis: A major reduction in cardiovascular morbidity and mortality in end-stage renal disease patients can be achieved with vitamin E supplementation.

Source: Boaz M, et al. Lancet 2000;356:1213-1218.

The antioxidant hypothesis remains the focus of considerable attention. Pathogenetic mechanisms for many vascular disease phenomena are readily attributed to oxidant stress, which also contributes to inadequate nitric oxide bioavailability. The concept that widely available vitamins with antioxidant properties might be an inexpensive, safe, and effective therapy for primary or secondary cardiovascular prevention, is attractive. However, recent randomized, clinical trial data are very sobering with respect to supporting the antioxidant hypothesis: More than 20,000 patients in the HOPE and GISSI-P trials1,2 showed no benefit for vitamin E in patients with cardiovascular disease or in high-risk diabetics without overt cardiac disease. The CHAOS trial data, published several years ago, did suggest a benefit with vitamin E supplementation in patients with coronary artery disease undergoing angioplasty.3 However, this study has been widely criticized for a variety of reasons and is not considered to be reliable.

The SPACE trial (Secondary Prevention with Antioxidants of cardiovascular disease and end stage renal disease), was carried out in five Israeli dialysis centers in high-risk end stage renal disease (ESRD) subjects with documented cardiovascular disease. The hypothesis was that vitamin E supplementation would decrease new vascular events due to decreases in oxidant stress and oxidation of LDL cholesterol. Stable dialysis patients between the ages of 40-75 and a documented history of cardiovascular disease were randomized to vitamin E 800 IU/d or matching placebo in this double-blind, randomized trial. An event rate of 30% over two years in the placebo group was postulated. The primary end point was a composite consisting of myocardial infarction (MI), stroke, unstable angina, and new peripheral vascular disease. The patient groups appeared to be well matched at baseline, although the cohorts were rather small, consisting of just under 100 patients each.

The results indicated a robust reduction in the primary end point of 54% (P = 0.014). If sudden death was included, there was a 46% reduction in relative risk, (34 end points with placebo vs 18 in the vitamin E group). An adjustment was made for smoking, as the vitamin E cohort had a greater percentage of active smokers (24% vs 14%). The results in smokers were similar to the overall primary outcome. The major end point affected by vitamin E was new MI, which demonstrated a 70% reduction (P = 0.016). When sudden deaths were considered as a fatal MI, the reduction was 55% (P = 0.04). Non-fatal was reduced by 66% (P = 0.08). No obvious variables could be detected to indicate a different baseline in those individuals who did or did not have an MI during this study. The Kaplan Meier event curve separation began at approximately 300 days, widening at study termination, with some patients being followed for 600-700 days.

Although there were trends toward a reduction in incident peripherial vascular disease, ischemic stroke, and unstable angina (62%, 49%, and 15% relative risk reduction, respectively), these were nonsignificant due to the small numbers of events and the small cohort size. Side effects were few, and not clearly related to vitamin E. The mean annual MI rate in the placebo patients in SPACE was 12.3%, substantially higher than CHAOS, GISSI, and HOPE. The latter two had MI rates of 2.5 and 3.8% per year, respectively, both lower than CHAOS. Overall mortality was not reduced in the SPACE study due to a "nonsignificant increase in non- cardiac mortality in patients who receive vitamin E." Vitamin E may have contributed to hemorrhage; otherwise, details were not provided. Boaz and colleagues point out that high-dose vitamin E supplementation has been shown in experimental studies to reduce the atherogenic profile, including monocyte super oxide release, lipid oxidation, platelet aggregation, smooth muscle cell proliferation, and possibly "stabilization of atherosclerotic plaque." They state, "our study is not the final word," and did not make specific therapeutic recommendations. Boaz et al call for a major clinical trial to further evaluate the vitamin antioxidant approach in ESRD patients.

Comment by Jonathan Abrams, MD

The results of SPACE, if validated by larger trials, are of considerable importance, indicating that a major reduction of cardiovascular morbidity and mortality in ESRD subjects can be achieved with vitamin E supplementa- tion. There are few obvious limitations of this study, other than the small number of patients followed over a short time period; the degree of relative risk reduction for MI and other end points are so robust that they cannot be ignored. Boaz et al believe that the baseline risk was equivalent in the two groups. However, information regarding left ventricular function, active myocardial ischemia, or lipid profiles are not provided. There is no question that the ESRD patients were high risk; 42% of the entire cohort had diabetes and a comparable number had hypertension. Prior MI was documented in 49% of the vitamin E cohort and 57% of the placebo patients; no quality of life or clinical classification was provided; therefore, it is possible that the two groups may have had a different burden of morbidity in spite of relatively comparable major prior cardiovascular disease. The group size of just under 100 patients each also presents a problem, allowing for the probability of a significant type I error. Given the large decrease in the primary end point, it is somewhat surprising that overall mortality was not affected by the study. The reasons for this are not clear and although unlikely, could possibly be due to an increased hazard with high-dose vitamin E. The dose used in this study was higher than in all three of the prior trials, although some patients in CHAOS did receive 800 IU.

What should the clinician do with these data? While the SPACE study is substantially less persuasive than HOPE, which recommended ACE inhibitors for all patients who meet the HOPE criteria, these data do suggest that ESRD patients who have a history of cardiovascular disease should be considered for high-dose vitamin E supplementation. In the SPACE trial, all such individuals were treated for hyperhomocysteinemia with vitamin B complex-folate supplementation. Given the enormous burden of morbidity and mortality in these patients and the lack of clear cut evidence that antioxidant vitamin supplementation or B vitamins carry any significant harm, it does appear reasonable for physicians to prescribe vitamin E in the doses used in SPACE. Many questions are unanswered regarding antioxidant vitamins, including whether synthetic or natural vitamins are better; whether the right vitamin molecule has been used in the major studies; whether vitamin C should be used in conjunction with vitamin E; and whether it is even reasonable to hypothesize that intake of vitamin supplements will be able to attenuate or even reverse the profound oxidant stress that is systemic in the vessels of individuals with many types of cardiovascular disease, including diabetes, hypercholesterolemia, and of course, ESRD patients.

A large clinical trial should be quickly launched testing vitamin E, and possibly the combination with vitamin C in ESRD. Given the differences in the patient populations in SPACE, GISSI-P, and HOPE, it is not appropriate to extend these observations to primary or secondary prevention in patients with coronary artery disease in the absence of renal failure.


1. Yusuf S, et al. HOPE trial. N Engl J Med 2000;342:154.

2. No author. GISSI-P trial. Lancet 1999;354:447.

3. Stephens NG, et al. Lancet 1996;347:781.