Correspondence
Antioxidant Vitamins and Coronary Disease
N Engl J Med 2002; 346:1092-1093April 4, 2002
- Article
To the Editor:
Brown et al. (Nov. 29 issue)1 report the nonsignificant effect of antioxidants on angiographic and clinical outcomes in patients with coronary disease and low levels of high-density lipoprotein (HDL) cholesterol. However, the study appears to be underpowered to demonstrate reliably that antioxidant therapy has no meaningful benefit in this population. Moreover, the authors neither provided power analysis in the Methods section nor acknowledged this limitation in the Discussion section.
In the Results section, the researchers affirm that among the 34 patients in the placebo group, the mean percentage of stenosis in proximal arteries increased by a mean (±SD) of 3.9±5.2 percent, whereas among the 39 patients who received only antioxidant vitamins, it increased by 1.8±4.2 percent, a statistically nonsignificant difference. A simple power calculation,2 assuming a population-wide standard deviation of 5.2 percent for the mean change in the percentage of stenosis (derived from the results for the placebo group), accepting as clinically useful a mean change of 2 percent in the percentage of stenosis with antioxidant therapy given the 3.9 percent change with placebo (an absolute difference of 1.9 percent), and aiming for a Bonferroni-adjusted two-tailed P value of 0.05 and 80 percent power in a trial with a two-by-two factorial design comparing placebo with antioxidant therapy, indicates that each of the two groups should have included at least 159 patients.
The study by Brown et al. was thus underpowered to reach any meaningful conclusion about the inefficacy of antioxidant therapy in this population of patients. As we strive to move forward in clinical research and practice, we should remember that a large, powerful, and simple trial is often the only reliable means for answering a relevant question about therapy.3
3 ReferencesGiuseppe G.L. Biondi-Zoccai, M.D.
Antonio Abbate, M.D.
Catholic University of the Sacred Heart, 00168 Rome, Italy
gbiondizoccai@tiscali.it Pierfrancesco Agostoni, M.D.
University of Verona, 37134 Verona, Italy1
Brown BG ,Zhao X-Q ,Chait A , et al. Simvastatin and niacin, antioxidant vitamins, or the combination for the prevention of coronary disease. N Engl J Med 2001;345:1583-1592
Full Text | Web of Science | Medline2
Armitage P, Berry G. Statistical methods in medical research. 3rd ed. Oxford, England: Blackwell Scientific, 1994.
3
Peto R ,Baigent C . Trials: the next 50 years: large scale randomised evidence of moderate benefits. BMJ 1998;317:1170-1171
CrossRef | Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: Biondi-Zoccai et al. correctly point out that small trials may not be sufficiently powerful to confirm or rule out the presence of small benefits. As they calculate, our recent angiographic trial should have enrolled 318 patients in order to confirm, with 80 percent power, the reduction of roughly 50 percent in the rate of progression of stenosis that we observed. Historically,1 it has been clear that angiographic trials are not ideally suited to the comparison of treatments whose effects are only moderately different. But they do provide insights into mechanisms of action, require many fewer subjects in order to determine whether a treatment affects the progression of disease, can be performed more quickly, cost 1/10th as much as larger clinical trials, and have generally not been proved wrong by the findings of larger trials.2
In this context, the nonsignificant slowing of the progression of disease by 50 percent with the use of antioxidants has seemed curious, particularly since the favorable effect appears to be limited to the subgroup of lesions with stenosis of at least 30 percent of the luminal diameter at base line. Perhaps antioxidants do slow the growth of established lesions yet do not reduce the risk of plaque rupture that results in clinical events. When viewed in the context of several negative trials of vitamin monotherapy, our composite three-year rates of clinical events — 24 percent with placebos and 21 percent with antioxidants alone (P=0.90) — clearly dampened our enthusiasm regarding this curious but weak trend in the progression of stenosis. The recently presented results of the Medical Research Council–British Heart Foundation Heart Protection Study3 demonstrated no effect whatsoever among more than 20,000 subjects randomly assigned either to an antioxidant cocktail similar to ours or to placebo. Thus, even in retrospect, we view our interpretation of our findings as appropriate.
The size of the sample needed to confirm the benefit from a therapy depends on the estimated rate of the end point in each treatment group, the variability of these rates, and the size of the therapeutic effect. For example, penicillin was found to be effective against pneumococcal pneumonia on the basis of results in fewer than 20 patients. In this regard, the combined simvastatin–niacin regimen in our trial is reminiscent of penicillin, although this finding seems to have been lost in the debate about antioxidants.
3 ReferencesB. Greg Brown, M.D., Ph.D.
Alan Chait, M.D.
University of Washington School of Medicine, Seattle, WA 98195Lloyd Fisher, Ph.D.
University of Washington School of Public Health, and Community Medicine, Seattle, WA 981951
Brown G ,Albers JJ ,Fisher LD , et al. Regression of coronary artery disease as a result of intensive lipid-lowering therapy in men with high levels of apolipoprotein B. N Engl J Med 1990;323:1289-1298
Full Text | Web of Science | Medline2
Hennekens CH, ed. Clinical trials in cardiovascular disease: a companion to Braunwald's Heart Disease. Philadelphia: W.B. Saunders, 1999:206.
3
MRC/BHF Heart Protection Study: preliminary results slideshow: 13 November 2001. (Accessed March 14, 2002, at http://www.ctsu.ox.ac.uk/~hps/slides/results/index.htm.)
