Simvastatin ‘Yes,’ Antioxidant ‘No’ — Two Important New Studies
January 1, 2002 9 minutes read
Simvastatin Yes,’ Antioxidant No’—Two Important New Studies
Abstracts & Commentary
Publication of the hdl atherosclerosis treatment study (HATS) in the New England Journal of Medicine follows presentation of preliminary data at the annual American College of Cardiology meeting in March 2001, as well as a companion publication regarding the effects of antioxidants on the niacin-induced elevation of HDL cholesterol.1
HATS is essentially a regression trial with a relatively small number of patients, testing the hypothesis that combination of aggressive lowering of LDL-cholesterol and elevation of HDL-cholesterol would be beneficial on slowing coronary atherosclerosis progression as well as reducing clinical events. Furthermore, providing a combination of antioxidant agents tested the antioxidant hypothesis. Lipid modification was achieved with a statin and niacin. The study used a 4 x 4 factorial design, with 160 patients enrolled between 1995 and 1997. Mean follow-up was 3 years. All patients had to have at least moderate coronary disease at baseline, with a minimum of 3 stenoses of > 30%, or one of > 50%. Lipid parameters included HDL < 35 mg/dL (men) or < 40 mg/dL (women), as well as LDL < 145 mg/dL and triglyceride < 400 mg/dL. Study medications included simvastatin 10-20 mg daily, depending on baseline LDL, with an increase of 10 mg if total LDL was > 90 mg/dL at 1 year. Niacin was given in slow-release formulation, with a target dose of 1 g b.i.d. Further increases of niacin were given if the HDL response was inadequate. The antioxidant vitamin cocktail included vitamin E 800 IU, 1000 mg vitamin C, 25 mg beta-carotene, and 100 mg of selenium. All patients received dietary counseling, food intake analysis, advice to stop smoking, and an exercise program. Multiple lipid moieties were measured as well as plasma levels of vitamin C, E, and beta-carotene. Resistance of LDL to oxidation was assessed. Standard quantitative angiography was carried out with paired films being read in a blinded fashion, using digital magnification and computer methodology. The primary angiographic end point was the mean change per patient in percent stenosis of the most severe lesion in each of 9 proximal coronary segments. The prespecified primary clinical end point was time to first major coronary event or stroke. Patient population was mostly male, relatively young (mean age 53), with approximately half having a prior myocardial infarction or having undergone prior angioplasty.
Results: Compliance was good, ranging between 80-95% in the 4 groups. The mean dose of simvastatin achieved was 13 ± 6 mg per day, and the dose of niacin was 2.4 ± 2 g per day. There were minimal changes in liver function tests with the simvastatin-niacin combination (SN). Flushing was reported with active niacin and placebo to a similar degree. Lipid levels on active therapy: elevation of HDL of 26% in those receiving SN alone, and 18% with SN plus antioxidants (P = 0.05). Baseline LDL decreased from 125 mg/dL by 42%, and triglycerides declined by 36%. Levels of HDL-2 and Lp(A-1) increased markedly, although the responses to SN were blunted by the addition of antioxidants. Plasma vitamin concentrations increased with active therapy. Resistance to LDL oxidation appeared to increase. Percent proximal stenosis was different among the 4 groups. Placebo patients experienced a 3.9% increase in mean % stenosis, compared to a 1.8% increase with antioxidant therapy alone. However, the SN group demonstrated a slight decrease in mean stenosis of 0.4% (P > 0.001); the combination of SN with antioxidants showed a slight, overall increase in mean proximal stenosis of 0.7% (P = 0.004).
Frequency of clinical end points: In the SN group, the composite primary clinical end point was 90% lower than in placebo-treated patients. There was a 60% overall reduction in risk of events in all SN subjects (P = 0.02) compared to "a small insignificant increase in risk" with antioxidants. In summary, the results demonstrate that the combination of simvastatin and niacin resulted in overall coronary stenosis regression, albeit slight, over more than 3 years; this was accompanied by a 90% decrease in the rate of major clinical events when compared to placebo patients (no antioxidants in either group). No benefits of the antioxidant cocktail were found.
The surprising aspect of this study was that the combination of antioxidants with SN appeared to be associated with diminished benefits; there was less regression and more clinical events. This group’s previous publication (Cheung et al, cited in the first paragraph of this article) documents an adverse interaction between the vitamins and elevation of HDL-2 subfractions. Thus, Brown and colleagues raise the possibility that the antioxidant cocktail may actually have impaired the beneficial rise of HDL-2, which was related to both niacin and simvastatin administration. Finally, Brown et al raise the possibility that combination lipid therapy, particularly in individuals with low HDL-C, could provide a major benefit if the large reduction of clinical events in the HATS trial of 60-90% can be documented by subsequent randomized trials. They conclude, "we see little justification for these antioxidant vitamins for the prevention of cardiovascular events." (Brown, G, et al. N Engl J Med. 2001;345:1583-1592.)
The MRC/BHF Heart Protection Study (HPS)
This study reported initial results at the American Heart Association Scientific Sessions in November 2001. The principal investigator, Rory Collins, MD, of Oxford, England, and the Heart Protection Study web site have provided the following synopsis.
HPS was a large trial testing the combination of statin therapy and an antioxidant cocktail in a cohort of individuals with documented coronary artery disease (CAD) or who were high-risk primary prevention subjects. The study was carried out in 169 hospitals in the United Kingdom, enrolling a wide range of patients, aged 40-80, with a total cholesterol of at least 135 mg/dL. No other lipid therapy criteria were necessary. Enrolled subjects had no obvious need for statin therapy or antioxidant vitamins. The population consisted of individuals with documented CAD, diabetes, or hypertension. Between 1994 and 1997, 20,536 individuals were enrolled. The study had a 4 x 4 factorial design, with half the participants receiving 40 mg daily of simvastatin or placebo, and half receiving antioxidant vitamins (600 mg vitamin E, 250 mg vitamin C, and 20 mg beta carotene) or matching placebo. The follow-up period was 5.5 years. Overall, Collins and associates estimated a two thirds compliance rate for the cohort, with 19% of individuals randomized to statin therapy not taking this medication by 5 years, and another 18% of subjects not randomized to statin who were given open-label lipid therapy by trial end by their physicians. The average total cholesterol decrease with simvastatin was approximately 40 mg/dL. Serum levels of administered antioxidants achieved were predicted levels. The HPS primary end point was all-cause mortality; coronary artery disease mortality; and non-CAD mortality. Secondary end points included strokes, CAD events, and non-CAD death.
Results: No antioxidant benefits were noted, with completely neutral findings for all primary and secondary end points. However, no hazard was noted. Lipid lowering was observed in more than 10,000 individuals randomized to the simvastatin, of which 81% were still on the drug after 5 years. There was a 17% decrease in CAD death, a 12% decrease in all-cause mortality, and a decrease in other vascular death. All other CAD events, revascularizations, and strokes were decreased by 24% (P = 0.001). Comparable end point reductions were seen in the approximately one third of the overall cohort who had no vascular disease at entry. All categories of baseline lipid levels experienced comparable event reductions, including those with an LDL under 100 mg/dL; 100-130 mg/dL, and > 130 mg/dL. There was virtually no liver or muscle toxicity. The actual rates of vascular events for placebo-treated patients were 25.4% vs. 19.9% in those who received simvastatin. Curve separation began between 1 and 2 years, and continued to diverge at the end of the study suggesting further benefits if individuals were to continue simvastatin. This large study had far more events than any other lipid-lowering RCT. Collins et al conclude that statin therapy in HPS was shown to be effective in a much wider range of individuals with increased risk for vascular disease than heretofore evaluated, including subjects who only had diabetes, a prior stroke, or peripheral vascular disease. Benefits were comparable in women and men and those individuals older than age 70. No "special" baseline lipid level could be identified for which there was not a benefit. The benefits in individuals with low total and LDL cholesterol remained robust. Collins et al conclude that "all individuals with a history of stroke, peripheral disease or diabetes, regardless of age, sex, or lipid level should be considered for simvastatin, as well as for any individual with a history of myocardial infarction, angina, stroke, or peripheral vascular disease." Collins et al state that the current guidelines or target levels for cholesterol lowering may no longer be appropriate, in that large numbers of individuals with LDL cholesterol levels of < 120 mg/dL benefited in HPS. (MRC/BHF Heart Protection Study. www.hpsinfo.org. Accessed Dec. 12, 2001.)
Comment by Jonathan Abrams, MD
HPS and HATS, particularly the former, provide major new information pertaining to the prevention of vascular diseases. The huge UK study is of great significance, as it expands the boundaries of those individuals who should be considered for statin therapy. This is the largest database of any statin trial in women. Furthermore, many patients in HPS were older than 70 or 80 years of age, and they demonstrated comparable benefit to those younger. These age groups have not been previously well represented in statin randomized controlled trials (RCTs). It is not a particular surprise that patients with hypertension or diabetes who otherwise do not have overt vascular disease benefit from lipid lowering, but this has not been documented before. Until the HPS trial results are published, it is difficult to know the precise risk factor profile of individuals with and without known vascular disease. It will be of great interest to see whether there are gradations of risk reduction depending on the baseline risk profile. Nevertheless, for the common type-II middle-age diabetic who has not had a vascular event, but may have hypertension or some degree of dyslipidemia, it now appears reasonable to add a statin to the preventive regimen (which should include an ACE inhibitor) with the hope of decreasing subsequent vascular events. HPS confirmed that stroke reduction with statins is a real phenomenon, previously suggested by the statin RCT, and now confirmed with much larger numbers of individuals. Collins et al are on somewhat shaky ground by extrapolating the risk reduction to one third, based on the proportion of individuals who went off statin therapy by trial end, and the proportion of placebo patients who went on open label statin. Thus, the overall nonfatal vascular event reduction was 24% in all categories, using an intention-to-treat analysis. One might wonder why the National Cholesterol Education Program Adult Treatment Panel (ATP-III) did not wait for the results of HPS before publishing the new guidelines, which are already somewhat out of date if one accepts the HPS results as the new gold standard. The antioxidant data from HPS are sobering but are of no surprise after the results of HOPE and GISSI-P, confirming no objective benefit of antioxidants on decreasing vascular events. In the aggregate, more than 35,000 patients have been randomized to antioxidants or placebo, with no benefit demonstrated, but also no harm. The latter is important because of the suggested results from the small HATS trial, which suggest an adverse effect of antioxidants on HDL-C metabolism and possibly on vascular protection. If the hypothesis of the HATS investigators is followed to the extreme, one would suspect that antioxidant-treated patients would do worse than placebo over a long period of observation; this has not been the case in HOPE, HPS, or GISSI-P, and many experts still believe that the oxidation hypothesis is alive and well, but that the appropriate and effective antioxidant moieties have not been identified. Furthermore, few would advocate that a healthy diet with leafy greens and fruit, the major source of natural antioxidants, should be abandoned simply because synthetic vitamin supplements have been a major disappointment.
The HATS data are important by demonstrating that the combination of aggressive LDL-C lowering and HDL-C elevation represents an extremely potent approach to vascular disease. The fact that there was global CAD regression in the cohort who received simvastatin and niacin alone with virtually no clinical events, is quite impressive. The addition of antioxidant cocktail to the simvastatin-niacin combination seemed to attenuate the benefits of the lipid active drugs alone. However, the cell sizes in HATS were small, and one must be cautious before extrapolating the HATS results universally. Brown et al’s gentle call for a larger trial makes sense in this regard. Nevertheless, they have apparently identified a mechanism whereby supplemental antioxidant vitamins may be detrimental to the desired effects on HDL physiology by blunting A1 and HDL-2 production. One would surmise that if this is a clinically important phenonomen, the huge HPS database comparing 10,000 subjects receiving antioxidants compared to placebo, would have suggested an adverse outcome on coronary events, and this was not found.
The HATS data should be looked at in light of the anticipated release of Advicor, a product that will include Niaspan in combination with lovastatin, and is expected to be released later next year. Studies have indicated that at high doses of lovastatin and Niaspan, very impressive decreases in LDL, as well as increases in HDL levels, can be achieved. The HATS findings suggest that this type of combination drug therapy could be important for clinical atherosclerosis.
Reference
1. Cheung M, et al. Artheroscler Thromb Vasc Biol. 2001; 21:1320-1326.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content