Department of Medicine (R.A.K.), University of South Alabama, Mobile, Alabama 36688-0002; and Department of Medicine (A.O.), Division of Preventive Medicine, The University of Alabama at Birmingham, Birmingham, Alabama 35205-4785
Address all correspondence and requests for reprints to: Robert A. Kreisberg, M.D., Department of Medicine, University of South Alabama, CSAB 170, Mobile, Alabama 36688-0002. E-mail: rkreisberg{at}usamail.usouthal.edu
Despite evidence from population studies and clinical trials dating to 1984 (1), The Lipid Hypothesis, the idea that a reduction in blood cholesterol reduced coronary heart disease (CHD) risk was viewed with skepticism. Subsequently, angiographic studies using a range of drug and lifestyle regimens demonstrated that progression of coronary atherosclerosis could be reduced and clinical cardiac events prevented. A new and expanded opportunity for the treatment of hypercholesterolemia and prevention of CHD emerged with the use of statins. The landmark Scandinavian Simvastatin Survival Study (4S) was the first randomized, controlled trial to convincingly demonstrate that coronary events and total mortality were decreased by a reduction in low-density lipoprotein cholesterol (LDL-C) (2). Within the past decade, major clinical endpoint trials (2, 3, 4, 5, 6, 7), enrolling more than 30,000 patients, have emphasized the value of LDL-C reduction. This reduction in CHD is particularly noteworthy because total and LDL-C reduction were accomplished with a variety of drugs (statins, fibrates, and bile acid-binding resins), dietary changes, lifestyle modifications, and surgery (ileal bypass), indicating that reduction in cholesterol and LDL-C was more important than how it was achieved.
CHD events, cardiovascular morbidity and mortality, and all-cause mortality can be prevented with statins or fibrates in patients with average or elevated LDL-C who have preexisting CHD
The totality of evidence derived from early angiographic studies and subsequent major clinical event trials with statins and adjunct dietary therapy unequivocally establishes the value of secondary prevention. Although there is considerable biologic plausibility for a direct effect of LDL-C in the atherogenic process (8, 9, 10), it is distinctly possible that other properties of statins and fibrates contribute to the reduction in CHD (11, 12).
Ileal bypass trial.
As early as 1990, the Program on the Surgical Control of the Hyperlipidemias (POSCH) (13) reported the efficacy of partial ileal bypass for reducing total cholesterol and LDL-C and CHD events. Ileal bypass reduced the LDL-C level by 38% and increased high-density lipoprotein cholesterol (HDL-C) by 4% compared with the control group in this randomized trial, whereas very low-density lipoprotein cholesterol (VLDL) and triglyceride levels increased 8 and 20%, respectively. Over 5 yr, there was a 21% decrease in overall mortality, a 28% decrease in CHD mortality, and a 35% decrease in the combined endpoint of CHD death and nonfatal myocardial infarction (MI). The POSCH trial also established a correlation between angiographic changes and clinical events (14).
Angiographic studies.
The number of patients enrolled in this type of study has ranged from less than 50 to more than 1000, and the intervals between angiographic evaluations have ranged from 1 to 10 yr. When collectively analyzed, these studies demonstrated plaque stabilization, reduced progression of atherosclerotic lesions, and fewer new lesions with cholesterol reduction (15, 16). Notably, in these studies clinical events were reduced regardless of the intervention used. Interestingly and importantly, the reduction in clinical events was much greater than expected from the demonstrated angiographic changes. This led to the concept that plaque composition was more important than size and that these regimens changed the biology of atherosclerosis by improving endothelial function and by reducing lipoprotein oxidation, foam cell formation, metalloproteinase activity, and inflammation, thereby stabilizing the nonocclusive, culprit lesion.
Statin trials.
The 4S (2), the first of the major statin trials, provided definitive evidence linking LDL-C reduction to secondary CHD prevention (Table 1). The 4444 participants with CHD and elevated LDL-C levels (mean, 188 mg/dl) received 20 to 40 mg of either simvastatin or placebo daily for 5.4 yr. LDL-C was reduced by 38%, all-cause mortality (the primary endpoint) by 30%, and major coronary events by 34%. The 4S clearly established the efficacy and safety of long-term statin treatment of patients with CHD. Shortly thereafter, the Cholesterol and Recurrent Events (CARE) and Long-Term Intervention with Pravastatin in Ischemic Disease (LIPID) trials extended these results by demonstrating benefit in patients with relatively average total cholesterol and LDL-C levels and CHD (4, 6). The CARE trial randomized 4159 participants with prior MI and LDL-C levels ranging from 115 to 175 mg/dl (4). Participants received 40 mg/d of pravastatin or placebo for 5 yr. Death from CHD or nonfatal MI decreased by 24%. Patients with a baseline LDL-C exceeding 150 mg/dl had a 35% reduction in coronary events, compared with a 26% reduction in those with baseline values of 125150 mg/dl and a 3% increase in events in those with baseline levels less than 125 mg/dl. Patients with previous unstable angina were as likely to benefit from pravastatin therapy as were participants with stable CHD (17). In similar fashion, the LIPID trial also evaluated treatment of coronary patients with moderately elevated LDL-C levels. Pravastatin (40 mg) was compared with placebo in 9014 CHD patients over a 6-yr period (6). Median LDL-C decreased from 150 to 112 mg/dl. Major coronary events were reduced by 29%, coronary deaths by 24%, and total mortality by 23%. The true long-term benefit of statin therapy is undefined but may be greater than achieved in these studies, which lasted approximately 5 yr.
|
There is universal agreement and acceptance that LDL-C lowering is cost-effective for secondary prevention of CHD, but its value for primary prevention is more controversial. Primary prevention of CHD is crucial because more than half of first coronary events are fatal, and half of these occur in persons not previously known to have CHD (18). Furthermore, secondary prevention is cost-effective only for those who survive their first event. Therefore, prevention of first events is crucial. Reluctance over reducing LDL-C for the purpose of primary prevention is due to concern about maintaining therapy for many years, the belief that cholesterol can be treated later in life, and lower cost effectiveness than with secondary prevention, as well as the mistaken impression (in our opinion) that hypercholesterolemia is unlikely to be important in older patients without clinical CHD. However, atherosclerosis begins in childhood or adolescence (19, 20) and progresses at a variable rate until it reaches the clinical threshold. The less severe stenotic lesions are perhaps the most dangerous. It is unlikely that later therapy will have as great an impact as earlier therapy. In fact data from the statin intervention trials reveal a death rate from CHD with therapy that is still much greater than observed in patients without hypercholesterolemia. There also is an excess risk of death from cardiovascular disease (CVD) and decreased life expectancy among those less than 40 yr of age with elevated cholesterol levels (21). It is now clear that there are patients without clinical CHD who are at even greater risk of events than those with established CHD. Assessment of global risk and nontraditional risk factors with available tools (22, 23) provides a more definitive assessment of risk and accentuates the benefit/risk ratio. Furthermore, because CHD is the leading cause of death in the elderly and older persons are at higher absolute risk, LDL-C reduction in this subgroup is associated with a greater reduction in absolute risk and consequently greater cost-effectiveness.
In the West of Scotland Coronary Prevention Study (WOSCOPS), 6595 men with LDL-C greater than 155 mg/dl on two occasions or greater than 175 mg/dl at least once were treated with 40 mg pravastatin or placebo for 5 yr (5). The primary endpoint of nonfatal MI or CHD death was reduced by 31%, and total mortality was reduced by 22% (P < 0.052). The Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS) evaluated the role of lovastatin (2040 mg) or placebo for primary CHD prevention among 6605 patients whose LDL-C concentration was mildly to moderately elevated (7). Entry criteria were an LDL-C of 130190 mg/dl or 125129 mg/dl with total cholesterol/HDL-C ratio greater than 6.0 and HDL-C less than 45 mg/dl in men and less than 47 mg/dl in women. The primary endpoint of fatal or nonfatal MI, unstable angina, or sudden cardiac death was reduced by 37%, and fatal or nonfatal MI alone was reduced by 40% after a mean of 5.2 yr. These risk reductions were independent of baseline LDL-C values, indicating that primary prevention can be accomplished in patients with mildly elevated as well as high LDL-C levels. If the AFCAPS/TexCAPS entry criteria are used to select from the general population, substantial numbers of individuals not currently recommended for drug therapy by the National Cholesterol Education Program (NCEP) guidelines would be eligible for treatment (24). A recent meta-analysis of randomized trials demonstrated the value of primary prevention in reducing coronary events (25). Total mortality also was less with treatment, but not significantly. However, the numbers did not provide adequate power.
The lower the LDL-C, the better. But the relation of LDL-C to CHD is curvilinear; the benefits diminish as LDL-C approaches lower levels
Unfortunately, the major clinical trials report conflicting results on desirable target levels for LDL-C. The relation between cholesterol levels and risk is curvilinear (26) so that the law of diminishing returns applies. It is highly unlikely that there is an absolute LDL-C threshold below which no benefit occurs. Actually, the curvilinear relationship between cholesterol and CHD consists of a family of individual doseresponse curves. Those at highest CHD risk experience the greatest benefit from the decrease in LDL-C and plateau at a lower LDL-C level than those at lower overall risk. Conversely, persons at the lowest risk have a less steep curve and plateau at higher LDL-C levels. Numerous prospective studies confirm the curvilinear relationship of LDL-C to CHD and support the concept that the optimal LDL-C value is 100 mg/dl (27, 28, 29). Furthermore, reduction of LDL-C to less than 100 mg/dl improves endothelial function and blood flow (30). In early prospective studies, the threshold of benefit occurred at a total cholesterol level of about 200 mg/dl, corresponding to a LDL-C of 130 mg/dl (26, 31). Other, larger studies (22) have demonstrated a continuous relationship over a wide range of cholesterol levels with progressive reduction of risk to a total cholesterol of 150 mg/dl, which corresponds to an LDL-C of 100 mg/dl. The POSCH trial clearly demonstrated that therapy directed exclusively at LDL-C was effective even when VLDL-C and triglyceride increased, both of which would be expected to reduce the benefit of LDL-C lowering (13). These results emphasize the value of reducing LDL-C, independent of the potential beneficial pleiotrophic properties of lipid-lowering agents (32). These conclusions are supported by an extensive body of experimental data indicating that LDL-C is atherogenic and that the inflammatory response of atherosclerosis depends on the atherogenicity of LDL-C, the number of LDL-C particles, and the pro-inflammatory properties of LDL-C (33, 34).
Other studies have demonstrated a direct relation between cholesterol level and endothelial-dependent vasodilation of coronary arteries without obstructive disease (10). Improved blood flow occurs when LDL-C is reduced from 130 to less than 100 mg/dl, providing support for the NCEP Adult Treatment Panel (ATP) III guidelines and goals of therapy in patients with established CHD (24). Statin trials using clinical endpoints also support the importance of lowering LDL-C. The highest correlation with coronary event reduction occurs with pretreatment LDL-C levels, especially those with concentrations greater than 160 mg/dl (27). In the 4S, the relationship between LDL-C lowering and a reduction in coronary events was curvilinear and continuous to concentrations of less than 100 mg/dl, although this was not observed in the CARE and LIPID studies.
The Post Coronary Artery Bypass Graft trial compared the effects of aggressive LDL lowering with modest LDL-C lowering in patients with previous coronary artery bypass surgery (35). The LDL-C levels achieved were 9397 mg/dl for 80 mg lovastatin and 132136 mg/dl with 2.55.0 mg lovastatin. Angiographic progression of arteriosclerosis in graft vessels was reduced by 39% in the aggressively treated group and by 27% in the less aggressively treated group. Extended follow-up data from this study at 7.5 yr revealed a 24% reduction in composite clinical endpoints and a 30% reduction in revascularization procedures in the aggressively treated group (36). Aggressive treatment with lovastatin in this study delayed graft atherosclerosis in women, patients older than 65 yr, and patients with diabetes mellitus, low HDL-C, hypertriglyceridemia, hypertension, and smoking. This study and the Cholesterol Lowering Atherosclerosis Study (CLAS) suggest that aggressive cholesterol reduction is more effective than mild to moderate reduction (37). The results of the Atorvastatin vs. Revascularization Treatment trial, though not conclusive, further support the concept that aggressive LDL-C lowering is better than modest lowering (38).
The "lower is better" concept for LDL-C is controversial because, in the CARE trial, coronary events decreased as LDL-C was reduced until a level of 125 mg/dl was reached (39). Below 125 mg/dl there was no additional risk reduction (4). An important limitation of the CARE analysis is that this was a postrandomization analysis (post hoc study) in which only 20% of the patients had on- treatment levels of LDL-C less than 125 mg/dl. Consistent with CARE, there was only a 16% reduction in risk in patients with baseline LDL-C less than 135 mg/dl in the LIPID trial (6). In the WOSCOPS, LDL-C was reduced by a mean of 25% from 192 mg/dl to 156 mg/dl with a 24% reduction in clinical events (5). No additional clinical benefit occurred in the subset of patients with greater LDL-C reduction.
There are many reasons that these studies appear to conflict. Differences in baseline LDL-C and global risk in these trials could influence outcomes, as could coexistent nonlipid risk factors, particularly when LDL-C is reduced to less than 130 mg/dl. It has been suggested that a reduction in LDL-C of about 25% produces maximum benefit in moderate risk populations, whereas greater reductions are required for high-risk dyslipidemic populations (40).
Clearly, this remains an important issue to resolve, but considering its controversial nature, it is prudent and reasonable to adhere to the goals set by NCEP ATP III (24). There seems to be little risk from aggressive lowering of LDL-C values to the levels proposed by ATP III, although subgroup findings from the Multiple Risk Factor Intervention Trial suggest a greater risk of hemorrhagic stroke in those with LDL-C less than70 mg/dl (41). The Honolulu Heart Study (42) also related hemorrhagic stroke to lower nontreatment cholesterol levels in hypertensives. Although this is obviously of some concern, these levels are rarely achieved, and the absolute increase in hemorrhagic stroke is insignificant compared with absolute reduction in CHD events. Even an LDL-C level of 100 mg/dl may be inadequate to reduce the risk of high-risk patients to that of individuals free of CHD (43). Results from unpublished studies currently in progress [The Study to Evaluate Additional Reduction of Cholesterol and Homocysteine, Treating to New Targets, and Oxford trials] of aggressive lipid lowering may answer this question. The issue of whether lower is better can be resolved only by properly designed randomized clinical trials that address risk reduction at designated LDL-C ranges and global risk levels.
Lipoproteins other than LDL-C can influence the development of coronary atherosclerosis and its response to therapy. It is no longer appropriate to direct treatment goals exclusively to total and LDL-C
The benefits of LDL-C lowering therapy are not uniform across a population of patients with CHD. This is evident by the fact that statins reduce CHD events by only approximately 30% and that there is a residual risk of approximately 70% among patients receiving therapy. Thus, many of the patients remain vulnerable to CHD events. For example, therapy in the 4S reduced the relative risk of CHD events over 5 yr by approximately 30% from an absolute risk of 28% to 19%. The event rate in the placebo group was a 5.2% per year and in the treatment group, 3.8%. Because a rate of 23% per year is considered high risk, the treatment group remained at an unacceptably high risk of CHD events. Further study is required to determine whether persisting high risk despite therapy is related to suboptimum LDL-C lowering, the presence of other untreated lipid abnormalities, inadequate treatment of coexistent standard CHD risk factors, or the presence of important but less well understood conditional risk factors (44, 45).
There is convincing evidence that HDL-C is a powerful negative risk factor (46). Plausible mechanisms for its protective effects include reverse cholesterol transport and antioxidative properties of HDL. Other lipids also have been linked to CHD, but the issue of causality is not yet resolved. Substantial evidence implicates the atherogenic properties of triglyceride-rich lipoproteins (TRLs) directly and indirectly, and that they should be targeted for intervention (47). Small TRL enter the intima and initiate events that promote the atherosclerotic process. Hypertriglyceridemia is associated with reduced levels of HDL-C and increased levels of small dense LDL particles. Of the three components of this atherogenic lipid triad (low HDL-C, high triglycerides, and small dense LDL), only increasing low levels of HDL-C has been subjected to a clinical trial and shown to reduce CHD events (48).
HDL-C.
The initial NCEP ATP (NCEP ATP I) recognized a low HDL-C as a CHD risk factor and a high HDL-C (>60 mg/dl) as protective (49). The NCEP ATP II did not propose HDL-C treatment guidelines because of the lack of data from therapeutic trials (50). Data from the Helsinki Heart Study (HHS), a 5-yr randomized trial, indicated that changes in LDL-C did not completely explain the CHD reduction (51). More than 80% of the risk reduction occurred in men whose LDL/HDL-C ratio exceeded 5 and whose triglyceride level was more than 200 mg/dl. HDL-C subgroup analyses from this early primary prevention trial suggested that a 1 mg/dl increase in HDL-C achieved with gemfibrozil treatment was independently associated with a 23% reduction in CHD endpoints (44, 52). This is similar to observational studies in which an increase of 1 mg/dl in HDL-C was associated with a 23% reduction in CHD. This has now been confirmed by the Veterans Affairs Cooperative Studies Program High Density Lipoprotein Cholesterol Intervention Trial (VA-HIT) (48).
Recently, major outcome trials using fibrates have assessed the impact of increasing HDL-C in coronary patients (53). The VA-HIT used 1200 mg of gemfibrozil or placebo in 2531 men with CHD for 5 yr (48). Gemfibrozil increased HDL-C levels by 6% (2.8 mg/dl) while reducing the primary endpoint of nonfatal MI and MI death by 22%. Eligibility criteria for the VA-HIT included an HDL-C level of no more than 40 mg/dl and an LDL-C of no more than 140 mg/dl. The mean baseline LDL-C level was 111 mg/dl and did not change over the 5.1 yr of treatment. The reduction of triglyceride by 31% in the VA-HIT raised the question of whether reduced CHD risk was due to a decrease in atherogenic VLDL (triglyceride) and/or a decrease in small dense LDL particles. A detailed analysis of these data revealed an independent association between increased HDL-C and CHD reduction but not decreased levels of triglycerides (54). Nonetheless, changes in HDL-C in the VA-HIT accounted for only 23% of the CHD reduction. In the Bezafibrate Coronary Atherosclerosis Intervention Trial, bezafibrate reduced progression of coronary atherosclerosis in arteries with less than 50% narrowing (55). The Bezafibrate Infarction Prevention trial, followed 3122 patients with serum cholesterol levels of 180250 mg/dl, normal triglyceride levels, and HDL-C less than 45 mg/dl. Although there was a favorable trend in coronary events, the difference after 3 yr was not statistically significant. The primary endpoint of CHD death or nonfatal MI was significantly reduced only in those patients whose initial triglyceride was greater than 200 mg/dl.
It is clear that any CHD benefit arising from the Bezafibrate Infarction Prevention trial and VA-HIT must be accounted for by factors other than the reduction in LDL-C or the increase in HDL-C. It is premature to attribute all of the benefit to increased levels of HDL-C because the absolute increase in HDL-C with therapy is not sufficient, based on epidemiologic data, the HHS, and the VA-HIT, to account for all of the reduction in CHD (1 mg/dl HDL-C associated with a 23% reduction in CHD). In the AFCAPS/TexCAPS, impressive clinical benefits occurred in the subgroup of patients who had low levels of HDL-C. In fact, in all studies except the Lipid Research Clinics-Coronary Primary Prevention Trial (3), patients with low HDL-C tend to have the greatest benefits from treatment directed at LDL-C reduction (52).
These observations are consistent with those discussed in the Lipoprotein and Coronary Atherosclerosis Study (LCAS) in which on-trial levels of apoprotein A-I and apoprotein B were the best predictors of progression of coronary atherosclerosis (56). LCAS is remarkable for several reasons. It enrolled only 339 patients with established CHD who were randomly assigned to diet plus placebo or diet plus fluvastatin. Progression of atherosclerosis was reduced the most by fluvastatin among 68 patients with HDL-C less than 35 mg/dl; clinical events also were significantly reduced only in those with HDL-C less than 35 mg/dl. Benefit from therapy occurred in LCAS whether the LDL-C was above or below 160 or 130 mg/dl. In the Lopid Coronary Angiographic Trial, gemfibrozil inhibited progression of coronary atherosclerosis in men with low HDL-C levels who had previously undergone coronary artery bypass grafting (57). Angiograms were performed at entry and after 32 months of therapy. Gemfibrozil reduced triglyceride by approximately 40% and LDL-C by approximately 12% and increased HDL-C by approximately 11% (4 mg/dl). These observations support the concept that an increase in HDL-C has important antiatherogenic benefits and is consistent with observations in the VA-HIT in which mean baseline LDL-C levels of 110 mg/dl were not changed by gemfibrozil.
A continuing question in angiographic studies and clinical trials is whether a selective increase in HDL-C reduces angiographic progression of coronary atherosclerosis or clinical CHD events. This is a difficult question to answer because none of the currently used pharmacologic agents acts exclusively on HDL-C. Statins reduce LDL-C and triglycerides and increase HDL-C slightly. Fibrates have a minor effect on LDL-C, reduce TRL and triglycerides, and modestly increase HDL-C in patients with hypertriglyceridemia but only slightly increase HDL-C when triglyceride is normal. Niacin reduces LDL-C, triglyceride, and lipoprotein (a) [Lp(a)] and also is the most potent agent for increasing HDL-C. Niacin was an integral component of three angiographic studies: CLAS (37), Familial Atherosclerosis Treatment Study (FATS) (58), and HDL-Atherosclerosis Treatment Study (HATS) (59).
In the CLAS, treatment with colestipol and niacin reduced LDL-C by 43% and increased HDL-C by 37%, reducing events by 25% (37). In FATS, conventional therapy was compared with treatment with lovastatin plus colestipol and to niacin plus colestipol in men with CHD and apoprotein B levels greater than 125 mg/dl (58). The multivariate equations in the FATS indicate that the change in HDL-C was as important as the change in apoprotein B for predicting improvement in stenosis. The HDL-Atherosclerosis Treatment Study was designed to test whether niacin plus simvastatin was better than simvastatin alone in patients with CHD and low HDL-C (59). It was a factorial trial evaluating the combination of niacin and simvastatin with and without antioxidant supplements. The niacin plus simvastatin combination reduced CHD events by 70%, whereas simvastatin alone reduced clinical events by 25 to 35% over 3 yr.
Triglyceride.
Whether hypertriglyceridemia is directly related to CHD or whether it is a marker for a cluster of CHD risk factors remains unsettled (60, 61). Nevertheless, treatment of hypertriglyceridemia and more specifically increased levels of TRLs makes better sense than treating more speculative lipid abnormalities. The POSCH trial demonstrated that coronary artery progression was related to VLDL-C, but that a reduction in events occurred despite an increase in triglyceride concentration (13). In the CARE trial, patients with baseline triglyceride concentrations less than 146 mg/dl had a greater reduction of clinical CHD events with pravastatin. In addition, the concentration of VLDL-C particles and of apo-CIII (an inhibitor of lipoprotein lipase) in VLDL-C and LDL-C were more specific indicators of risk than plasma triglyceride (62). If the major lipid trials are plotted to relate cholesterol lowering to reduction in CHD events, all the cholesterol lowering trials fall on a straight line (63) (Fig. 1). However, the fibrate trials reduce CHD events at a much lower reduction in cholesterol. Possible explanations for this divergence include the triglyceride/HDL-C changes, change in LDL particle size, or the effect on thrombotic factors. In CLAS, the content of apo-C III was directly related to progression of coronary atherosclerosis (37). Baseline triglyceride levels did not influence outcome in the 4S (2) but did in the CARE (4) and the VA-HIT study (54). These findings suggest that TRL and triglyceride may be of greater importance when LDL-C has been optimally reduced or is only mildly elevated at initiation of therapy.
|
Triglyceride influences LDL size with a shift from large buoyant to small dense particles primarily over the relatively narrow range of 80250 mg/dl (60). In some prospective studies, LDL particle size independently related to CHD (64, 65). In the St. Thomas Atherosclerosis Regression Study, patients with coronary atherosclerosis were randomized to receive usual care, or treatment with diet alone or diet plus cholestyramine (66). Angiograms were obtained before and after an interval of 38 months. Changes in coronary atherosclerosis (expressed as mean arterial width and minimum arterial width) correlated best with change in dense LDL. This is of considerable interest because the presence of dense LDL identified patients most likely to benefit from therapy in the Stanford Coronary Risk Intervention Program (67). Usual care was compared with multiple risk factor intervention. After 4 yr, progression of coronary lesions was 40% greater in those with dense LDL than with buoyant LDL. In contrast, progression of coronary atherosclerosis was reduced by 79% in patients with dense LDL who received multiple risk factor intervention. This suggests that patients with CHD and dense LDL may experience the greatest benefit from LDL-C-lowering therapy.
Lp(a).
Although some studies have not found a link between Lp(a) and CHD, most prospective, observational studies have found a significant relationship (23, 68, 69). Lp(a) is only weakly related to the major vascular risk factors, and its risk association is unlikely to be accounted for by those factors. Elevated baseline Lp(a) levels and reduction of Lp(a) in the Heart and Estrogen/Progestin Replacement Study (HERS) identified women who experienced delayed benefit from hormone replacement therapy (HRT) (70). Genetic factors are the major determinant of Lp(a) size and concentration. Because no current treatment is practical in reducing Lp(a) values, except nicotinic acid or HRT, randomized trials have not been conducted specifically to assess the causal relationship of Lp(a) to CHD.
Prompt therapy with statins may reduce morbidity and mortality after an acute coronary event
The old atherosclerosis paradigm conceptualized the plaque as a focal stenosis that progressed to an occlusive lesion and, ultimately, to an acute coronary event. Angiography has shown that most of the patients with a MI have predisposing lesions that are nonocclusive (<50% of the coronary artery lumen) (71, 72). Vulnerable plaques are nonobstructive lesions that are asymptomatic until the cap ruptures or erodes. The pathologic characteristics of these unstable plaques now are well understood and have been described in detail (8, 72, 73). Angiographic studies following lipid modification have demonstrated that reductions in clinical events occur with minimal change in luminal stenosis. The reduction in CHD events has been attributed to compositional changes and stabilization of these nonobstructive lesions preventing rupture with subsequent thrombus formation (74, 75). Lipid-lowering therapy also can rapidly improve endothelial function and reverse ischemia (10, 76, 77). These data have stimulated interest in treating acute coronary syndromes (unstable angina, nonQ wave MI, and Q wave MI) with statins (10). The majority of CHD patients have dyslipidemia, although the LDL-C decreases during an acute coronary event and may take several months to stabilize. Early studies suggested that statins may stabilize these plaques and prevent rupture with subsequent thrombus formation. An observational study revealed a 25% reduction in mortality after acute MI in patients started on statin therapy before discharge from the hospital (78). In the Pravastatin with Thrombolytic Therapy Trial, patients received thrombolytic therapy plus 40 mg pravastatin or placebo (79). After 6 months, patients receiving pravastatin had a lower rate of nonfatal MI and in-hospital mortality than the group on thrombolytic therapy alone. Recently, two comprehensive studies of the early initiation of statin therapy in acute coronary syndromes have been reported, the Fluvastatin on Risk Diminishing After Acute Myocardial Infarction (80) and Myocardial Ischemia Reduction with Aggressive Cholesterol Lowering (81) trials. In the Fluvastatin on Risk Diminishing After Acute Myocardial Infarction trial, 540 patients with acute MI were randomized to fluvastatin or placebo. Major adverse cardiac events and residual ischemia were similar in those treated with fluvastatin or placebo after 6 wk or 12 months. However, the study was not powered for this purpose (80). The Myocardial Ischemia Reduction with Aggressive Cholesterol Lowering study included 3086 patients with nonQ wave infarction randomized to atorvastatin or placebo within 4 d of infarction (81). The composite endpoint of death, recurring MI, cardiac arrest with resuscitation, or worsening of angina requiring hospitalization at 16 wk was 14.8% in the atorvastatin group and 17.4% in the placebo group, a statistically significant 16% reduction in risk. It should be noted that the reduction in the prevalence of worsening angina was the only component of this primary endpoint reaching statistical significance.
Further ongoing randomized trials will amplify the interpretation of these encouraging studies and hopefully provide data on the implications for management. Unpublished ongoing acute coronary syndrome trials include the Aggrastatin to Zocor Trial in which the A phase analyzes anticoagulants and the Z phase analyzes treatment with Zocor. The Pravastatin or Atorvastatin Evaluation in Infection Therapy was a factorial trial designed to show the clinical equivalence of pravastatin and atorvastatin in reducing all-cause mortality or major cardiovascular events after treatment of approximately 24 months in patients with acute coronary syndromes. Another primary objective was to demonstrate the value of the antimicrobial agent gatifloxacin compared with placebo in reducing the major outcomes. The Prevention of Reinfarction by Early Treatment of Cerivastatin Study will test the effects of placebo and two doses of cerivastatin in 3000 patients hospitalized for an acute MI. The placebo arm will be discontinued after 3 months, and patients will be placed on one of two doses of cerivastatin and followed for 2 yr.
Nonlipid properties of the statins may play a role in plaque stabilization and reduction of CHD events
The concept that statins have multiple direct or nonlipid antiatherosclerotic properties is gaining wide acceptance. Effects unrelated to reduction of LDL-C are termed pleiotropic. The search for pleiotropic effects of statins was stimulated by the findings that statin therapy can promptly change clinical outcomes and that the reduction in clinical events is much greater than would be expected from the change in luminal size. As a result, it was suggested that statins act mainly on plaque stability and progression. Pleiotropic properties are vasodilative, antioxidant, anti-inflammatory, immunomodulatory, antithrombotic, antiproliferative, and plaque stabilizing (11). The potential impact of these pleiotropic effects on the atherosclerotic plaque is extremely complex, and it is unclear which one(s) is or are important. The direct or nonlipid effects of statins may complement those due to cholesterol reduction. The inhibition of 3- hydroxy-3-methylglutaryl coenzyme A reductase by statins occurs at an early step in the metabolic pathway of cholesterol synthesis, thereby influencing the synthesis of important metabolic intermediates beyond mevalonate that regulate prenylation. In a substudy of CARE, it was shown that pravastatin was associated with improved endothelium-dependent vasodilation (82). Lovastatin, pravastatin, and simvastatin reduce transient myocardial ischemia, further supporting improved endothelial function. Other studies clearly demonstrate the anti-ischemic effects of statin therapy (reducing the size and severity of perfusion abnormalities and reversal of ambulatory ECG ischemic changes) (16, 76, 77). Reduced susceptibility to ex vivo oxidation of LDL-C has been demonstrated with lovastatin (83), simvastatin (84, 85), pravastatin (84), and fluvastatin (86).
Inflammation.
Inflammation is an integral part of the atherosclerotic process (8). A number of markers of inflammation, such as highly sensitive C-reactive protein (hsCRP), serum amyloid A, white-cell count, interleukin-6, and other cytokines may allow more precise assessment of individual risk. In the CARE trial, a subgroup of 391 participants who developed recurrent MI were compared for hsCRP levels with age- and sex-matched participants who did not (87). Those in the highest quintile of hsCRP had a relative risk of a recurrent ischemic event that was 75% higher than those in the lowest quintile. Pravastatin therapy reduced hsCRP by a median of 17%. More importantly, those participants with a high level of hsCRP experienced a 32% reduction in events compared with the overall reduction of 28%. The benefit of statin therapy was unrelated to LDL-C and occurred almost exclusively in CHD patients whose hsCRP was in the uppermost range. These results imply that pravastatin has an anti-inflammatory effect that provides clinical benefit. In addition, baseline quartile of hsCRP in the AFCAPS/TexCAPS was related to risk of acute coronary events; lovastatin effectively reduced hsCRP. Atorvastatin, cerivastatin, pravastatin, and simvastatin all reduce hsCRP (88, 89). Pravastatin reduces killer T-cell cytotoxicity in patients receiving cardiac transplants (90). This immunomodulatory effect was potentiated by cyclosporin and associated with decreased activity of cytotoxic T-lymphocytes. Although transplant vasculopathy is different from atherosclerosis, the same inflammatory mediators may determine plaque vulnerability. Other experimental studies have demonstrated that statins have anti-inflammatory properties. Recent data on the immunologic effects of statins corroborates the theory that statins decrease CVD risk independent of lipid-lowering effect (91).
Antithrombotic effects.
An occlusive thrombus not only causes acute MI but also underlies the pathogenesis of unstable angina. Accumulating evidence supports an important role for lipids and lipoproteins in the blood coagulation cascade. Lipid lowering drugs are believed to further reduce CHD risk by altering the prothrombotic state that accompanies disruption of a plaque. Statins influence tissue factor expression and platelet aggregation, but the effects differ among the agents used (12, 75). Reports on changes in fibrinogen concentration with statin therapy have been inconsistent. A recent head-to-head study revealed that lovastatin, pravastatin, simvastatin, and atorvastatin had little effect on fibrinogen (92). Fibrates, with the exception of gemfibrozil, reduce serum fibrinogen, suggesting that these changes may be independent of triglyceride lowering (12, 93). Yet, gemfibrozil reduces serum viscosity in types IV and V hyperlipoproteinemia (93). These antithrombotic effects may account for the reduction in CHD events despite minimal LDL-C lowering. The influence of statins on PAI-1 levels appears to be highly variable (75). The differential effects of the lipid-lowering agents, despite similar effects on lipids, suggest unique properties that require a better understanding of the molecular mechanisms involved in thrombosis. At this time, it is unclear how statins influence these complex pathways. Consequently, there is no clear picture of how the balance between the deposition of fibrin and fibrinolysis is achieved.
Plaque stabilization and cell proliferation.
The vulnerability of the small culprit atherogenic plaque to rupture is related to the size of the lipid core, the thickness of the fibrous cap, and the presence of macrophages and foam cells. Many of the changes in the composition and nature of the plaque that occur with statins are due to a reduction in LDL-C and lipid depletion of the lesion. All statins except pravastatin inhibit smooth muscle cell replication and migration (75). The significance of smooth muscle cell replication or apoptosis relative to plaque stability has not been clearly determined and at this point is conjectural. The accumulation of cholesterol in macrophages is reduced as a result of reduction in LDL-C. A direct effect of statins on the biologic behavior of lesions has been proposed. Importantly, statins directly inhibit metalloproteinase secretion by macrophages (94, 95). Metalloproteinases digest collagen and other supporting tissues in the plaque. Inhibition of this activity stabilizes the plaque and decreases susceptibility for rupture. It is unclear at this time whether statins differ in their direct effects on the vessel and, if so, which differences are important. It is likely that statins have direct antiatherosclerotic properties, but how much these properties contribute to the overall reduction in CHD events remains to be determined. The benefits of statins in the clinical trials generally do not become apparent until 1824 months, but accumulating data now suggest that benefits occur more rapidly. Failure to appreciate such early benefits may reflect the use of protocols that were designed to detect late rather than early reduction in clinical events and death.
Statins reduce atherothrombotic stroke and progression of carotid wall atherosclerosis
Despite the lack of a clear association between cholesterol levels and atherothrombotic stroke in observational studies and early lipid-lowering trials with nonstatin therapies, all of the major statin trials, particularly secondary prevention trials (96, 97), have shown reduced risk of stroke. Meta-analyses of statin trials demonstrated a 2430% decrease in the risk of stroke in secondary CHD prevention trials and a 1520% reduction (nonsignificant) in primary prevention trials (96). Post hoc analysis of the 4S revealed a 22% reduction in stroke alone (nonsignificant) and a 28% reduction in the combined endpoint of stroke and transient ischemic attack (98). In the CARE and LIPID trials, stroke was a prespecified endpoint. Pravastatin reduced the risk of stroke by 32% over a period of 5 yr in CARE (3.8 vs. 2.6%; P < 0.03) despite the use of aspirin by most patients (99). Pravastatin reduced the risk of atherothrombotic stroke by 23% (4.4 to 3.4%) in the LIPID trial (100). Thus, pravastatin consistently reduced the risk of nonhemorrhagic stroke in patients with previous MI or unstable angina. When the results of CARE and LIPID were combined (102,559 person-years of follow-up), there was a 22% reduction in total stroke (P = 0.01) and a 25% nonsignificant reduction in nonfatal stroke (101). This effect was consistent across all major subgroups and was attributed to a reduction in nonfatal atherothrombotic stroke. In addition, a predefined LIPID substudy assessed carotid atherosclerosis among 552 coronary patients using B-mode ultrasound of the common carotid artery (102). After 4 yr, carotid wall thickness increased in the placebo group and decreased in the pravastatin group independent of the baseline total cholesterol level. Surprisingly, in a study of acute coronary syndromes of only 16 wk, 80 mg atorvastatin significantly decreased fatal and nonfatal stroke events (81).
In the VA-HIT, gemfibrozil reduced stroke by 25%, transient ischemic attack by 59%, and carotid endarterectomy by 65% in patients with mean LDL-C and HDL-C levels of 104 and 32 mg/dl, respectively (48). The reduction of stroke with statins and gemfibrozil may be due to mechanisms other than lipid lowering. Pravastatin decreases lipid oxidation and inflammation, as well as increases the collagen content of carotid plaques (94). Possibilities include improved endothelial function, reduced inflammatory response, greater plaque stability, and less thrombus formation (103). It also is possible that the reduction in CHD attenuated the development of strokes.
Statin therapy reduces CHD events in all major subgroups of patients
Subgroup analyses consistently have shown that statins reduce coronary risk regardless of age, gender, or the presence of hypertension, cigarette smoking, diabetes, or low HDL-C. Taken collectively, the statin trials, as well as other lipid-lowering trials, and observational studies strongly support the concept that all patient subgroups benefit from reduction of cholesterol and more specifically LDL-C. Because the atherosclerotic process is similar in these subgroups, there is little reason to believe otherwise. In the Pravastatin Pooling Project, which included the large, randomized trials using 40 mg pravastatin, the relative CHD risk in patients above and below 65 yr, men and women, and with and without diabetes, hypertension, and cigarette smoking benefited from therapy. Other trials corroborate this efficacy of statins across broad subgroups (104).
Elderly.
Eighty-five percent of the deaths from CHD occur in those age 65 or older (105). Clinical CHD events are reduced in older patients with cholesterol lowering in a fashion similar to younger patients (Table 2). Age, the most powerful of the major risk factors, is a surrogate for atherosclerotic burden, but the relationship is imperfect. Older persons also are at higher absolute risk of CHD because of an increase in other risk factors that predispose to CHD, especially impaired glucose tolerance and/or systolic hypertension. Because absolute risk unequivocally increases with age, treatment of the asymptomatic older person should be thought of as secondary prevention. The lag period for benefit from treatment is well within the life expectancy of healthy older individuals. In the 4S, the relative risks for clinical events in patients older than 65 yr who were treated with simvastatin were 0.66 for deaths, 0.57 for CHD mortality, and 0.66 for major coronary events (106). Older patients in the CARE and LIPID trials benefited as much from statin therapy as younger patients. In the LIPID trial, total mortality was reduced in older patients treated with pravastatin (107).
|
Women.
There is increasing interest in CHD in women. CHD accounts for about 30% of the mortality in women, as contrasted with breast cancer, which accounts for only 3% (110). Furthermore, prevention of CHD events in women is particularly important because of higher morbidity and mortality in women than men following MI (111). Of the two major primary prevention trials, only AFCAPS/TexCAPS enrolled women. These women experienced a 46% reduction in the major endpoint compared with 37% for men. However, the difference was not significant due to the small number of major coronary events (7). In CARE, women in the treatment group had a 43% risk reduction in the primary outcome, a 46% reduction for combined coronary events, and a reduction of 48% for percutaneous transluminal coronary angioplasty and 40% for coronary artery bypass grafting (112). The magnitude of absolute risk reduction for women exceeded that for men, and the evidence of benefit tended to occur earlier. In the 4S, the 24% reduction in coronary events among women was similar to that seen in men, as was the risk reduction for the collective endpoints of clinical CHD events and need for revascularization (106). Women in the LIPID trial had a nonsignificant 11% reduction in CHD death or nonfatal MI, whereas men had a 26% reduction (6). In all of the major statin trials, women were as likely to benefit as were men. In the statin trials women have experienced up to a 46% reduction in risk of major coronary events (113). An angiographic trial also provided evidence of less progression and fewer new lesions in women treated with statins (114). With the continuing controversy surrounding postmenopausal hormone therapy, lipid therapy has become increasingly important for older women.
The traditional wisdom is that HRT is cardioprotective for postmenopausal women. Improvement in the surrogate markers of atherosclerosis, lipid/lipoprotein levels, and endothelial function in estrogen-deficient postmenopausal women with HRT is consistent with this concept. However, the belief that E replacement is beneficial is derived from observational studies in which the "healthy user" concept has been emphasized to explain the bias that might occur with regard to the cardioprotective effects of HRT. The HERS was the first randomized controlled trial of HRT in postmenopausal women with preexisting CHD (secondary prevention trial) (70). Women were treated with placebo or conjugated E plus medroxyprogesterone acetate for 5 yr. Despite the favorable HDL-C and LDL-C changes, cardiovascular events were 50% higher in women receiving HRT than in those receiving placebo during the first year of the study. Subsequently, clinical events decreased, and in the fourth and fifth years of the study were lower in the HRT group than the placebo group, but there was no overall evidence of cardioprotection. As expected, there was an increased risk of venous thromboembolism. It has been suggested that the early adverse effects of HRT were due to the prothrombotic properties of E, whereas lipoprotein changes and other antiatherosclerotic properties were responsible for the delayed benefit. A number of criticisms have been directed at this study, suggesting that the benefit of E was inhibited by the use of medroxyprogesterone acetate, that E alone or in combination with a different progestin might have produced entirely different results, or that the use of HRT on a continuous basis to prevent or minimize vaginal bleeding obscured the benefits that would have been seen had medroxyprogesterone been used cyclically. The lack of cardioprotection in the HERS has now been confirmed in a randomized controlled angiographic trial, the Estrogen Replacement Atherosclerosis study (115). This study was conducted in postmenopausal women with CHD who received placebo, E alone, or E plus medroxyprogesterone acetate. Angiograms were performed before and after a mean of 3.2 yr of treatment in 309 women. HRT did not prevent angiographic progression of CHD, nor did it prevent clinical CHD events in this study. Regardless of the criticism(s) of these studies, the facts remain that the combination of conjugated E and medroxyprogesterone acetate is the most common hormone replacement combination used in the United States and that these two studies, one an endpoint trial and the other an angiographic trial, are in agreement. Based on this information and the fact that statin therapy reduces risk in women with average or elevated LDL-cholesterol levels, HRT is not recommended for CHD prevention in postmenopausal women. Furthermore, because women do not always adhere or persist with HRT, it has been suggested that many postmenopausal women experience the acute CHD risks without the delayed benefits. The Womens Health Initiative, a primary prevention study of HRT (116), also reported in a letter to participants an increased risk for CHD during the first 2 yr of the study. Despite the evidence to the contrary, some have suggested that HRT is cardioprotective in postmenopausal women without CHD and could be used in such women, whereas others have suggested that HRT not be used for cardioprotection in any postmenopausal women until there is supportive evidence from properly designed trials (117). HRT is not an alternative to therapy of known efficacy for treatment of specific CHD risk factors in any woman.
Diabetes mellitus.
Patients with diabetes benefit as much or more than nondiabetics from lipid-lowering therapy with statins or fibrates (48, 118, 119). The issue of modifying CHD risk among patients with diabetes is important because they have increased early and late mortality after MI and worse outcome after revascularization (120). In the Diabetes Atherosclerosis Intervention Study (DAIS), fenofibrate reduced the angiographic progression of coronary lesions in type 2 diabetes (121). Fewer clinical endpoints occurred in the fenofibrate group, but the difference was not significant because of the small number of events. Three large statin clinical trials (4S, CARE, and LIPID) and two fibrate trials (HHS and VA-HIT) enrolled patients with diabetes (2, 4, 6, 48, 51). The relative and absolute reductions in risk of CHD events were comparable or better in patients with diabetes (Table 3). Studies enrolling large numbers of patients with diabetes are under way, but results will not be available until 20012005. It is obvious that lipids/lipoproteins are as important or more important in diabetics as in patients without diabetes. The CARE trial showed a significant 25% reduction in fatal and nonfatal MI and revascularization in patients with diabetes (119). Data from the 4S, including patients with fasting glucose levels of no less than 126 mg/dl without a clinical history of diabetes, confirm these findings (118). In this combined group of diabetics treated with simvastatin, the risk of major coronary events was reduced by 42%. Treated diabetics had survival curves that were virtually equivalent to those of patients without diabetes. Using absolute risk reduction rather than relative risk reduction, the NNT in the 4S was 12 for those with normal fasting glucose, 8 for impaired fasting glucose, and 7 for diabetes. In addition, another subgroup identified by the 1997 ADA diagnostic criteria as having impaired fasting glucose had 38% fewer major coronary events with simvastatin therapy. With this broad group of diagnostic criteria, generalization to the population at large appears reasonable. In the more recent LIPID trial, pravastatin treatment reduced CHD by 19% in patients with diabetes compared with 25% in those without diabetes (6). Although the LIPID study contained more patients with diabetes than either the 4S or CARE trials, the results were not statistically significant. In the AFCAPS, there were only 264 diabetics, and the results were not statistically significant although the magnitude of reduction in major coronary events was similar to that in the 4S (7). Haffner et al. (122) have demonstrated that the incidence of MI in diabetic patients without diagnosed CHD is similar to that for nondiabetic patients with a previous MI. As a consequence, the American Diabetes Association has recommended more aggressive management for all adults with diabetes, an LDL-C goal of no more than 100 mg/dl (123). This recommendation is further supported by the high rate of CHD events in patients with diabetes, especially those who already have sustained an MI. The NCEP considers diabetes to be a CHD risk equivalent (24).
|
Long-term lipid-lowering therapy with statins or fibrates is safe
Deaths.
The overall reduction in all-cause mortality with statins has mainly been from the decrease in CVD deaths. Earlier data on increased mortality from suicide, accidents, and trauma have been refuted (125). Deaths from noncardiovascular causes were basically unchanged by therapy with the exception of the CARE trial (4) and AFCAPS/ TexCAPS (7). There was a significant increase in the number of breast cancer deaths in the pravastatin group of CARE; but this did not occur in any of the other pravastatin trials, and it is likely that the difference between placebo and treated groups was due to an exceptionally low rate of breast cancer deaths in the placebo group. In the AFCAPS/TexCAPS, there was a significant (P = 0.04) decrease in melanoma in the group treated with lovastatin (7).
Morbidity.
Original concerns about the possibility of lens opacities with statins were never substantiated, and the requirements for ocular screening were discarded. Data from earlier nonstatin clinical trials suggested the possibility of a higher risk of death from trauma, suicide, and homicide (125, 126). Subsequent data from multiple studies indicate no significant association between low cholesterol reduction and psychological variables (125, 126, 127). Recently, data from the LIPID trial expanded these findings to demonstrate no significant difference by treatment group in measures of anxiety, depression, anger expression, or impulsiveness and no difference in the proportion of subjects with excessive alcohol consumption or adverse life events (127). The investigators concluded that long-term reduction of serum cholesterol with pravastatin has no adverse effect on psychological well being, confirming the findings of these earlier assessments by the randomized trial of statins. Curiously, it has now been reported that statins may prevent dementia (128). Some prospective observational studies (42, 129), but not all (130), have reported an increase of hemorrhagic stroke in persons with very low cholesterol levels, but these were not treated patients; consequently the cholesterol levels may have reflected underlying health problems.
Transient and reversible alanine aminotransferase (ALT) elevations can be seen with all hypolipidemic agents. The number of patients with any drug-related ALT elevation generally is greater in the groups treated with statins, but hepatic toxicity is rare. Sustained-release nicotinic acid appears to convey a higher risk for hepatotoxicity (131). Increased ALT usually is dose related and generally occurs within the first year of therapy. Prescribing information for the statins as a class indicates that the incidence of elevated serum transaminases is generally less than 2%, with the exception of the highest doses of atorvastatin, simvastatin, and fluvastatin, which are 2.3, 2.1, and 2.7%, respectively (132). The incidence of acute liver failure is estimated at 0.2 per 100,000 (131). Myopathy is an uncommon but serious side effect with the statins. When rhabdomyolysis is defined as an increase in creatine kinase to at least 10 times the upper limit of normal value, with compatible symptoms, the occurrence is less than 0.1% (133). The situations in which abnormal liver function tests and myalgia with elevated creatine kinase are clinically meaningful generally occur with near maximal statin dose, concomitant use of drugs affecting the cytochrome P450 system (134), and/or the combination of lipid-lowering drugs in the elderly in whom the serum creatinine underestimates impaired renal function.
Fibrates also appear to be well tolerated, with few serious adverse effects with gemfibrozil, fenofibrate, or bezafibrate, although much fewer data are available. However, concerns remain because of the adverse mortality experience with clofibrate (135). Fibrates can cause a reversible deterioration in renal function in patients with renal insufficiency or with a renal transplant. Fibrates can be used safely in combination with statins, but this should be done with great care and only in patients with mixed dyslipidemia who are at high CHD risk and have persisting lipid/lipoprotein abnormalities after therapeutic lifestyle changes and monotherapy. The combination of cerivastatin specifically with gemfibrozil is contraindicated.
Diet composition influences CHD risk independent of changes in blood lipid values
Nutritional approaches to preventing CHD have received less attention because of difficulty in implementing and maintaining low-fat diets over prolonged periods and because of minimal changes in LDL-C. Yet dietary intervention studies show a remarkable decrease in coronary events without necessarily a change in LDL-C or major risk factors. Regardless of its effect on LDL-C, dietary therapy should be considered for primary prevention as well as a necessary adjunct to drug therapy for secondary prevention. Recent dietary intervention trials (Table 4) for secondary prevention of CHD show striking reductions in the recurrence rate of coronary events independent of changes in major risk factors including cholesterol (136). Men with heart disease who were randomized to eat fatty fish or take fish oil supplements as opposed to no dietary advice were less likely to have cardiovascular events (137). After 2 yr, there was a reduction in all-cause and CVD mortality, although neither was statistically significant. The Lyon Diet Heart Study, a randomized, controlled secondary prevention trial, evaluated a Mediterranean-style Step I diet over a 46-month period (138). The dietary intervention group experienced a 70% reduction in CVD morbidity and mortality compared with a control group of patients who received dietary advice similar to concurrent NCEP guidelines. Risk reduction was observed despite no differences in cholesterol values between the two treatment groups. The Mediterranean-style diet emphasized olive oil and a canola oil margarine as the fat sources replacing butter and cream. Overall the diet was high in
- linolenic acid and contained more bread, root and green vegetables, fruit, fish, and poultry (139). The Gruppo Italiano per lo Studio della Sopravvivenza nellInfarto Miocardio (GISSI) Prevenzione trial reported that coronary patients receiving N-3 polyunsaturated fatty acids had 1015% (P < 0.05) fewer primary endpoints of death, nonfatal MI, and stroke and 30% fewer CVD deaths (140). As in the Lyon Diet Heart Study, the observed benefits were independent of lipids and other major risk factors. This reduction in cardiovascular events could not be attributed to any other risk factors. Collectively, these dietary data indicate that diet composition independent of change in total or LDL-C reduces CHD. Currently, there is little clinical trial evidence of the benefit of antioxidant vitamin supplementation (59, 141), but perhaps different types of trials are needed (142).
|
Our understanding of the basic biology of atherosclerosis and the impact of therapy on clinical CHD has expanded at an accelerated pace in the past decade. We are on the verge of understanding the basic mechanisms of the development of atherosclerosis, leading to development of new therapeutic options. Perhaps we will see the fulfillment of the prophecies of Brown and Goldstein (143) that CHD as a major public health problem may be eliminated early in the 21st century.
There is now no real disagreement that lipid modification with statins or fibrates leads to reduced coronary morbidity and mortality, as well as the need for revascularization procedures. The value of statins for the acute coronary syndrome currently is under investigation. Yet, the use of lipid-lowering medication for primary CHD prevention, despite the trial evidence, has not been translated satisfactorily into general medical practice. Optimum assessment of risk and identification of high-risk patients without CHD should allow appropriate primary intervention. The clinical trials have unequivocally shown that statin therapy can reduce cardiovascular events with few adverse effects. The efficacy of statins in older persons, diabetics, and women seems to be well established, although specific clinical trials must corroborate these subgroup analyses. An unanticipated reduction in stroke, especially in the secondary prevention studies, has emerged as an added benefit for older persons. HRT has proven not to be an alternative to treatment of specific CHD risk factors in postmenopausal women with clinical CHD. Cholesterol reduction does not cause psychological problems or excessive noncardiac death and actually may improve quality of life. The efficacy and safety of these agents for primary and secondary prevention has been demonstrated across a broad range of populations with average to high cholesterol levels. How much of the reduction in CHD is due to LDL-C lowering or accompanying lipid changes, and how much is due to the pleiotropic effects of these drugs, is unresolved. A therapeutic strategy that focuses on increasing HDL-C while reducing LDL-C raises the possibility of even greater reduction in CHD events than the 2535% reduction obtained with statins. This can be accomplished by using drug combinations. The role of TRLs and triglyceride in atherosclerosis is plausible but lacks data from clinical trials. Pleiotrophic effects of statins and fibrates most likely complement the documented lipid changes through unexplained mechanisms. Management of other nonlipid risk factors in conjunction with favorable lipoprotein changes almost certainly will further reduce coronary events. In addition, we need to clarify the pleiotrophic effects of statins and fibrates in well designed studies to determine whether meaningful differences exist that potentially influence clinical decisions. Trials are under way with new agents to modify lipoproteins, which should lead to further improvement in our understanding and management of the atherosclerotic process.
We still need properly designed trials to address the efficacy of the following strategies on CHD: 1) combinations of statins plus fibrates or nicotinic acid in patients with combined hyperlipidemia as well as other lipoprotein abnormalities; 2) novel therapies to increase HDL-C in patients with low HDL-C; 3) triglyceride-lowering in patients with and without increased LDL-C; 4) therapy tailored to modify the size of the LDL particle; and 5) lipid-modifying therapy in adolescents and young adults to prevent the development of early atherosclerotic lesions. This will require new sensitive noninvasive imaging techniques.
Footnotes
Abbreviations: 4S, Scandinavian Simvastatin Survival Study; AFCAPS/ TexCAPS, Air Force/Texas Coronary Atherosclerosis Prevention Study; ALT, alanine aminotransferase; ATP, Adult Treatment Panel; CARE, Cholesterol and Recurrent Events; CHD, coronary heart disease; CLAS, Cholesterol Lowering Atherosclerosis Study; CVD, cardiovascular disease; DAIS, Diabetes Atherosclerosis Intervention Study; FATS, Familial Atherosclerosis Treatment Study; HDL-C, high-density lipoprotein cholesterol; HERS, Heart and Estrogen/Progestin Replacement Study; HHS, Helsinki Heart Study; hsCRP, highly sensitive C- reactive protein; LCAS, Lipoprotein and Coronary Atherosclerosis Study; LDL-C, low-density lipoprotein cholesterol; LIPID, Long-Term Intervention with Pravastatin in Ischemic Disease; Lp(a), lipoprotein (a); MI, myocardial infarction; NCEP, National Cholesterol Education Program; NNT, number needed to treat; POSCH, Program on the Surgical Control of the Hyperlipidemias; TRL, triglyceride-rich lipoprotein; VA-HIT, Veterans Affairs Cooperative Studies Program High Density Lipoprotein Cholesterol Intervention Trial; VLDL, very low-density lipoprotein; WOSCOPS, West of Scotland Coronary Prevention Study.
Received May 16, 2001.
Accepted July 18, 2001.
References