TABLE OF CONTENTS
Randomized Controlled Trial Summaries
Return to the Clinical Case and Question
Prevention of Myocardial Infarction And Coronary Death
With HMG-CoA Reductase Inhibitors
Internal Medicine Resident Grand Rounds
By Mark Napoli, MD.
January 18, 1999
Earlier in this century, the reduction of low-density lipoprotein (LDL) cholesterol has been shown effective in slowing the progression and improving the clinical outcomes of coronary artery disease. In the 1950s and 1960s, many cholesterol-lowering agents were introduced into clinical use. Many of these agents were modestly successful in cholesterol reduction but had several disagreeable side effects such as gastrointestinal upset, angina, and even feminizing effects.
In 1971, a biochemist named Akiro Endo began searching for a microbial by-product that would inhibit HMG-CoA reductase under the assumption that certain microorganisms would use such a compound as a weapon against other microorganisms that required sterols or other isoprenoids for growth. By the end of 1973, Endo and his colleagues isolated 23 milligrams of an HMG-CoA reductase inhibitor they named mevastatin from 600 liters of culture filtrate from the mold Penicillium citrinum. Mevastatin was proven safe and effective in lowering cholesterol in the early 1980s. By 1990, several drugs referred to collectively as statins such as lovastatin, pravastatin, and simvastatin were derived from mevastatin and marketed successfully in many countries.
HMG-CoA reductase inhibitors, or statins, block the production of LDL cholesterol by inhibiting the enzyme HMG-CoA reductase as the name suggests. Statins are also known to affect in vivo levels of other lipid parameters; but this as well as the specific biochemical pathways through which statins operate is beyond the scope of this discussion.
Clinical trials released before data on statins established that cholesterol reduction led to improved clinical outcomes. A leading theory was that reducing cholesterol leads to a reduction in the presence or size of coronary cholesterol plaques thereby decreasing in coronary events. However what these studies found using careful angiographic analysis was that there was a minimal regression at best and usually only a slowed progression of the size of cholesterol plaques within coronary arteries. These disappointing findings were incongruous with evidence that aggressive reduction in cholesterol reduces the risk of MI and death from coronary artery disease. With the appearance of statins researchers used this class of drug in similar studies. One such example was the Pravastatin Limitation of Atherosclerosis in Coronary Arteries trial (PLAC 1).
This trial published by Pitt et al used pravastatin as the sole intervention. Investigators powered the study to detect regression in coronary plaque size. What they found, however, was only a slowed progression in size. The secondary endpoint, coronary events, was dramatically out of proportion to the primary endpoint. Pravastatin decreased myocardial infarction and coronary events significantly and more so than other non-statin trials with similar findings. These results cannot be explained by a slowed progression of coronary plaques and they led researchers down a different path to explain the role of statins and cholesterol reduction in decreasing coronary events.
Healthy endothelium releases paracrine factors such as nitric oxide, which cause vasodilatation in response to stress. In vitro studies have shown that LDL and oxidized LDL directly injure coronary endothelium leading to decreased production of paracrine factors. Elevated LDL cholesterol has also been shown to stimulate the endothelium to generate superoxide radicals that directly inactivate nitric oxide.
Clinical trials investigating the effects of LDL reduction on endothelium used the established finding that acetylcholine dilates normal healthy coronary arteries and it vasoconstricts coronary arteries that are diseased regardless of the size of atheromatous plaques. Several studies in the late 1990s demonstrated that LDL reduction by a variety of methods repeatedly results in improved vasomotion of coronary arteries in response to acetylcholine challenge. This observation has implications for explaining LDL reduction lowering the incidence of adverse coronary events by stabilizing endothelium from vasoconstriction and possible rupture at areas of compromised integrity like sites of atheroma.
In 1993, The Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults released the second report of the National Cholesterol Education Program (NCEP). Research up to this point had established that reduction of LDL cholesterol in patients with known coronary artery disease (CAD) results in decreased coronary events and CAD mortality in primary and secondary prevention. Also by this time statins had yet to be proven safe or effective in preventing coronary events or mortality although they were generally well accepted as highly effective for reducing LDL.The NCEP report reviews accepted CAD risk factors in addition to high LDL cholesterol as age greater than 45 for men or 55 for women, any post-menopausal women, family history of premature CAD, current cigarette smoking, hypertension, low HDL cholesterol, and diabetes mellitus. The only negative risk factor noted was high HDL cholesterol.
The panel denotes three general risk categories for patients with high LDL cholesterol to be taken into consideration for therapy: (1) highest risk patients for future CAD events because of previous CAD events or other present atherosclerotic disease such as PVD or symptomatic carotid disease; (2) intermediate risk patients without evident CAD who are at high risk because of multiple CAD risk factors together with high LDL; (3) low risk patients with high cholesterol but no other risk factors for CAD.
Furthermore the panel categorized levels of LDL into risk categories for those without known CAD: (1) high risk LDL is greater or equal to 160 mg/dL; (2) borderline high risk LDL is 130-159; (3) desirable LDL is less than 130. Those patients with known CAD should have desirable LDL cholesterol below 100 to significantly prevent further CAD events.
The NCEP report makes several concessions after setting forth these aggressive guidelines for prevention. The panel advocates reserving drug therapy for otherwise healthy young men and premenopausal women who have very high LDL levels (>220 mg/dl.). They admit that for lack of current evidence at the time there are no clear guidelines for the treatment of post-menopausal women. And lastly the panel acknowledges the high incidence of CAD events in the elderly but surprisingly cautions against aggressive drug therapy in the group as a whole siting several reasons. Those listed are not only advanced age and comorbidity, but also intolerance to medication side effects.
As stated above these guidelines were set forth before any randomized controlled trials of HMG-CoA reductase inhibitors had been performed exploring the safety and efficacy in preventing disease and mortality as single agents.
Tessie is a 70-year-old widowed white woman of Mediterranean descent who has no significant past medical history and is in general good health. She takes vitamins and exercises. She presents to the emergency room with a dull ache associated with shortness of breath, which was present for months only with exertion but now occurs at rest. She has no significant EKG changes and becomes pain free with nitroglycerine and heparin in the emergency room. Tessie is admitted to the cardiology service where she remains pain free however her troponin peaks above the normal range at 4.0 ng/mL. The diligent physicians maximize her medical therapy, and the patient returns home wishing to continue this conservative approach. After a follow-up appointment in which a fasting lipid profile is drawn, Tessie undergoes a non-invasive stress imaging study which reveals normal left ventricular ejection fraction and contractility and no inducible ischemia. The lipid profile returns and reveals total cholesterol of 170, a LDL of 115, and a HDL of 34 mg/dL.
Randomized Controlled Trial Study Summaries
The trials below are arranged in chronological order in which they appeared in published form. All reported data below was significant, having a p value of always below and usually much less than 0.05 unless otherwise specified. All patients in the trials underwent dietary therapy prior to measuring pre-trial cholesterol levels. During the study periods, reductions in LDL cholesterol ranged from 25% to 38% in patients in the experimental groups. None of the placebo groups had significant reductions in LDL cholesterol. Very few patients withdrew from any trial because of adverse drug reactions, and no study demonstrated significant increases non-cardiovascular mortality in the experimental groups. Only endpoints pertinent to this discussion are reported or discussed below; therefore these summaries are in no way complete reports of the trials, but rather focused details of data within the scope of this discussion.
The Scandinavian Simvastatin Survival Study (4S), November 1994.
This randomized, double blind, multicenter, placebo-controlled trial lasted 5.4 years and compares men and women aged 35-70 with prior MI (79%) or stable angina taking either simvastatin or placebo. The primary endpoint of the study is total mortality. Secondary endpoints are major coronary events including CAD death and MI. The trial includes patients with a history of angina or prior MI with a total cholesterol of 5.5-8.0 mmol/L (mean=261mg/dL) and total triglycerides of <=7.5mmol/L. Exclusion criteria are many. The 4S trial excludes patients who were pregnant; had unstable angina (USA), tendon xanthomata, planned CABG or PTCI, myocardial infarct more recently than six months prior, poor mental functioning, or competing cardiac or other significant disease process.
Over 7000 patients were recruited and 4444 met criteria for participation. They are randomized with very similar baseline characteristics and the dose of simvastatin or placebo is titrated to 40 mg as tolerated. Follow-up in the study was excellent and was intention-to-treat.
Briefly the results of the 4S trial are as follows: 11.5% in the control group died versus 8.2% in the simvastatin group yielding a relative risk reduction (RRR) of 28.8%. Approximately 22.6 percent of patients in the control group suffered a non-fatal MI; and sixteen percent of the simvastatin group had a non-fatal MI giving a RRR of 29.6%. All-cause coronary artery disease death in the control group was 8.5% while CAD death in those patients taking simvastatin was 5%, which gave a RRR of 41.2%%.
This very large comprehensive trial yields very promising results for those patients with coronary disease significant enough to cause symptoms or result in prior infarction. The drug was very well tolerated in the experimental group with very few withdrawals from side effects. Threats to the validity of the trial are that the Merck Corporation funded the study and several Merck employees participated directly in the analysis of data. Only 19% of the participants were women and less than half of the patients were taking aspirin. About 60% of patients took beta-blockers.
Prevention of Coronary Heart Disease with Pravastatin in Men with Hypercholesterolemia (The West of Scotland Study), November 1995.
The next randomized, double blind, placebo-controlled trial using statins to appear was the West of Scotland Coronary Prevention Study (WOSCOPS). The trial lasted five years and compares pravastatin 40 mg. with placebo in a population of men with elevated LDL cholesterol and no prior myocardial infarction. The primary endpoint is non-fatal MI and CAD death. Men (mean age 55yo) included in the study had LDL level greater than 155 mg/dL and no evidence of prior MI although about five percent of patients had stable angina at the time of recruitment. Exclusion criteria included Q waves or other serious EKG abnormalities or arrhythmias.
About 160,000 men in primary care clinics were recruited and 6595 were entered in this intention-to-treat analysis. Patients were randomized to pravastatin or placebo and the two groups had very similar demographic data.
Over the five-year study 7.5% of patients taking placebo suffered a non-fatal MI as compared to 5.5% taking pravastatin demonstrating a RRR of 26.7%. Coronary artery disease death occurred in 1.9% of patients in the control group while 1.2% of the experimental group suffered CAD death for a RRR of 36.8%. The study also reported a significant risk reduction from total cardiovascular mortality.
Although the results of this trial are not as striking as with those of secondary prevention in the 4S trial, they did show a significant advantage in using statins in primary prevention of MI. The prevention of CAD death is somewhat questionable according to this study. In fact, more smokers died from coronary events during the study than non-smokers did --pravastatin or not. Also the study included only men and no information about their other medication such as aspirin was available.
WOSCOPS Kaplan-Meier Analysis of Time to Coronary Event
The Effect of Pravastatin on Coronary Events after Myocardial Infarction in Patients with Average Cholesterol Levels (The CARE trial), October 1996.
The CARE trial was the first large RCT to investigate patients with average or normal cholesterol in secondary prevention of MI. The multicenter study had a five-year average follow-up. Men and women aged 21 to 75 included in the study were randomized in a double-blind fashion to either 40mg of pravastatin or placebo. Inclusion criteria were total cholesterol of less than 240 mg/dL, LDL levels between 115 and 174 mg/dL, and a myocardial infarction 3-20 months prior to randomization. Exclusion criteria were fasting triglycerides greater than 350, fasting glucose greater than 220, an EF less than 25% or symptomatic CHF. Primary endpoints were fatal coronary events and non-fatal MI.
Randomization included 4159 patients to experimental or control in an intention-to-treat analysis. Follow-up was near 100% and 94% of the pravastatin group took the drug for the entire study period, slightly more than the placebo group. Baseline characteristics were very similar in both groups. Most patients were taking aspirin but less than half were taking beta-blockers even though all had had prior MI.
The results of the study showed 8.3% of placebo patients suffered non-fatal MI while 6.5% of pravastatin patients did with a RRR of 21.7%. About 5.7% of placebo patients died from coronary causes while 4.6% died while taking pravastatin yielding a RRR of 19%.
This third large randomized controlled trial of statins proves a significant although less-than-striking advantage in favor of the use of statins in the prevention of myocardial infarction and CAD death. The trial revealed no significant benefit on secondary prevention in patients whose pre-trial LDL cholesterol levels were less than 125 mg/dL. This contrasts with the guidelines set forth in the NCEP. Moreover, investigators concluded that the benefit of statin use increased along with pre-trial levels of LDL.
CARE Trial Kaplan-Meier Curve
Primary Prevention of Acute Coronary Events with Lovastatin in Men and Women with Average Cholesterol Levels (AFCAPS/TexCAPS), May 1998.
This randomized, double blind, placebo-controlled, trial of lovastatin also lasted five years. The study population was men aged 55 to 73, and post-menopausal women aged 45 to 73, who had no known or clinically evident CAD and normal cholesterol levels. The primary endpoint was the first acute major coronary event and included myocardial infarction, unstable angina, or sudden death. Inclusion criteria consisted of total cholesterol levels 180 - 264 mg/dL, LDL cholesterol 130 - 190 mg/dL, HDL cholesterol equal or less than 45 mg/dL, and triglycerides less than 400 mg/dL. Exclusion criteria were prior MI or CAD, uncontrolled hypertension, secondary increased hyperlipidemia, uncontrolled diabetes mellitus, or morbid obesity.
Researchers screened 102,800 potential participants and randomized 6605 men and women to either lovastatin or placebo at two sites in Texas in an intention-to-treat analysis. Most patients completed the study adhering to the drug regimen, however 2138 participants withdrew from the trial before completion. Baseline characteristics in each group were similar.
The placebo group experienced more events included in the primary endpoint than did the lovastatin group, 5.5% versus 3.5% respectively (RRR=36.4%). In the placebo group 2.9% suffered fatal or non-fatal MI while 1.7% of those taking lovastatin suffered the same fate (RRR=41%). This study was not able to show a significant decrease in coronary artery disease related mortality.
The AFCAPS/TexCAPS study demonstrated a small but significant decrease in the combined occurrence of MI, USA, or sudden cardiac death in patients who were relatively healthy. Researchers pointed out that only 17% of the patients studied in the trial qualified for aggressive drug treatment as set forth by the NCEP guidelines. Authors of the trial lauded the inclusion of a large number of women in the study, however the event rate in women participants was small (20/997). This study was also funded solely by the Merck Corporation.
Prevention of Cardiovascular Events and Death with Pravastatin in Patients with Coronary Heart Disease and a Broad Range of Initial Cholesterol Levels --The Long- Term Intervention with Pravastatin in Ischaemic Disease Study Group (LIPID trial), November 1998.
The latest published randomized, double blind, multicenter, placebo-controlled trial on statins. The primary outcomes were CAD death and myocardial infarction. The secondary outcome was total mortality. The trial length was 6.1 years. The study included a patient population of men and women with known CAD in the form of prior MI or unstable angina (USA). The listed inclusion criteria were men and women aged 31 to 75; prior MI or hospitalized USA three to 36 months prior to randomization; and total cholesterol of 155 to 271 mg/dL. Mean LDL was 150 mg/dL. The study excluded those with any significant medical or surgical event less than three months prior to randomization; history of congestive heart failure, renal, or hepatic disease; or those taking other cholesterol lowering medicines.
The study randomized 9014 patients with similar background characteristics to take pravastatin 40 mg. or placebo. The investigators followed all patients but one to completion of the trial in an intention-to-treat basis. Non-compliance rates were similar in both groups.
The LIPID trial reported significant reductions in CAD-related and total mortality, as well as non-fatal and fatal myocardial infarction among other endpoints. About 14.1% of placebo patients died in the study while 11.0% of pravastatin patients died (RRR=22%). Of those suffering MI, 10.3% were in the placebo group and 7.4% were in the pravastatin group (RRR=28.2%). Death attributed to coronary disease numbered 8.3% of those taking placebo versus 6.4% of those taking pravastatin (RRR=22.9%). Furthermore, investigators compared risk reduction within several subgroups. They reported more benefit in higher pretrial cholesterol levels and LDL levels.
The LIPID trial showed definite benefit for the use of statins in secondary prevention of death and MI. The trial was not able to demonstrate a definite benefit for women over men, likely to the small number of events in the female population. Again this known-CAD study population was on less than optimal therapy with less than half taking beta-blockers although 82% received regular aspirin although standard treatment was much better than in 4S. The study was reportedly performed independently of Bristol Myers Squibb, the makers of pravastatin; however they paid for the study.
LIPID Study Kaplan-Meier Curve of CAD Mortality
When following these trials as they appeared historically it may be difficult to gain insight into how to apply the results into real-world practice. A more pragmatic approach may be to view the studies based upon the patient population in which they were directed. Overall the studies have used patients from risk categories similar to those in the NCEP: 1) patients with high levels of LDL cholesterol and prior MI or significant CAD; 2) patients with average LDL and prior MI or significant CAD; 3) patients with high levels of LDL with no prior evident CAD; and finally 4) patients with normal LDL and no prior evident CAD.
Aside from the patient population, these major trials are strikingly similar in methods, analysis, length, and endpoints. Also, the trials all report similar relative risk reductions of MI and CAD death between 20 and 40% regardless of the dissimilarity of the baseline incidence and prevalence of disease of each population. Thus comparison of benefit among cohorts is difficult. When the results can be converted into the number- needed-to-treat, the disparity between risk categories becomes more evident.
To prevent myocardial infarction in the high LDL and prior myocardial infarction, the 4S trial and the LIPID study group show the number-needed-to-treat is 12 and 34 respectively. The NNT in 4S for non-fatal MI is 15. The CARE trial focuses on a group with prior MI but normal cholesterol, and the NNT for non-fatal MI is 55. To prevent non-fatal MI in those with high LDL and no prior MI or CAD, the West of Scotland trial yields a NNT of 50. The AFCAPS trial shows a NNT of 83 to prevent one MI --fatal or not-- in a population of people who have normal cholesterol and no evident CAD. The escalation of the number-needed-to-treat is clear as the incidence of CAD and risk factors decline (table 1).
Table 1. Prevention of Myocardial Infarction
|
Cohort (RCT) |
Relative Risk Reduction |
Number Needed to Treat |
|
LDL, prior MI(4S,LIPID) |
31% and 28%§ |
12 and 34§ |
|
Avg LDL, prior MI(CARE) |
21% |
55 |
|
LDL, no MI(WOSCOPS) |
26% |
50 |
|
Avg LDL, no MI(AFCAPS) |
41%§ |
83§ |
§ represents both fatal and non-fatal MI
To prevent death from coronary disease in patients with high LDL and prior MI, the 4S and LIPID trials show the NNT as 29 and 53 respectively. The CARE trial, that includes patients with normal LDL and prior MI, shows a NNT of 91. The West of Scotland trial focuses on patients with high LDL cholesterol but without prior MI and mostly no evident CAD. The number needed to treat this cohort to prevent a coronary related death is 143. Finally the AFCAPS study cannot demonstrate a significant NNT to prevent a coronary disease related death in the disease free population with normal cholesterol. Fatal coronary events are within the combined primary endpoint including MI, unstable angina, and sudden cardiac death. Again, when the evidence these trials present is shown in a different light it becomes clear that the benefit of statins in primary and secondary prevention largely depends on the risk category of the patient (table 2).
Table 2. Prevention of Coronary Artery Disease Death
|
Cohort (RCT) |
Relative Risk Reduction |
Number Needed to Treat |
|
LDL, prior MI(4S,LIPID) |
41% and 23% |
29 and 53 |
|
Avg LDL, prior MI(CARE) |
19% |
91 |
|
LDL, no MI(WOSCOPS) |
36% |
143 |
|
Avg LDL, no MI(AFCAPS) |
ns© |
ns© |
© not significant
Overall the randomized controlled trials using statins published thus far have mainly included middle aged men with very little data presented in important subgroups that physicians see in clinical practice. Small subgroup analyses have been published concerning the elderly, and diabetics that show a trend towards increased benefit in the elderly, and at least equal benefit in diabetics; however data on post-menopausal women is largely lacking. Evidence for the benefit of statins in patients with competing illnesses like congestive heart failure, too, is notably absent. This fact may be because meta-analyses reviewing general cholesterol lowering trials earlier this decade found that major competing illnesses like CHF erased all benefit of lowering cholesterol.
HMG-CoA reductase inhibitors are by far the most efficacious medicine for lowering total and LDL cholesterol a significant degree. Evidence shows that dietary therapy alone and non-statin therapy reduces cholesterol 9.2% and 14.6% respectively; while statin therapy reduces cholesterol about 24% on average.
Statins appear very safe for use in the general population, with few side effects or contraindications. So far in evidence available, the benefit of use of statins seems to far outweigh the risk. Given the fact that coronary atherosclerosis is a lifelong disease process, the length of these major trials may greatly underestimate either benefit or risk.
Statins in the secondary prevention of myocardial infarction and coronary artery disease related death is highly efficacious. The 4S, LIPID, and CARE trials have supported this consensus even despite study participants on less than optimal therapy like aspirin and beta-blockers. Statin trials like West of Scotland for the primary prevention of MI and CAD death of patients with risk factors show modest benefit though not as high as in patients with prior disease. Statins used for primary prevention of MI and CAD death in the AFCAPS/TexCAPS population is questionable and as of yet no guidelines exist. Overall the guidelines set forth in the NCEP from 1993 appear to remain applicable with little exception. When applied to the elderly, post-menopausal women, and diabetics in each risk category; the above conclusions appear to hold true within limited study.
Based on available evidence including patients with competing illnesses present such as congestive heart failure, the use of statins does not appear to alter prognosis. This fact, however, should be carefully considered in light of emphasis on correct treatment of CHF as well as recent advances in therapy.
Finally, when the findings of risk reduction in statin trials are applied to a given patient, it is important to consider the nature of the cohort in which he dwells. Several factors including incidence and prevalence of disease within the patient’s risk category should be weighed by doctor and patient as individuals against the prospect of lifelong treatment.
Return to the Clinical Case and Question
I believe that the decision of whether or not to begin Tessie on an HMG-CoA reductase inhibitor can be based on the guidelines and evidence reviewed above. But, of course, no hard and fast rules apply and the physician must rely on the entire biopsychosocial model. The LDL level of 115 mg/dL is between the NCEP guidelines for initiation of therapy but below the level found by the CARE trial showing benefit. The clinical case is that of an elderly post-menopausal white female with no competing illnesses and with evident but now stable coronary artery disease. She was released from the hospital on standard-of-care medical therapy for post-MI patients. She presently has adequate financial resources. The patient lives an active and full life with a very involved and close family. You be the judge.
Bibliography