Coronary Artery Disease: TheScandinavian Simvastatin Survival

Study ExperienceTerje R. Pedersen, MD

Although hyperlipidemia is a known risk factor for cor-onary artery disease, lipid-lowering agents were notused widely until recently because evidence was lackingthat they could prolong life. In 1987, a large clinicaltrial, the Scandinavian Simvastatin Survival Study (4S),was designed to test whether such therapy could de-crease all-cause mortality in patients with documentedcoronary artery disease. The prospective, randomized,multicenter trial included 4,444 patients who had hadangina pectoris or myocardial infarction (MI), serumtotal cholesterol of 213–310 mg/dL, and serum trigly-cerides <221 mg/dL. Patients received either simvasta-tin 20–40 mg/day or placebo and were followed for amedian of 5.4 years. Therapy decreased total choles-terol an average of 25%; low-density lipoprotein (LDL)cholesterol, 35%; and triglyceride levels, 10%. Therapyincreased high-density lipoprotein (HDL) cholesterol lev-els 8%. Although noncardiac death rates were similar

among the groups, the relative risk of mortality (fromany cause) was decreased 30%, and the relative risk ofcoronary mortality was decreased 42% in the simvasta-tin arm. The mortality risk reductions were profound inpatients >60 years of age. Treatment also significantlydecreased the relative risk of coronary events and theneed for bypass surgery or coronary angioplasty. Pa-tients with diabetes also benefited significantly fromsimvastatin therapy. The reductions in relative risk ofmajor coronary events were achieved irrespective ofsuch baseline risk factors as hypertension and smokingand such medication factors as aspirin, b-blocker, andcalcium-antagonist use. Simvastatin therapy has beenshown to be cost-effective, decreasing per-patient hos-pitalization costs by 31% or $3,872 in 1995 dollars.Q1998 by Excerpta Medica, Inc.

Am J Cardiol 1998;82:53T–56T

Coronary artery disease remains the leading causeof death in Scandinavian nations, just as it does in

the United States. Moreover, coronary artery diseasemortality is similar in Scandinavia and the UnitedStates.1 In some parts of Scandinavia, such as northernFinland, coronary artery disease mortality still ranksamong the highest in the world, although the mortalityis decreasing.

Hyperlipidemia is a well-recognized risk factor forcoronary artery disease, and multiple clinical trialshave shown that cholesterol-lowering interventionscan decrease coronary artery disease events. Clinicalguidelines in the United States and Europe have es-tablished a role for using interventions, especiallylipid-lowering drug therapy.2–5 Lipid-lowering agentshave not been widely used until recently, however,because of a lack of evidence that the agents couldprolong life. Moreover, some evidence suggested thatany cardiac mortality benefits of lipid-lowering ther-apy might be offset by increases in noncardiac mor-tality.6–8 Scandinavian physicians have been particu-larly skeptical about the overall benefits of therapy todecrease high cholesterol levels.

In 1987, a large clinical trial was designed to testthe hypothesis that aggressive lipid-lowering therapywith simvastatin would decrease all-cause mortality inpatients with documented coronary artery disease. TheScandinavian Simvastatin Survival Study (4S)9 pro-vided the evidence that many physicians demanded as

justification for regular use of lipid-lowering therapyin patients with coronary artery disease.

TRIAL BACKGROUNDInvestigators at 94 clinical centers in Scandinavia

prospectively randomized to treatment or placebo4,444 patients who had a history of angina pectoris ormyocardial infarction (MI). Inclusion criteria includeda serum total cholesterol level of 213–310 mg/dL andserum triglyceride level of#221 mg/dL. Investigatorsexcluded patients who had a recent (,6 months) his-tory of MI, stroke, antiarrhythmic therapy, congestiveheart failure requiring therapy, or who planned toundergo coronary bypass surgery or coronary inter-ventional procedures.

Patients received either simvastatin (20–40 mgdaily) or placebo and were followed for a median of5.4 years. All patients received dietary counselingconsistent with the guidelines of the European Ath-erosclerosis Society. Patients were evaluated every 6weeks for the first 18 months and every 6 monthsthereafter.

The primary endpoint of the trial was total mortal-ity. The secondary endpoint was any major coronaryevent, defined by the time to a first event. Majorevents comprised coronary death, definite or probablenonfatal MI, resuscitation from cardiac arrest, andoccurrence of silent MI, as documented by electrocar-diogram.

In the study population,.80% were men, and halfthe patients were$60 years of age. About 80% had ahistory of MI, and hypertension and current smokinghistory were documented in a quarter of the patients.

From the Department of Cardiology, Aker Hospital, Oslo, Norway.Address for reprints: Terje R. Pedersen, MD, Department of Car-

diology, Aker Hospital, N0514 Oslo, Norway.

53T©1998 by Excerpta Medica, Inc. 0002-9149/98/$19.00All rights reserved. PII 0002-9149(98)00727-9

About a third of the patients were taking aspirin atstudy entry, and 57% usedb blockers.

At baseline, serum total cholesterol averaged 261mg/dL. The mean low-density lipoprotein (LDL) cho-lesterol concentration was 188 mg/dL. High-densitylipoprotein (HDL) cholesterol levels averaged 46 mg/dL, and the serum triglyceride mean was 134 mg/dL.

Patients in the placebo and active treatment groupswere well matched, and there were no major clinicalor demographic differences between the 2 groups.

PRIMARY RESULTSOver the whole course of the study, treatment with

simvastatin decreased total cholesterol levels by anaverage of 25%. LDL-cholesterol levels decreased anaverage of 35% and triglyceride levels 10%. Themean serum HDL-cholesterol level increased 8%. Ta-ble I shows the changes after 1 year of simvastatintherapy. In contrast, lipid levels did not change appre-ciably in the placebo group, except for a trend towardhigher triglyceride levels over the course of the study.

Simvastatin therapy resulted in significant reduc-tions in relative risks of mortality and disease (TableII). A total of 256 (11.5%) patients in the placebocohort died during the study, compared with 182(8.2%) in the simvastatin group. The difference trans-lated into a significant 30% reduction in total mortalityrelative risk in the simvastatin treatment arm(p 5 0.0003). Coronary mortality declined to an even

greater degree in the group that received active treat-ment. The placebo group had 189 coronary deaths, or74% of all deaths in the group. In the simvastatingroup, 111 coronary deaths occurred. The differencereflected a highly significant 42% reduction(p 5 0.00001) in the relative risk of coronary heartdeath in the simvastatin group.

The 4S trial provided some resolution to the con-troversy surrounding an apparent increased incidenceof noncardiac death in previous trials of lipid-loweringtherapy. Similar numbers of patients in each groupdied violently (suicide and trauma combined), 7 in theplacebo group versus 6 with simvastatin. A total of 35placebo patients developed fatal cancers during thestudy, compared with 33 in the simvastatin group.Twelve patients receiving placebo and 9 receivingsimvastatin had fatal cancers that arose in the gastro-intestinal tract. The incidence of cerebrovasculardeath also was similar in the 2 groups, as was thenumber of deaths due to other cardiovascular causes.

Simvastatin therapy significantly decreased the rel-ative risk of major coronary events (coronary death,nonfatal MI, silent MI, resuscitated cardiac arrest). Atotal of 622 major events occurred in the placebogroup (28%), compared with 431 in the simvastatincohort (19%), reflecting a 34% reduction in relativerisk (p,0.00001). The relative risk of definite orprobable nonfatal MI was decreased 37%(p ,0.00001). The 6-year probability of remainingfree of major coronary events was 70.5% for theplacebo group and 79.6% for the simvastatin group.

Treatment with simvastatin also had a favorableand statistically significant impact on the risk of anycoronary event, decreasing the relative risk 27%(p ,0.00001). The 6-year probability of escaping anycoronary event was 56.7% in the placebo group and66.6% in the simvastatin group.

Other statistical analyses showed additional bene-fits of simvastatin therapy. The reduction in relativerisk of death or any atherosclerotic event in the sim-vastatin cohort (p,0.00001) was 26%. The probabil-ity of avoiding both types of events over 6 years was53.0% in the placebo group and 62.9% in the simva-statin group. The active treatment group also had adecreased need for coronary bypass surgery or coro-nary angioplasty. Overall, 383 patients in the placebogroup and 252 in the simvastatin group required by-pass surgery or angioplasty during the trial, reflectinga 37% reduction in relative risk (p,0.00001). Finally,a post hoc analysis showed that 98 placebo patientshad fatal or nonfatal cerebrovascular events, com-pared with 70 in the simvastatin group. The differencereflected a significant 30% reduction in the relativerisk of such events (p5 0.024).

The reduction in the risk of a major coronary eventwas proportional and almost linearly related to thereduction in serum levels of LDL cholesterol after 1year. In one third of the simvastatin-treated patients,LDL cholesterol was decreased by,34%; in thesecond third, it was decreased by 34–44%; and in thelast third, it was decreased by 44–70%. Subsequentrates of major coronary events were 18.2%, 13.8%,

TABLE I Baseline and 1-Year Serum Cholesterol andTriglycerides for Simvastatin Therapy Patients* in the 4SStudy

Baseline*(mg/dL)

Year 1(mg/dL)

Change(%)

Total cholesterol 260 188 228LDL cholesterol 188 118 237HDL cholesterol 46 48 15Triglycerides 131 107 218

*Baseline values were not significantly different from the placebo group.4S Study 5 Scandinavian Simvastatin Survival Study; HDL 5 high-density

lipoprotein; LDL 5 low-density lipoprotein.

TABLE II Relative Risk Reduction with Lipid-Lowering Therapywith Simvastatin in the 4S Study

Reduction(%) p Value

Total mortality 30 0.0003Coronary heart mortality 42 0.00001Major coronary events* 34 0.00001

Women 35 0.01Patients $60 yr 29 0.0001

Any coronary event 27 ,0.00001Death or any atherosclerotic event 26 ,0.00001Cerebrovascular event 30 0.024Need for coronary bypass or angioplasty 37 ,0.00001

4S Study 5 Scandinavian Simvastatin Survival Study.*Major coronary event defined as coronary death, definite or probable

nonfatal myocardial infarction, resuscitation from cardiac arrest, and occur-rence of silent myocardial infarction.

54T THE AMERICAN JOURNAL OF CARDIOLOGYT VOL. 82 (10B) NOVEMBER 26, 1998

and 10.6% in these tertiles, respectively. For eachadditional percent point reduction in LDL cholesterol,the risk was decreased by 1.7% (p,0.00001).10

Collectively, the findings demonstrated major, sta-tistically significant reductions in prespecified end-points, including primary and secondary events.

MAJOR SUBGROUP ANALYSESThe size and duration of the 4S trial, combined

with the large number of “hard” endpoints, permittedmeaningful statistical analyses of various patient sub-groups.11 Of particular interest was the potential effectof lipid-lowering therapy among women and olderpatients, who were among the patient subgroups spec-ified for statistical analysis in the original study de-sign. Simvastatin significantly decreased the incidenceof major coronary events in both subgroups.

A total of 52 women died during the study, too fewdeaths to permit a statistical evaluation of simvasta-tin’s possible effect on mortality. The deaths com-prised 25 women in the placebo cohort and 27 in thesimvastatin group. Thirty of the 52 deaths resultedfrom coronary artery disease, 17 in the placebo groupand 13 in the simvastatin group. Simvastatin signifi-cantly decreased the relative risk of major coronaryevents in women, 59 versus 91 for placebo. The prob-ability that a woman would remain free of majorcoronary events was 77.7% with placebo and 85.1%with simvastatin, resulting in a 35% reduction in rel-ative risk (p5 0.01).

Simvastatin had a pronounced effect on mortalityand major clinical events in older patients, defined as$60 years of age. In the placebo group, 167 deathsoccurred in patients aged at least 60 years, comparedwith 127 in the simvastatin group. The differencereflected a 27% reduction in relative risk (p5 0.01).With respect to major coronary events, 319 occurredin the placebo group and 243 in the simvastatin co-hort, a 29% reduction in relative risk (p5 0.0001).

Other analyses revealed a consistency of simvas-tatin’s effect across a wide range of clinical variables.Notably, a post hoc analysis showed that patients withdiabetes derived significant benefit from lipid-lower-ing therapy.12 Although patients with diabetes had ahigher risk of major coronary events than those with-out the disease, treatment with simvastatin decreasedthe relative risk for patients with diabetes by 55%(p 5 0.002). Their relative risk of death from anycause during the study period was decreased by 43%(p 5 0.087), and their relative risk of any atheroscle-rotic event was decreased 37% in the simvastatincohort (p5 0.018).

Lipid-lowering therapy with simvastatin substan-tially decreased the relative risk of major coronaryevents irrespective of baseline risk factors such asdiabetes, hypertension, and smoking.11 Patients withhypertension had a 37% lower relative risk with sim-vastatin, and smokers and ex-smokers had a 31%reduction in relative risk. The pattern was similar forbaseline medication use: simvastatin therapy was as-sociated with a 34% reduction in the relative risk ofmajor coronary events among users of aspirin, a 36%

reduction in patients takingb blockers, and a 38%reduction among those taking calcium antagonists atentry to the study.

Only 21% of 4S patients had no history of MI atbaseline, a subgroup presumably at lower risk thanthose with a history of previous infarction. A post hocanalysis revealed a 26% reduction in the relative riskof major coronary events in this lower-risk group.11

Finally, simvastatin decreased the risk of majorcoronary events irrespective of baseline total, LDL-,or HDL-cholesterol level. Particularly noteworthy,simvastatin decreased the relative risk by 35% amongpatients who were in the lowest LDL quartile at base-line and by 36% in patients who had the highest LDLlevels at study entry.13

COST-EFFECTIVENESS ANALYSESProspectively collected hospital admissions from

4S have been analyzed to evaluate the impact ofsimvastatin on healthcare resource utilization.14 In theplacebo group, 937 patients required 1,905 hospital-izations for acute cardiovascular events or for coro-nary revascularization procedures. In contrast, 720simvastatin patients required 1,403 hospitalizationsfor the same causes (p,0.0001).

The per-patient cost of hospitalization for cardio-vascular causes was estimated by use of diagnosis-related group (DRG) allowances from the US Medi-care system. During the median 5.4-year follow-up in4S, simvastatin decreased the cost for cardiovasculardisease by 31%, or $3,872 per patient. At the time theanalysis was performed, the average daily cost ofsimvastatin therapy in the United States was $2.30.Using that figure, the effective cost of simvastatinwould be decreased by 88%, to $0.28 per day, throughdecreased use of healthcare resources.

Another analysis of prospectively collected datafrom 4S included a model of projected survival for the20 years after completion of the trial.15,16 The modelwas based on the survival benefits observed during thedouble-blind phase of the trial. In this analysis, thecost per life-year saved by use of simvastatin therapywas approximately $7,300. By comparison, a similarassessment of the cost-effectiveness of antihyperten-sive therapy in Sweden revealed a cost per life-yearsavings of $4,400 to $17,000 in patients aged 45–69years and with diastolic blood pressure of 95–99 mmHg.15

Any cost-effectiveness analysis of preventive treat-ment must take into account 3 basic variables: (1) thecost of therapy, (2) the risk reduction, and (3) theabsolute risk of the population treated. In the 4S trial,the annual risk of a major coronary event was 4.7% inthe placebo group.

In lower-risk populations (e.g., no symptoms ofcoronary artery disease and only 1 risk factor), thecost per life-year saved will likely be much higherthan the costs calculated from 4S. In patients whohave no coronary artery disease but do have a clusterof risk factors (high cholesterol, family history, smok-ing, hypertension), the cost per life-year saved will besimilar to the 4S calculations.

A SYMPOSIUM ON CHOLESTEROL AND CORONARY DISEASE 55T

DISCUSSIONThe 4S trial demonstrated that aggressive lipid-

lowering therapy with simvastatin can significantlydecrease the relative risk of total mortality, coronarymortality, and major coronary events. The effective-ness of simvastatin was related to the level of LDLcholesterol achieved with therapy. The percent reduc-tion of LDL cholesterol exhibited a near-linear rela-tion with the relative risk reduction for coronary arterydisease events. No evidence was seen of a thresholdLDL level below which no benefit occurred. The drugwas well tolerated, with only 5.7% of patients in eachrandomized group discontinuing treatment because ofadverse events.

The impact of simvastatin on mortality and coro-nary artery disease events emerged during the first 1–2years of therapy, and the survival and coronary eventcurves continued to diverge for the duration of follow-up. The data indicate that adding simvastatin, at thedoses used in 4S, to the treatment of 100 coronaryartery disease patients would prevent 4 of the 9 cor-onary deaths that would be expected to occur over 6years, the approximate follow-up period in 4S. Withinthat same time frame, simvastatin therapy would pre-vent 7 of the expected 21 nonfatal MIs and allow 6 of19 patients to avoid myocardial revascularization pro-cedures.9

The 4S trial was the first clinical trial to demon-strate that aggressive lipid-lowering therapy can de-crease all-cause mortality. Importantly, the studyshowed that lipid-lowering therapy is not associatedwith an increased risk of noncardiovascular mortality.

1. American Heart Association. Heart and Stroke Facts 1997. Statistical Supple-ment. Dallas, TX: American Heart Association, 1997.2. Lowering blood cholesterol to prevent heart disease: consensus conference.JAMA 1985;253:2080–2090.

3. Study Group, European Atherosclerosis Society. Strategies for the preventionof coronary heart disease: a policy statement of the European AtherosclerosisSociety.Eur Heart J1987;8:77–88.4. Expert Panel on Detection, Evaluation, and Treatment of High Blood Choles-terol in Adults. Summary of the second report of the National CholesterolEducation Program (NCEP) Expert Panel on Detection, Evaluation, and Treat-ment of High Blood Cholesterol in Adults (adult treatment panel II).JAMA1993;269:3015–3023.5. Pyorala K, DeBacker G, Graham I, on behalf of the Task Force. Prevention ofcoronary heart disease in clinical practice. Recommendations of the Task Forceof the European Society of Cardiology, European Atherosclerosis Society andEuropean Society of Hypertension.Eur Heart J1994;15:1300–1331.6. Muldoon MF, Manuck SB, Matthews KA. Lowering cholesterol concentra-tions and mortality: a quantitative review of primary prevention trials.Br J Med1990;301:309–314.7. Rossouw JE, Lewis B, Rifkind BM. The value of lowering cholesterol aftermyocardial infarction.N Engl J Med1990;323:1112–1119.8. Ravnskov U. Cholesterol lowering trials in coronary heart disease: frequencyof citation and outcome.Br J Med1992;305:15–19.9. Scandinavian Simvastatin Survival Study Group. Randomized trial of choles-terol lowering in 4444 patients with coronary heart disease: the ScandinavianSimvastatin Survival Study (4S).Lancet1994;344:1383–1389.10. Pedersen TR, Olsson AG, Faergeman O, Kjekshus J, Wedel H, Berg L,Wilhelmsen L, Haghfelt T, Thorgeirsson T, Pyo¨rala K, Miettinen TA, Christ-ophersen B, Tobert JA, Musliner TA, Cook TJ. Lipoprotein changes and reduc-tion in the incidence of major coronary heart disease events in the ScandinavianSimvastatin Survival Study (4S).Circulation 1998;97:1453–1460.11. Kjekshus J, Pedersen TR, for the Scandinavian Simvastatin Survival StudyGroup. Reducing the risk of coronary events: evidence from the ScandinavianSimvastatin Survival Study (4S).Am J Cardiol1995;76(suppl):64C–68C.12. Pyorala K, Pedersen TR, Kjekshus J, Faergeman O, Olsson AG, ThorgeirssonG. Cholesterol lowering with simvastatin improves prognosis of diabetic patientswith coronary heart disease. A subgroup analysis of the Scandinavian SimvastatinSurvival Study (4S).Diabetes Care1997;20:614–620.13. Scandinavian Simvastatin Survival Study Group. Baseline serum cholesteroland treatment effect in the Scandinavian Simvastatin Survival Study (4S).Lancet1995;345:1274–1275.14. Pedersen TR, Kjekshus J, Berg K, Olsson AG, Wilhelmsen L, Wedel H,Pyorala K, Miettinen T, Haghfelt T, Faergeman O, Thorgeirsson G, Jo¨nsson B,Schwartz JS. Cholesterol lowering and the use of healthcare resources: results ofthe Scandinavian Simvastatin Survival Study.Circulation 1996;983:1976–1802.15. Jonsson B, Johannesson M, Kjekshus J, Olsson AB, Pedersen TR, Wedel H.Cost effectiveness of cholesterol lowering: results from the Scandinavian Sim-vastatin Survival Study (4S).Eur Heart J1996;17:1001–1007.16. Johannesson M, Jo¨nsson B, Kjekshus J, Olsson AG, Pedersen TR, Wedel H.Cost-effectiveness of simvastatin treatment to lower cholesterol levels in patientswith coronary heart disease. Scandinavian Simvastatin Survival Study Group.N Engl J Med1997;336:332–336.

56T THE AMERICAN JOURNAL OF CARDIOLOGYT VOL. 82 (10B) NOVEMBER 26, 1998