Clin. Cardiol.

23,682-688 (2000)

Comparison of the Efficacy and Tolerability of Sirnvastatin and Atorvastatin in

the Treatment of Hypercholesterolemia
CESAR s.RECJD n, M.D., M.D.,* ADRIAN DOBS,
STELLA ACOSTA, M.D.,?

Philippine Heart Center, Quezon City, Philippines;* Service of Endocrinology, Hospital de San Jose, Santafe de Bogota, Colombia;
?Department of Medicine, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA

Summary atorvastatin 10 mg and 20 mg. The mean percent reductions

for LDL cholesterol from baseline to Week 6 ranged from
Background: Simvastatin and atorvastatin are effective 3542% for the entire study cohort. An LDL cholesterol level
statinsfor treating hypercholesterolemia. I130mddl(3.4 mmol/l) was achieved in approximately70%
Hypothesis: The study was undertaken to compare the effi- of patients treated with both drugs in this study. Simvastatin
cacy and tolerability of simvastatin 20 and 40 mdday and ator- and atorvastatin were well tolerated at the doses studied.
vastatin 10and 20 mdday. Conclusion: In patients with hypercholesterolemia, the
Methods: In this multinational, open-label,crossoverstudy, most commonly used doses of simvastatin and atorvastatin
258 patients with primary hypercholesterolemia were random- produced similarchanges in LDL cholesteroland achieved an
ized after 4 weeks of diet plus placebo to once-daily adminis- LDL cholesterol level I130 mddl(3.4 mmoM) in a similar
tration of a starting dose sequence of simvastatin (20 mg) or number of patients. Both statins were well tolerated.
atorvastatin (10 mg), or a higher dose sequence of simvastatin
(40mg) or atorvastatin (20 mg) for 6 weeks. Patients were then
switched after a 1-week washout to the correspondingstarting Key words: simvastatin, atorvastatin, 3-hydroxy-3-methyl-
or higher dose of the alternatedrug for a second 6-week period. glutaryl coenzyme A reductase inhibitors, hypercholesterol-
The primary endpoint was the mean percent changefrom base- emia, low-density lipoprotein cholesterol, triglycerides
line to Week 6 in low-density lipoprotein (LDL) cholesterol;
percent changes from baseline in total cholesterol,high-densi-
ty lipoprotein cholesterol, triglycerides, and apolipoprotein B Introduction
were also compared. Safety was assessed through adverse ex-
periences and laboratory measurements. The strong positive relationship between plasma choles-
Results: Both statins produced statisticallysignificant im- terol and coronary heart disease (CHD) incidence extends
provements in all measured plasma lipids and lipoproteins. over a wide range of cholesterol concentrations.1-3 Effective
The main treatment comparison showed no statisticallysignif- lipid-loweringtreatmenthas thus assumed an importantrole in
icant difference in changes in LDL cholesterol and triglyc- the prevention of atherosclerosisand coronary artery disease
erides, whereby the overall effects were comparable when (CAD)." Of the lipid-lowering agents currently available, the
doses of 20 mg and 40 mg of simvastatinwere compared with 3-hydroxy-3-methylglutaryl coenzyme A (HMGCoA) reduc-
tase inhibitors (statins) are the most effective in lowering plas-
ma levels of total and low-density lipoprotein (LDL) choles-
terol and are generally well tolerated? The therapeutic effects
of these agents are related to their ability to inhibit the rate-lim-
iting enzyme HMG CoA in cholesterol synthesis. Cholesterol
~~
synthesis is thereby decreased and receptor-mediated catabo-
This study was sponsored by a grant from Merck & Co., Inc. lism of LDL cholesterol is enhanced.
Simvastatin (ZocoP, Merck & Co., Inc., Rahway, N.J.),
Address for reprints:
an Hh4G-CoA reductase inhibitor, has been demonstrated to
Dr. Cesar S. Recto be effective in lowering LDL cholesterol and triglycerides as
Philippine Heart Center well as in raising high-densitylipoprotein (HDL) cholesterol?
East Avenue It was also the first HMG CoA reductase inhibitor to demon-
Quezon City, Philippines strate an unequivocal benefit on survival in hypercholes-
Received: August 24, 1999 terolemic patients with documented CAD. In the Scandina-
Accepted with revision: November 4, 1999 vian Simvastatin Survival Study (4S), treatment with simva-
 C. S. Recto e t a / . :Comparison of simvastatin and atorvastatin 683

statin (20to40 mg/day) for an average of 5.4 years significant- periods were separated by a 1-week washout period during
ly reduced the risk of all-cause mortality by 30%,coronary which patients did not receive any cholesterol-loweringagent.
mortality by 42%, and nonfatal myocardial infarction by Throughout the study, patients were counseled to follow a
37%.7 In this large, long-term study, the safety profile of sim- standard lipid-lowering diet that had a daily intake restriction
vastatin was similar to that of placebo. Furthermore, results of 50%carbohydrates, 20% protein, 30% fat (polyunsaturat-
from an additional 2-year follow-up of all patients completing edsaturated fatty acid ratio of 1/1) and 1300 mg cholesterol.
the original 4s study show that the survival curves continue to The protocol was approved by local ethics committees, and
diverge, demonstrating a continued prevention of mortality in all patients gave informed consent in accordance with the
patients with CHD for a period of at least 8 years.8 Long-term Declaration of Helsinki as revised in 1983. Simvastatin and
treatment with simvastatin has also been shown to slow the placebo were manufactured by Merck & Co., Inc., Rahway,
progression of diffuse and focal coronary atherosclerosis in New Jersey, while atorvastatin was provided as commercially
patients with hypercholesterolemia and CAD.9 available 10 and 20 mg tablets.
Atorvastatin (LipitoP, Parke-Davis Div. of Warner-Lambert
Co., Moms Plains, N.J.) is a new HMG CoA reductase inhib- Patient Selection
itor. It has been shown to be effective in lowering LDL choles-
terol and triglycerides.l03 Its effects on cardiovascular mor- Participants in this study were men and women aged 2 1 to
bidity and mortality and its long-term safety are still unknown. 70 years who met the 1ipidAipoproteinentrance criteria and
Simvastatin and atorvastatin are considered to be the most were in general good health based on medical history, physical
effective agents of the HMG-CoA reductase inhibitors in treat- examination, and laboratory evaluations. LipidAipoproteinen-
ing hypercholesterolemia. Using a randomized, open-label, trance criteria consisted of a plasma LDL cholesterol concen-
crossover design, the present study was to provide a direct tration 2 3.4 mmoVl(130 mg/dl) and a plasma triglyceride lev-
comparison of the lipid-lowering efficacy and tolerability of el 14.0 mmoVl(350 mg/dl); these criteria must have been met
typical starting doses of simvastatin and atorvastatin and sin- after 2 weeks of diet plus placebo treatment.
gle titration doses in approximately 260 patients with primary Patients with any of the following were excluded from
hypercholesterolemiadespite dietary compliance. study participation: secondary hyperlipoproteinemia; types I,
111, IV, or V hyperlipidemia; myocardial infarction, coronary
angioplasty or coronary bypass surgery within 3 months of
Methods trial entry; acute coronary insufficiency; active liver disease;
renal insufficiency; partial ileal bypass; obesity (body weight
Trial Design > 50% of ideal); uncontrolled or insulin-dependent diabetes
mellitus; uncontrolled hypertension; and excessive alcohol
This multicentered, randomized, open-label, crossover consumption (> 10 dnnkdweek). To qualify for study enroll-
study had a split-plot design and consisted of a 4-week diet ment, patients must not have received therapy with a lipid-
plus placebo baseline period, followed by two 6-week active lowering agent for at least 6 weeks (8 weeks for fibrates and
treatment periods. The study was conducted at 20 centers in 12 months for probucol). Concurrent use of cimetidine, im-
14 countries. The design of the study is shown in Figure I . munosuppressants, anticoagulants, and systemic antifungal
During the first treatment period, eligible participants compli- agents of the mole class was disallowed throughout the study.
ant with placebo and a lipid-lowering diet were randomized to
once-daily treatment with a starting dose sequence of simvas- Study Variables
tatin (20 mg) or atorvastatin (10 mg), or a higher dose se-
quence of simvastatin (40 mg) or atorvastatin (20 mg). At the Clinic visits were scheduled at initiation of diet plus place-
end of the first treatment period, patients were switched to the bo therapy (Week -4), after 2 weeks of placebo treatment
corresponding starting or higher dose of the alternate drug for (Week -2), upon randomization to active treatment (Day 1,
an additional 6 weeks of treatment. The two active treatment Week 1) and after each 6-week active treatment period

Starting dose Simvastatin 20 rng - Atorvastatin 10 mg

Atorvastatin 10 rng -

I
Simvastatin 20 mg

Randomization -
Simvastatin 40 mg - Atorvastatin 20 mg
Higher dose Atorvastatin 20 rng - Simvastatin 40 mg

Study week 1-6 7 8-1 3

FIG.1 Study design. Subjects were washed out from drug during Week 7.
684 Clin. Cardiol. Vol. 23, September 2000

(Weeks 6 and 13).Patients were contacted by telephone at the tatin 40 mg versus atorvastatin 10 mg. Within-group compar-
conclusion of the washout period (Week 7). Plasma total isons of the percent change from baseline were done using
cholesterol, HDL cholesterol, triglyceride, and apolipopro- paired t-tests.
tein B levels were measured by a central laboratoryfrom 12-h The proportions of patients with clinical and laboratory ad-
fasting blood samplescollected at each clinic visit. Low-den- verse experiences were compared among the simvastatin
sity lipoprotein cholesterol levels were calculated using the (20-40 mg) and atorvastatin (10-20 mg) treatments using
Friedewald equation.’* McNemar’s test; four pairwise comparisons were subsequent-
Adverse events were monitored throughout the trial. A ly performed using McNemar’s test for paired subjects and
blood sample was obtained at each clinic visit for evaluation of Fisher’s exact test for unpaired subjects. Possible treatment
hematologic parameters (hemoglobin, hematocrit, platelet differences in the percent change from baseline in plasma fib-
count, white blood cell count, and differential) and blood rinogen were evaluated using an analysis of covariance that
chemistryparameters [aspartateaminotransferase(AST), ala- included baseline fibrinogen levels as covariate.The median
nine aminotransferase (ALT), alkaline phosphatase, total changes from baseline in liver function parameters (AST,
bilirubin, C-reactive protein, creatine phosphokinase (CPK), ALT, CPK), serum testosterone, and serum estradiol, pooled
creatinine, glucose, fibrinogen, serum electrolytes]. At these across the two periods, were evaluated using an ANOVA (on
visits, a urine sample was also collected and analyzed for spe- ranked values) with treatment, dose, treatment-by-dose inter-
cific gravity, protein, and cells, and a physical examination action, period, investigator,and patient (nested within investi-
which included measurement of vital signs and body weight gator-by-dose)as factors; changes from baseline within each
was performed. A 12-leadelectrocardiogram (ECG) was ob- group were assessed using the Wilcoxon signed-rank test.
tained within 3 months of study initiation and again at its com- Changes from baseline in vital signs and body weight were
pletion if clinically indicated. analyzed using an ANOVA model similarto that described for
analysis of plasma lipids and lipoproteins.
Statistical Methodology All statistical comparisons were two-sided, and statistical
significance was defined as p c 0.05, except for interaction
The hypotheses of the study were that simvastatin (20-40 terms and evaluationsof carryovereffects which were tested at
mg) and atorvastatin (10-20 mg) will have a similar effect on a signtficancelevel of 0.10.
LDL cholesterol(primary) and triglycerides (secondary).The
study was designed for 2 12 evaluable patients to estimate the
differencein LDL cholesterol reduction between simvastatin Results
(20-40 mg) and atorvastatin(10-20 mg) with an absolutepre-
cision of 1.9% and a degree of confidence equal to 95%. With Patient Characteristics
this sample size, the study would have 95% power to detect a
3.5% difference in the mean percent change from baseline in
LDL cholesterol between the simvastatin and atorvastatin Of the 258 patients randomized to active treatment, approx-
groups assuminga within-patientvariability of 9.9% (correla- imately half (51%) were male with a median age of 55 years.
tion = 0.5) and significancelevel = 0.05,using two-sidedtests. Approximately two-thirds of the patients (64%)were Cauca-
Therefore,if the difference is not significant and the precision sian, 16% were Hispanic, 12% were Asian, 5% were Mestizo,
of the estimate equals 1.9%because of the high power of the and 2% were black. Over one-half (54%) of the patients had
trial, it is likely that the differencewould be small and provide received prior lipid-loweringtherapy, of which HMG CoA re-
strong evidence that the treatments are similar. ductase inhibitors were the most common. Baseline lipid and
lipoprotein values for all randomized patients are summarized
Efficacy analyses were performed using an intention-to-
treat approach and included all randomized patients with a in Table I. In all, 247 (96%) patients completed the 13-week
baseline measurement and at least one post-treatment mea- study. Of the 11 patients who discontinued the study prema-
surement in each of the two active treatment periods. Percent turely, 8 withdrew for adverse experiences, 1 was uncoopera-
tive, 1 discontinued for a protocol violation, and 1 withdrew
changes from baseline in lipids and lipoproteins after 6 weeks
of treatment were analyzed for comparative efficacy between for an insufficienttherapeuticresponse.Treatmentcompliance
the simvastatin (20-40 mg) and atorvastatin (10-20 mg) rates in the study were high and comparablebetween simvas-
tatin and atorvastatin.
groups analysis of variance (ANOVA) with factors for treat-
ment, dose, treatment-by-dose interaction, period, investi-
gator, and patient (nested within investigator-by-dose).The Efficacy
baseline value was defined as the last measurementbefore ad-
ministration of study drug in the first active treatment period. A total of 249 patients had efficacy measurements from
In addition to the overall simvastatin-atorvastatincomparison, complete treatment sequences and were included in the inten-
four preplanned painvise comparisons were performed using tion-to-treat analyses for the primary endpoint, LDL choles-
contrasts of the treatment effects: simvastatin 20 mg versus terol. No si@cant (p > 0.10) carryover effects or treatment-
atorvastatin 10 mg, simvastatin 40 mg versus atorvastatin 20 by-investigatorinteractions were observed for any of the lipid
mg, simvastatin 20 mg versus atorvastatin20 mg, and simvas- efficacy variables.
 C. S. Recto et al.: Comparison of simvastatin and atorvastatin 685

TABLE
I Baseline summary for efficacy variables-all patients weeks resulted in mean overall reductions in LDL cholesterol
of 35 to 4296, median overall decreases in plasma triglycerides
Variable N Mean SD Median Range
~~ ~~ levels of 21 to 25%, and mean overall elevations in HDL
LDL cholesterol, 257 193.4 53.8 183.0 105.0 429.0 cholesterol of 8 to 9%. None of the pairwise comparisonsbe-
mg/d (mmoVI) 5.0 1.4 4.7 2.7-1 1.1 tween simvastatin 20 mg and atorvastatin 10 mg, or simvas-
Total cholesterol, 257 277.4 58.2 269.0 184.0-657.0 tatin 40 mg and atorvastatin20 mg, was statistically significant
mg/dl (mmoVI) 7.2 1.5 7.0 4.8-17.0 for any lipid or lipoprotein variables. Results of these pairwise
HDL cholesterol, 257 47.2 12.2 46.0 10.0-96.0 comparisons are displayed for total cholesterol, LDL choles-
mg/d (mmoVI) 1.2 0.3 1.2 0.3 to 2.5 terol, HDL cholesterol, and triglycerides in Figure 2. As ex-
Triglycerides, 257 181.3 79.6 166.0 52.0-526.0 pected, a doubling of the daily dose from 20 to 40 mg for sim-
mg/dl (mmoM) 2.0 0.9 1.9 0.6-5.9
vastatin or from 10 to 20 mg with atorvastatin resulted in a
Apolipopmtein B, 257 177.4 42. I 171.0 98.0-501 .O
further mean reduction of 6% in LDL cholesterol.Overall,ap-
mkw proximately 70% of simvastatin-and atorvastatin-treatedpa-
LDL-C I HDL -C 257 4.5 3.0 3.9 1.642.9
tients achieved an LDL cholesterol level of 5 130 mgldl(3.4
Total-C /HDL-C 257 6.4 4.2 5.8 2.9-65.7
mmoV1) after 6 weeks of treatment (Fig. 3). The overall com-
Abbreviations: N = number of patients, SD = standard deviation, parison between simvastatinand atorvastatinand the four pair-
LDL-C = low-density lipoprotein cholesterol,HDL-C = high-densi- wise treatment comparisons were not statisticallydifferent in
ty lipoprotein cholesterol. the proportion of patients achieving this goal.
In all four treatment groups, the reduction in triglycerides
was p a t e r in patients whose baseline plasma triglyceridelev-
el was 1200 mg/dl (median reductions range from 33 to 39%
There were statistically significant (p < 0.001) reductions at 6 weeks) than in those whose baseline level was < 200 mgldl
from baseline in plasma levels of LDL cholesterol,total choles- (median reductions range from 14 to 17%).In the absence of a
terol, triglycerides, and apolipoprotein B, and in the ratios placebo group, aportion of this effect may be attributedto a re-
of LDL-to-HDL cholesterol and total-to-HDL cholesterol at gression to the mean.
Week 6 in all treatment groups. Plasma levels of HDL choles- The effects of gender (male, female), race (Caucasian, oth-
terol were significantly elevated (p c 0.001) from basehe to er) and age (I 55 years, > 55 years, based on median age of the
Week 6 in all groups. The mean percent changes from baseline total population) on the percent change from baseline in LDL
to Week 6 for each of the four treatment groups are shown in cholesterolwas assessed using an ANOVA model. Gender or
Table II; also summarized is the p value for the overall simvas- race had no effect on the observed treatment effectsof simvas-
tatin versus atorvastatincomparisons. tatin and atorvastatin in this study. A somewhat larger reduc-
No significant differences in percent reduction in LDL tion in LDL cholesterol was observed in patients > 55 years
cholesterol or triglycerides were observed between simvas- compared with those I 5 5 years of age; this difference was ap-
tatin and atorvastatin based on the primary intention-to-treat parent in all four treatment groups, and the treatment-by-age
analysis. Treatment with simvastatin or atorvastatin for 6 interaction was not statistically sigmficant.

TABLE
II Mean (+ SD)percent change from baseline to Week 6 in plasma lipids and lipoproteins
Simvastatin Atorvastatin p Value
20 mg 40 mg lOmg 20 mg Simvastatin
Lipid/lipoprotein (N= 125) (N = 126) (N = 125) (N = 126) vs.
parametera (%I (%I (%I atorvastatin
LDL cholesterol -34.8 f 14.0 -41.0+ 15.9' -36.7k 13.3 -42.1 k 15.6' 0.102
Total cholesterol -25.7 + 10.7 -30.0+ 12.6 -26.8 + 10.4 -31.1 + 12.3 0.122
HDL cholesterol 8.7 k 13.4 9.3 + 13.7 8.1 k 12.4 8.5 k 12.8 0.407
Triglycerides -21.5k23.6 -21.4 + 23.3 -22.0k28.5 -25.4226'4 0.793
Apolipoprotein B -27.3k 14Sd -32.9+ 12.6" -29.3k 12.4d -34.0k 13.7' 0.084
LDL/HDL cholesterol -39.3 + 14.5 -45.5+ 15.oc -40.7 + 14.1 -46.2 + 14.7 ' 0.214
Total/HDLcholesterol -30.8k11.8 -35.3 k 12.3 -31.6k 11.8 -35.8+ 12.5 0.346
a Within treatment group changesfor each lipid/lipoprotein were statistically significant (p < 0.001).
Statistical significancedefined as p < 0.05.
C N =124.
N = 123.
Median.
Abbreviationsas in Table I.
686 Clin. Cardiol. Vol. 23, September 2000

FIG.3 Percentage of patients achieving target low-density lipopro-

tein cholesterol (LDL-C)level of I130mddl(3.4 mmoV1) by Week
6. El = Simvastatin 20 mg, 0= atorvastatin 10 mg, = simvastatin
40 mg, I= atorvastatin 20 mg.

riods with a I 1% incidence in any treatment group. There was

no statistically significant difference (p = 0.585) between sim-
vastatin and atorvastatin with respect to the incidence of drug-
related adverse experiences. The adverse experiences occur-
ring during the washout period were counted under the pre-
vious therapy taken. There was also no statistically significant
difference between simvastatin and atorvastatin in the overall
frequency of laboratory adverse experiences (p = 0.629)or in
FIG.2 Mean percent change from baseline and standard error in
lipids and lipoproteins following 6 weeks of once-daily treatment frequency of drug-related adverse laboratory experiences (p =
with simvastatin (20 or 40 mg) or atorvastatin (10 or 20 mg). 0.344). Only one patient in this study had an elevation in A I T
> 3 times the upper limit of normal, which occurred following
treatment with simvastatin 40 mg (ALT = 78 mU/ml, upper
limit of normal range = 25 mU/ml). At a follow-up evaluation
Clinicaland Laboratory Safety Evaluations 2 weeks later, ALT levels in this patient had returned toward
the normal range. One patient discontinued the study prema-
Both HMG CoA reductase inhibitors were well tolerated turely due to drug-related clinical adverse experiences (arthral-
during the active treatment period (Table III). There was no gia, headache, nausea) while receiving simvastatin (20 mg)
statistically significant difference (p = 0.170) in the frequency compared with five patients who withdrew from atorvastatin
of clinical adverse experiences between simvastatin and ator- therapy for the same reasons (two with 10 mg dose and three
vastatin. Table IV shows the drug-related clinical adverse ex- with 20 mg dose). Nausea andor abdominal pain were the
periences reported during the active treatment or washout pe- most common reasons for discontinuation of atorvastatin ther-

TABLEIII Patients reporting adverse experiences

Simvastatin Simvastatin Atorvastatin Atorvastatin
20 mg (n = 128)(%) 40 mg (n = 127)(%) 10mg (n = 127)(%) 20 mg (n = 129)(%)
One or more clinical AE 48 (37.5) 33 (26.0) 46 (36.2) 50(38.8)
Drug-related clinical AEs 13 (10.2) 4(3.1) lO(7.9) 12(9.3)
Withdrawn due to drug-relateda AEs l(0.8) 0 2(1.6) 3 (2.3)
Seriousclinical AEs 0 0 0 2 (1.6)
Simvastatin Simvastatin Atorvastatin Atorvastatin
20 mg (n = 127)(%) 40 mg (n = 126)(%) 10mg (n = 127)(%) 20 mg (n = 128)(%)
One or more laboratory AE 5 (3.9) 9 (7.1) 6 (4.7) 5 (3.9)
Drug-relateda laboratory AEs 3 (2.4) 5 (4.0) 2(1.6) 2(1.6)
ALT or AST elevations > 3 times upper limit of normal 0 l(0.8) 0 0
CPK elevations > 10times upper limit of normal 0 0 0 0
aRated as possibly, probably, or definitely drug-related by investigator.
' h o patients were hospitalized for angina.
Abbreviations: AE = adverse experiences,ALT or AST = alanine or aspartate aminotransferase, CPK = creatine phosphokinase, n = number of
patients.
 C. S. Recto et al.: Comparison of simvastatin and atorvastatin 687

TABLEIV Summary of common drug-related clinical adverse experiences(21% incidence)

Simvastatin Simvastatin Atorvastatin Atorvastatin
Adverse 20 mg 40 mg 10 mg 20 mg
experience (n = 128)(%) (n = 127)(%) (n = 127)(%) (n = 129)(%)
Asthenidfatigue 2(1.6) 2(1.6) 2(1.6) 0
Abdominal pain 0 0 l(0.8) 2(1.6)
Flatulence 3 (2.3) 0 0 0
Nausea 2(1.6) 0 2(1.6) 2(1.6)
Arthralgia 2(1.6) 0 0 0
Myalgia l(0.8) 0 2(1.6) 0

apy (n =4); the remaining patient discontinued atorvastatinfor LDLcholesterol goal of < 130mddl(3.4 mmol/l) was not sta-
muscle cramps. Two patients were hospitalized for angina tistically different at approximately50%for each treatment."
while receiving atorvastatin.Neither patient was withdrawn Results of the present large, multinational, randomized,
for the event and neither event was considered drug related. crossover study evaluating fixed-dosecomparisonsof simvas-
There were no statistically significant differences between tatin and atorvastatinindicate that simvastatinand atorvastatin
treatment groups in changes in fibrinogen, C-reactive protein, provide comparablereductions in plasma lipid and lipoprotein
or serum estradiol (data not shown). Significant (p ~ 0 . 0 5de-
) parameters among patients with hypercholesterolemiawhen
creases from baseline in serum testosterone levels were ob- administered at commonly used therapeutic doses. After 6
served with both atorvastatin and simvastatin treatment, with weeks of once-dailyadministration,both simvastatinand ator-
the decreasesbeing statistically(p<O.001)smaller for atorvas- vastatin produced comparable reductions in plasma levels of
tatin than for simvastatin. The changes in serum testosterone total cholesterol, LDL cholesterol, and triglycerides. Follow-
were small (mean 0.3 to 1 ng/ml) and considered clinically in- ing treatment with either Hh4G CoA reductase inhibitor,LDL
significant(average0.3to 1.O ng/ml). Small changes in serum cholesterol levels were reduced by 35 to 42%, and a doubling
alkaline phosphatase that were significantly (p < 0.OOl) dif- of the daily dose of either drug was associated with only a 6%
ferent were observed between atorvastatin and simvastatin. incremental reduction in LDL cholesterol. Notably, approxi-
While simvastatin produced median changes in serum alka- mately 70% of patients treated with simvastatinor atorvastatin
line phosphatase that were essentially unchanged or slightly in this study achieved an LDL cholesterollevel of I130mg/dl
reduced, atorvastatin produced significant (p < 0.01) median (3.4 mmol/l) recommended by the National Cholesterol Ed-
increases from baseline of 1 mU/ml. ucation Program (NCEP) ATP I1 Expert Panel for individuals
with more than two risk factors.I4
A positive correlation between plasma triglycerides and
Discussion CHD risk has been observed,15,l 6 although some have ques-
tioned whether triglyceride levels represent an independent
Very little comparative clinical data between atorvastatin risk factor for CHD.I7 The triglyceride-lowering activity of
and simvastatin existed at the time the present study was con- simvastatin'8-2iand atorvastatin'()has been well documented.
ducted. Early fixed dose comparisons of the lipid-loweringef- In the present study, simvastatin and atorvastatin produced
ficacy of simvastatinand atorvastatin,such as the CURVES of similar reductions in plasma triglyceride levels of 17 to 2 1 %,
efficacy for various doses of statins on LDL-cholesterol re- and there was an indication that the efficacy of both drugs in
duction in the study (CURVES),had relatively small numbers lowering triglycerides was enhanced in patients with triglyc-
of patients in several of the treatment groups and these studies eride levels >200 mg/dl. Observations of this baseline triglyc-
were not adequately designed or powered to allow cross-dose eride-dependenteffect of triglyceride lowering has been made
comparisons.Furthermore,most other comparativetrials with by others,18but due to a lack of a placebo control, this observa-
simvastatin and atorvastatin have involved a dose titration tion may be considered a regression to the mean effect.
treatmentcomparison, making it difficultto assess the compa- In the present study, the magnitude of the changes in plasma
rable dosage efficacy of each product. In addition, these stud- lipids and lipoproteinsfollowing treatment with simvastatinor
ies frequently evaluated only part of the dosage range avail- atorvastatin were as expected for dosage range studied. In ad-
able (e.g., simvastatin 80 mg excluded). However, several dition to producing clinically meaningful improvements in
studies have provided insight regarding the comparativeeffi- plasma lipids and lipoproteins,results of the study show no sig-
cacy of atorvastatin and simvastatin to have patients achieve nificant difference between simvastatinand atorvastatin in the
LDL cholesterol goals. For example, in a 1-year titrate-to-goal incidence of clinical adverse experiencesafter short-term ther-
study comparing commonly prescribed doses of atorvastatin apy. Although not statistically significant, numerically more
(10-20 mg) and simvastatin (10-20 mg) in patients with hy- patients discontinued therapy with atorvastatin (n = 5 ) than
percholesterolemia, the percentage of patients to reach an with simvastatin(n = 1) for drug-relatedclinical adverse expe-
688 Clin. Cardiol. Vol. 23, September 2000

riences.Four of the five patients who discontinuedatorvastatin 3. Chen Z, Pet0 R, Collins R, MacMahon S, Lu J, Li W Serum
therapy did so for abdominal pain andor nausea. Nausea ac- cholesterol and coronary heart disease in populations with low
cholesterol concentration.BrMed J 1991;303:276-282
companied by headache and arthralgia resulted in the prema- 4. National Cholesterol Education Program Expert Panel: Summary
ture discontinuationof only one patient receiving simvastatin. of the Second Report of the National Cholesterol Education Pro-
In general, there was no difference in laboratory changes gram (NCEP) Expert Panel on Detection, Evaluation and Treat-
between simvastatin and atorvastatin in this study. In particu- ment of High Blood Cholesterol in Adults (Adult Treatment Panel
lar, both HMG CoA reductase inhibitorsproduced similarele- II). JAm MedAssoc 1993;269:3015-3023
5 . Illingworth DR, Bacon SP, Larsen KK: Long-termexperience with
vations in serum transaminase and plasma CPK levels. Both HMG CoA reductase inhibitors in the therapy of hypercholes-
showed small decreases from baseline in testosterone, and terolemia.Atheroscler Rev 1988;18:161-1 87
atorvastatin produced elevationsin alkaline phosphatase. 6. ZOCORTM(simvastatin) tablets. Product information (1998)
Merck & Co., Inc., Whitehouse Station, N.J., USA
7. Scandinavian Simvastatin Survival Study Group: Randomisedtrial
Conclusion of cholesterol lowering in 4,444 patients with coronary heart dis-
ease: ScandinavianSimvastatin Survival Study (4s). Lancet 1994;
344: 1383-1389
Based on the primary intention-to-treatanalysis, the results 8. Pedersen TR, Wilhelmsen L, Faergeman 0, Thorgeirsson TG,
of this study demonstratethat simvastatin and atorvastatin pro- Troedsson L, KristiansonJ, Cook TJ, Kjekshus J: Extended follow-
vide similar low-density lipoprotein cholesterol and triglyc- up of patients in the ScandinavianSimvastatin Survival Study (4s)
eride reductions at commonly used doses in patients with pri- shows increased survival benefit of simvastatin therapy (abstr).
mary hypercholesterolemia.Over short-term use, both drugs Circulation 1998;98(17)(suppl)I-451
9. MAAS Investigators:Effect of simvastatin on coronary artheroma:
had similarsafety and tolerabilityprofiles. The Multicentre Anti-Atheroma Study (MAAS). Lancer 1994;
344:633-638
10. Nawrocki JW, Weiss SR, Davidson MH, Sprecher DL, Schwartz
Acknowledgment SL, Lupien PJ, Jones PH, Haber HE, Black DM: Reduction of
LDL-cholesterol by 25% to 60% in patients with primary hyper-
The authors wish to thank and acknowledge the other cholesterolemia by atorvastatin, a new HMG-CoA reductase in-
participating investigators:StellaAcosta, Hosp. San Jose, Co- hibitor. Arterioscler Thromb Vasc Bioll995; 15:678482
lumbia; Carlos Aguilar, Instituto Nacional De La Nutricion, 11. Bertolini S, Bon GB, Campbell LM, Famier M, Langan J, Mahla
G, Pauciullo P, Sirtori C , Egros F, Fayyad R, Nawrocki J W Ef-
Mexico; Oscar Kisen Briger, Inst. Modelo Card., Argentina; ficacy and safety of atorvastatin compared to pravastatin in patients
Elizabeth Cavallero, Hosp. Ramaos MejiafCntr., Argentina; with hypercholesterolemia.Atherosclerosis 1997;130 191-197
Jerome Cohen, St. Louis University Health Sciences Center 12. Friedewald WT, Levy RI, Fredrickson DS: Estimation of the con-
School of Medicine, U.S.; Adrian Dobs, Johns Hopluns Uni- centration of low-density lipoprotein cholesterolin plasma, without
use of preparative ultracentrifuge. Clin Chem 1977;18:499-502
versity, U.S.; Jose Ernestos dos Santos, Brazil; Mario Rul
13. Best J, Dart A, Hamilton Craig I, Jerums G, Nicholson G, Sullivan
Ferreira, Centro HospitalarDe Coimbra,Portugal;Jiri Frolich, D, Tallis G, West M, dEmden M, Bracs P A One Year Study
St. Paul’s Hospital,Canada; Claude Gagne, Centre Hospitalier Comparing Atorvastatin and Simvastatin in Patients with Hyper-
Universite Laval, Canada; Raul Gamboa, Clinica San Felipe, cholesterolernia. 66th Congress of the European Atherosclerosis
Peru; Pedro Garcia Hernandez, Hospital Universitario “Dr. Society,July 1996
14. Summary of the Second Report of the National Cholesterol Edu-
Jose EleuterioGonzalez”De Monterrey, Mexico; Helge Istad, cation Program (NCEP) Expert Panel on Detection, Evaluation,
Theresegaten Legesenter, Norway; Fernando Flores Lozano, and Treatment of High Blood Cholesterol in Adults (Adult
Hospital “Dr. Angel Leano”, Mexico; Leiv Ose, Lipidklinik- Treatment Panel 11). J A m MedAssoc 1993;23:3015-3023
ken-Rikshospitalet, Norway; Mark MacMahon, Chiltern In- 15. Hulley SB, Rosenman RH, Bow01 RD, Brand RJ:Epidemiology as
ternational Ltd., U.K.; Mary McGowan, New Hampshire a guide to clinical decisions: The association between triglyceride
andcoronary heart disease. NEnglJMed 1980;302:1383-1389
Heart Institute, U.S.; Dirk Myburgh, Institution of Aviation 16. Castelli WP: The triglyceride issue: A view from Framingham.Am
Medicine, South Africa; Trevor Orchard, Rangos Research H e m J 1984;112:432-437
Clinic, U.S.; Cesar Recto, Heart Center, Philippines; Miguel 17. Austin MA: Plasma triglyceride and coronary heart disease.
Sanchez-Palacios, Clinica Ricardo, Peru; Chee Eng Tan, Sing- Arterioscler Thromb Vasc Biol1991;11:2-14
apore General Hospital Singapore, Singapore; Mary Vino- 18. Stein EV, Lane M, Laskarzewski P Comparison of statins in hy-
pertriglyceridemia.Am J Cardiol I998;8 1(4A):66B49B
cour, del Hospital San Juan de Dios, Costa Rica; Alexei 19. Keech A, Collins R, MacMahonS, ArmitageJ, Lawson A, Wallend-
Yurenov, Cardiology Research Center, Russia. szus K, Fatemian M, Kearney E, Lyon V, Mindell J, Mount J, Painter
R, Parish S, Slavin B, SleightP, Youngman L, Pet0 R, for the Oxford
CholesterolGroup: Three-year follow-up of the Oxford Cholesterol
References Study: Assessment of the efficacyand safety of simvastatinin prepa-
ration for a large mortality study. EurHeat? J 199415:255-269
I . Levine GH, Keaney JF Jr, Vita JA: Cholesterolreduction in cardio- 20. Neuman MP, Neuman HR, Neuman J: Significantincrease of high-
vascular disease: Clinical benefits and possible mechanisms (re- density lipoprotein2-cholesterolunder prolonged simvastatintreat-
view). NEnglJMed 1995;332:512-521 ment.Atherosclerosis 1991;91:Sll-S19
2. Anderson KM, Catelli WP, Levy D: Cholesterol and mortality 30 21. The SimvastatinPravastatin Study Group: Comparison of the effi-
years of follow-up from the Framingham study. J A m Med Assoc cacy, safety and tolerability of simvastatin and pravastatin for hy-
1987;257:2176-2 180 percholesterolemia.Am J Curdiol1993;71: 1408-1414