See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/236315241

Relation of Gemfibrozil Treatment and Lipid Levels With Major Coronary

EventsVA-HIT: A Randomized Controlled Trial

Article in JAMA The Journal of the American Medical Association · March 2001
DOI: 10.1001/jama.285.12.1585

CITATIONS READS

905 111

12 authors, including:

Dorothea Collins Janet Wittes

65 PUBLICATIONS 10,219 CITATIONS

Statistics Collaborative, Inc.
260 PUBLICATIONS 44,645 CITATIONS
SEE PROFILE
SEE PROFILE

Vasilios Papademetriou Prakash Deedwania

U.S. Department of Veterans Affairs UCSF University of California, San Francisco
512 PUBLICATIONS 17,811 CITATIONS 639 PUBLICATIONS 34,186 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Janet Wittes on 26 May 2014.

The user has requested enhancement of the downloaded file.

 ORIGINAL CONTRIBUTION
Relation of Gemfibrozil Treatment and Lipid
Levels With Major Coronary Events
VA-HIT: A Randomized Controlled Trial
Sander J. Robins, MD
Dorothea Collins, ScD
Janet T. Wittes, PhD
Vasilios Papademetriou, MD
Prakash C. Deedwania, MD
Ernst J. Schaefer, MD
Judith R. McNamara, MT
Moti L. Kashyap, MD
Jerome M. Hershman, MD
Laura F. Wexler, MD
Hanna Bloomfield Rubins, MD, MPH
for the VA-HIT Study Group
C
ONSIDERABLE EPIDEMIO-
logic data show that a low
concentration of plasma high-
density lipoprotein choles-
terol (HDL-C) is a major risk factor for
coronary heart disease (CHD).1-4 In the
United States, a low HDL-C concen-
tration is the most prevalent lipid ab-
normality in men with known CHD.5,6
Moreover, a low HDL-C level better
distinguishes populations with and
without CHD than does a high level of
low-density lipoprotein cholesterol
(LDL-C).7,8 The VA High-Density Li-
poprotein Intervention Trial (VA-
HIT) was undertaken to test the hy-
pothesis that drug therapy to increase
a low HDL-C level would decrease the
incidence of major CHD events.9
A total of 2531 men with known CHD
and low levels of HDL-C and LDL-C were
recruited for VA-HIT and treated with
either the fibric acid derivative gemfi-
brozil or placebo (FIGURE 1). Treat-
©2001 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013
Context A low plasma level of high-density lipoprotein cholesterol (HDL-C) is a ma-
jor risk factor for coronary heart disease (CHD). A secondary prevention study, the
Veterans Affairs High-Density Lipoprotein Intervention Trial (VA-HIT), demonstrated
that CHD events were significantly reduced during a median follow-up of 5.1 years
by treating patients with the fibric acid derivative gemfibrozil when the predominant
lipid abnormality was low HDL-C.
Objective To determine if the reduction in major CHD events with gemfibrozil in
VA-HIT could be attributed to changes in major plasma lipid levels.
Design Multicenter, randomized, double-blind, placebo-controlled trial conducted
from September 1991 to August 1998.
Setting The Department of Veterans Affairs Cooperative Studies Program, in which
20 VA medical centers were participating sites.
Participants A total of 2531 men with a history of CHD who had low HDL-C levels
(mean, 32 mg/dL [0.83 mmol/L] ) and low low-density lipoprotein cholesterol (LDL-C)
levels (mean, 111 mg/dL [2.88 mmol/L]).
Intervention Participants were randomly assigned to receive gemfibrozil, 1200 mg/d
(n=1264), or matching placebo (n = 1267).
Main Outcome Measure Relation of lipid levels at baseline and averaged during
the first 18 months of gemfibrozil treatment with the combined incidence of nonfatal
myocardial infarction and CHD death.
Results Concentrations of HDL-C were inversely related to CHD events. Multivari-
able Cox proportional hazards analysis showed that CHD events were reduced by 11%
with gemfibrozil for every 5-mg/dL (0.13-mmol/L) increase in HDL-C (P=.02). Events
were reduced even further with gemfibrozil beyond that explained by increases in HDL-C
values, particularly in the second through fourth quintiles of HDL-C values during treat-
ment. During gemfibrozil treatment, only the increase in HDL-C significantly pre-
dicted a lower risk of CHD events; by multivariable analysis, neither triglyceride nor
LDL-C levels at baseline or during the trial predicted CHD events.
Conclusions Concentrations of HDL-C achieved with gemfibrozil treatment predicted
a significant reduction in CHD events in patients with low HDL-C levels. However, the
change in HDL-C levels only partially explained the beneficial effect of gemfibrozil.
JAMA. 2001;285:1585-1591 www.jama.com
ment with gemfibrozil at a dosage of 1200 dence of nonfatal myocardial infarction
mg/d resulted in a 22% reduction in the (MI) and CHD death, during a median
primary end point, the combined inci- follow-up of 5.1 years.9 The relative
Author Affiliations and Financial Disclosures are listed Corresponding Author and Reprints: Sander J. Rob-
at the end of this article. ins, MD, Section of Endocrinology, Nutrition, and Dia-
A listing of the members of the Veterans Affairs High- betes, Evans 204, Boston University School of Medi-
Density Lipoprotein Intervention Trial (VA-HIT) was cine, 88 E Newton St, Boston, MA 02118 (e-mail:
published previously (N Engl J Med. 1999;341:417-418). sjrobins@bu.edu).
(Reprinted) JAMA, March 28, 2001—Vol 285, No. 12 1585
 GEMFIBROZIL, LIPIDS, AND CORONARY EVENTS
trials have also had relatively high con-
Figure 1. Participant Flow Through the
Study centrations of total cholesterol or LDL-C
and, when measured, concentrations of
163 490 Subjects Invited for Screening HDL-C that have not been in a dis-
tinctly low range. In sharp contrast to
24 233 Screened
subjects in previous trials, participants
in VA-HIT were recruited who had low
LDL-C (#140 mg/dL [3.63 mmol/L]) as
21 702 Excluded for Lipid Levels Outside
Desired Range or Other Reasons
well as low HDL-C (#40 mg/dL [1.04
mmol/L]) levels. Participants also had a
broad range of triglyceride values (#300
2531 Randomized mg/dL [3.38 mmol/L]), representative of
triglycerides in 75% to 80% of men with
1267 Assigned to Receive 1264 Assigned to Receive
CHD in the United States.6
Placebo Gemfibrozil Gemfibrozil, like other fibric acid de-
1267 Received Therapy 1264 Received Therapy rivatives, has a wide range of poten-
as Assigned as Assigned
tially favorable effects on lipoprotein me-
tabolism.19 The most consistent plasma
275 Major CHD Events 219 Major CHD Events
lipid changes that result from gemfibro-
zil treatment are an increase in HDL-C
303 Withdrew or Died 307 Withdrew or Died and a decrease in triglycerides, changes
270 Adverse Events 265 Adverse Events
148 Cancer 136 Cancer
that are in great part reciprocally re-
102 Non-CHD Death 105 Non-CHD Death lated and secondary to the activation by
20 Other 24 Other gemfibrozil of the lipolysis of triglycer-
24 Only Vital Status 33 Only Vital Status
Known Known ide-rich lipoproteins. Gemfibrozil may
0 Lost to Follow-up 3 Lost to Follow-up also increase plasma HDL-C by decreas-
ing cholesteryl ester transfer protein–
964 Completed Trial 957 Completed Trial mediated cholesterol exchange from
HDL20,21 and by directly stimulating he-
CHD indicates coronary heart disease. patic HDL synthesis and secretion.22 In
addition to changes in HDL that are
reduction in these CHD events with gem- closely linked to triglycerides, like other
fibrozil was virtually the same as reported fibric acids gemfibrozil will activate per-
for recent secondary prevention trials oxisome proliferator–activated nuclear
with statin therapy in patients with mod- receptors that may result in other favor-
estly high LDL-C concentrations.10,11 able effects on vascular function, which
At baseline in VA-HIT, gemfibrozil and may be largely independent of changes
placebo groups were evenly matched for in HDL concentration.23
clinical characteristics and laboratory val- The present analysis was performed to
ues and had the following mean fasting determine if the reduction in major CHD
lipid values: total cholesterol, 175 mg/dL events in VA-HIT could be correlated
(4.53 mmol/L); triglycerides, 162 mg/dL with the concentrations of lipids at base-
(1.82 mmol/L); LDL-C, 111 mg/dL (2.88 line, during the trial, and the changes in
mmol/L); and HDL-C, 32 mg/dL (0.83 these lipids with gemfibrozil treatment.
mmol/L). As previously reported,9 in this
population treatment with gemfibrozil, METHODS
compared with placebo, produced after The general design and procedures of
1 year an average increase in HDL-C of VA-HIT have been previously re-
6%, a decrease in triglycerides of 31%, ported.24 The trial was conducted from
and no change in LDL-C. September 1991 through August 1998
Other clinical end point trials have in 20 VA medical centers. The pri-
been conducted with drugs that have a mary combined end point, nonfatal MI
prominent effect on HDL-C and triglyc- and CHD death, was adjudicated by an
erides, most particularly fibric acids12-18 independent committee blinded to
or niacin.15,16 However, subjects in these treatment assignment. A central labo-
1586 JAMA, March 28, 2001—Vol 285, No. 12 (Reprinted) ©2001 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013
ratory performed all lipid analyses once
subjects had been randomized. Labo-
ratory personnel were blinded to treat-
ment and participant identification.
This study was approved by the hu-
man rights committee of the Coopera-
tive Studies Program Coordinating Cen-
ter and by each study site’s institutional
review board. Each participant gave
written informed consent.
Lipid Measurement
Before medication was dispensed, lipid
levels were determined twice and av-
eraged to obtain a baseline value. Ma-
jor lipid concentrations (consisting of
HDL-C, triglycerides, and LDL-C) were
determined during the trial at 4, 7, 12,
18, 24, 36, 48, and 60 months. Values
obtained at the 4- through 18-month
visits were averaged and used for trial
results in accord with our original pro-
tocol design that postulated a 2-year lag
in treatment benefit and with the ra-
tionale that these earliest samples were
obtained before an appreciable num-
ber of new CHD events had occurred.
If an end point occurred before 18
months, only values up to the time of
an end point were averaged to avoid a
lipid change that might have been in-
fluenced by a new CHD event.
Lipids were measured after subjects
had fasted for 12 to 14 hours. Blood
samples were prepared using EDTA as
an anticoagulant and plasma was mailed
frozen from individual study sites to the
central laboratory. Total cholesterol,
HDL-C, and triglycerides were deter-
mined by standardized automated en-
zymatic methods25 and LDL-C was cal-
culated by the Friedewald formula.26
Levels of LDL-C were not calculated if
triglycerides were higher than 400 mg/dL
(4.51 mmol/L). HDL-C was isolated by
precipitation with dextran-Mg2+.27 Apo-
lipoprotein (apo) A-I, apoB, HDL2-C, and
HDL3-C concentrations were deter-
mined at baseline and during the trial at
the 12-month follow-up visit.
The mean coefficient of variation for
the measurements of cholesterol, triglyc-
erides, and HDL-C was ,2%, ,3%, and
,4%, respectively. Levels of apoA-I and
apoB were determined by immunotur-

bidometric precipitation methods28,29 and
HDL subfractions were separated by dif-
ferential polyanion precipitation.30 All re-
sults were sent to the VA Cooperative
Studies Coordinating Center (West Ha-
ven, Conn) and entered into a VA-HIT
centralized database.
Statistical Analysis
All analyses that relate plasma lipids and
apolipoproteins at baseline and during
the trial to the development of a new
CHD event were performed according to
the principle of intention-to-treat and us-
ing the combined incidence of nonfatal
MI and CHD death, the primary VA-
HIT outcome measure. The incidence of
CHD events was obtained using the
Kaplan-Meier survival method for ter-
tile divisions of baseline and quintile di-
visions of trial concentrations of HDL-C,
triglycerides, and LDL-C. Relative risks
(RRs) for concentrations and changes in
lipids were calculated from Cox propor-
tional hazards models31 adjusting for
treatment category and the major CHD
risk factors of age, diabetes, hyperten-
sion, smoking, and body mass index.
Compliance with therapy, as assessed by
pill counts,24 did not significantly pre-
dict a primary CHD event and was not
used in risk calculation. Models with
baseline and trial lipid levels were sepa-
rately constructed to assess the signifi-
cance of individual lipids and apolipo-
proteins as univariable predictors of risk
and, for the major plasma lipids (HDL-C,
triglycerides, and LDL-C), as a multi-
variable set of variables to predict risk.
Interaction between treatment and
plasma lipids or apolipoproteins at base-
line or during the trial were tested 1 at a
time using separate Cox models. Rela-
tive risks with 95% confidence inter-
vals (CIs) and corresponding P values are
shown for all Cox results. Quintile com-
parisons were performed using log-
rank tests to compare survival curves.
RESULTS
Lipid data were available for 2521 men
at baseline (99.6% of the cohort) and
for 2375 subjects from at least 1 of the
4 follow-up visits. The distribution at
baseline of plasma HDL-C, triglycer-
©2001 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013
GEMFIBROZIL, LIPIDS, AND CORONARY EVENTS
Table 1. Levels of Lipids and Apolipoproteins at Baseline and During the Trial*
Values, Mean (SD), mg/dL†
No. of Patients
With With
Variable Placebo Gemfibrozil Baseline Placebo Gemfibrozil P Value‡
Total cholesterol 1180 1185 175 (25) 177 (25) 168 (25) ,.001
HDL-C 1180 1185 31.5 (5.3) 31.7 (5.3) 33.4 (5.8) ,.001
LDL-C 1174 1183 111 (22) 113 (23) 113 (22) .71
Triglycerides 1180 1185 151 (68) 156 (70) 101 (54) ,.001
Cholesterol/HDL-C 1180 1185 5.7 (1.2) 5.7 (1.1) 5.2 (1.2) ,.001
HDL2-C 1073 1063 5.3 (2.5) 5.0 (2.2) 4.6 (2.1) ,.001
HDL3-C 1073 1063 26.5 (4.7) 26.7 (5.2) 28.9 (5.9) ,.001
Apolipoprotein B 1074 1064 95.9 (21.3) 93.0 (18.2) 88.3 (18.8) ,.001
Apolipoprotein A-I 1074 1064 106.5 (17.1) 108.9 (16.9) 108.6 (16.9) .93
*HDL-C indicates high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol.
†Triglyceride values and cholesterol/HDL-C values are medians. Data shown reflect the average of 4- to 18-month
values; trial HDL-C subfractions and apolipoproteins are the values at 12 months. To convert HDL-C and its sub-
fractions, total cholesterol, and LDL-C to mmol/L, multiply values by 0.02586. To convert triglycerides to mmol/L,
multiply values by 0.01127.
‡P values for comparison of placebo vs gemfibrozil, using Cox proportional hazards model.
ide, and LDL-C values was similar in placebo while HDL 2 -C, the minor
the placebo and gemfibrozil treatment subfraction, was decreased.
groups.32 In this population, selected to
have a restricted range of lipid values, Relation of CHD Events
mean (SD) values were nearly identi- to Baseline Lipid Values
cal to median values for HDL-C (mean, The relation of CHD events in placebo
32 [5] mg/dL [0.83 {0.13} mmol/L] and and gemfibrozil groups to baseline con-
median, 31 mg/dL [0.80 mmol/L]) and centrations of HDL-C, triglycerides, and
LDL-C (mean, 111 [22] mg/dL [2.88 LDL-C is shown in FIGURE 2. Lipids
{0.57} mmol/L] and median, 112 mg/dL were stratified as tertiles. In the pla-
[2.90 mmol/L]), and closely corre- cebo group, CHD events were higher
sponded to median values for triglyc- than the overall mean event rate for this
erides (mean, 161 [68] mg/dL [1.81 group in the lowest tertile of HDL-C and
{0.77} mmol/L] and median, 151 mg/dL the highest tertiles of triglycerides and
[1.70 mmol/L]). As previously shown,9 LDL-C. With gemfibrozil, there was a
with treatment the mean values of reduction in the RR of a CHD event for
HDL-C, triglycerides, and LDL-C in each tertile of each lipid, except for the
both placebo and gemfibrozil groups re- lowest tertile of LDL-C.
mained nearly constant from 12 months
to the end of the study at 60 months. Relation of CHD Events to Lipid
Values During the Trial
Lipid Changes With Gemfibrozil Subjects in the placebo and gemfibrozil
Baseline and trial lipid and apolipopro- groups were subdivided into quintiles by
tein values are shown in TABLE 1. Most trial values of HDL-C, triglycerides, and
notable was the significant increase in LDL-C to relate the concentrations of
HDL-C, the decrease in triglycerides, theselipidstothe5-yearincidenceofnon-
and the absence of a change in LDL-C fatal MI or CHD death (FIGURE 3). With
concentrations with gemfibrozil placebo,theincidenceofCHDeventswas
therapy. During the study, plasma apoB inversely related to trial HDL-C levels
concentrations were lower with gem- (log-rank test, P=.01 comparing quintile
fibrozil therapy than with placebo. In 1 with 5) but was unrelated to levels of
contrast, apoA-I concentrations were triglycerides (log-rank test, P=.93) and
the same in both groups. With gemfi- LDL-C (log-rank test, P=.49 for compari-
brozil, HDL3-C, the major subfraction son of quintiles 1 and 5).
of HDL-C isolated by polyanion pre- With gemfibrozil, there was a marked
cipitation, was increased compared with reduction in CHD events compared with
(Reprinted) JAMA, March 28, 2001—Vol 285, No. 12 1587
 GEMFIBROZIL, LIPIDS, AND CORONARY EVENTS

placebo for the second through fourth nificant predictors of a new CHD
quintiles of HDL-C levels (P = .02), event, whereas during the trial, only
whereas event rates in the 2 treatment concentrations of HDL-C significantly
groups did not differ at the lowest and predicted a CHD end point. Neither
highest HDL-C quintiles. With respect the change by concentration nor by
to triglycerides, event rates in the gem- percentage in HDL-C, the HDL 2 -C
fibrozil group did not differ across the subfraction, triglycerides, LDL-C, or
lowest 4 quintiles and were decidedly apolipoproteins were significant pre-
lower than for subjects taking placebo. dictors of CHD risk. Although the trial
Only with triglycerides in the highest concentration of HDL3-C did not pre-
quintile range with gemfibrozil was the dict CHD risk (P=.07), the percentage
CHD event rate the same as with pla- change in the HDL3-C subfraction was
cebo. Finally, with gemfibrozil, the in- significantly related to the probability
cidence of CHD events did not differ of a CHD event (P=.01).
across the quintile range of LDL-C lev- TABLE 3 shows the results of multi-
els. For all levels of LDL-C, subjects in variable analyses performed with the 3
the gemfibrozil group had a lower CHD major lipid variables at baseline and dur-
event rate than those taking placebo. ing the trial. These models were con-
structed to exclude apolipoproteins and
RR of CHD Events From Cox HDL subfractions because these closely
Proportional Hazards Models related components of the 3 major lipid
The RR of a new primary CHD event variables could be expected to change the
was determined for individual lipid predictive relationship of a major lipid.
and apolipoprotein variables at base- Adjustment was made for treatment cat-
line and during the trial by Cox mod- egory, which showed no significant in-
els, adjusted for treatment group and teraction with any of the major lipids, and
for major CHD risk factors (TABLE 2). for CHD risk factors. There was no sig-
At baseline, the concentrations of nificant interaction between HDL-C and
HDL-C, triglycerides, the HDL3-C sub- triglycerides either at baseline or dur-
fraction, apoA-I, and apoB were all sig- ing the trial. Although HDL-C, triglyc-
eride, and LDL-C concentrations at base-
line were not significantly related to the
development of a CHD event, the con-
centration of HDL-C achieved with
therapy was strongly related to a reduc-
tion in CHD events. The RR reduction
in CHD end points for a 5-mg/dL (0.13-
mmol/L) increase in HDL-C with gem-
fibrozil was 11%.
We estimated that, together, the con-
centrations of HDL-C, triglycerides, and
LDL-C achieved with therapy made a
relatively small contribution to the over-
all decrease of a primary CHD event
with gemfibrozil. That is, the RR of a
CHD event with treatment as the only
variable in regression analysis was 0.78
(95% CI, 0.66-0.94). The RR of a CHD
event attributed to treatment itself
should be lessened by the inclusion of
other variables in this kind of analysis
that would provide explanation for the
benefit of treatment. In a regression
model that included trial lipid values
with treatment category, the RR of a
CHD event was 0.83 (95% CI, 0.68-
1.02), which would indicate that 77%
([1−0.83]/[1−0.78]) of the benefit of
gemfibrozil was unexplained or that, at
most, the lipid concentrations achieved

Figure 2. Relation of the 5-Year Incidence of CHD Events to Baseline Lipid Values

A HDL-C Placebo B Triglycerides C LDL-C

Gemfibrozil
28 28 28
Overall Event Rate
Placebo
Gemfibrozil
22 22 22
5-Year CHD Event Rate, %

16 16 16

10 10 10

4 4 4

≤29 29.1-33.5 >33.5 ≤124 124.1-180 >180 ≤103.5 103.6-121 >121

Baseline Concentrations, mg/dL
RRR, % 23 25 17 15 22 28 –5 24 38
(95% CI) (42 to –2) (46 to –4) (40 to –14) (37 to –15) (44 to –9) (46 to 4) (33 to –42) (44 to –5) (55 to 15)

Incidence rates were obtained from Kaplan-Meier survival curves for each tertile division of each lipid for placebo or gemfibrozil. For each treatment and at each tertile
an interquartile range is shown for the median event rate. The overall event rate for subjects taking placebo is shown by the upper dashed lines and for gemfibrozil by
the lower dashed lines. The relative risk reduction (RRR) of coronary heart disease (CHD) events with 95% confidence intervals (CIs) with gemfibrozil was calculated
using Cox proportional hazards analysis for each tertile and is shown under the range of values for each tertile. HDL-C indicates high-density lipoprotein cholesterol;
LDL-C, low-density lipoprotein cholesterol. To convert mg/dL to mmol/L, multiply HDL-C and LDL-C values by 0.02586 and triglyceride values by 0.01127.

1588 JAMA, March 28, 2001—Vol 285, No. 12 (Reprinted) ©2001 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

 GEMFIBROZIL, LIPIDS, AND CORONARY EVENTS

with gemfibrozil could account for only HDL-C. Treatment with gemfibrozil re-
23% of the treatment benefit. sulted in a significant reduction in CHD
events during a 5-year period of fol-
COMMENT low-up in conjunction with an in-
VA-HIT was undertaken to determine crease in HDL-C, a decrease in triglyc-
if raising HDL-C concentrations would erides, and no change in LDL-C levels.9
decrease major coronary events in a We now show that in VA-HIT the re-
high-risk group of men with low duction in nonfatal MI and CHD death
was strongly correlated with treat-
ment concentrations of HDL-C but not
Figure 3. Relation of the 5-Year Incidence triglycerides or LDL-C. In multivari-
of CHD Events to HDL-C, Triglyceride, and
LDL-C Values Achieved With Placebo or able analysis, adjusting for the CHD risk
Gemfibrozil factors of diabetes, hypertension, smok-
ing, age, and body mass index, the only
A HDL-C major lipid to predict a significant re-
25
Placebo duction in CHD events was HDL-C.
5-Year CHD Event Rate, %
A notable finding of VA-HIT was that
with multivariable analysis, neither base-
line nor treatment triglyceride levels pre-
dicted CHD events (Table 3). A similar
conclusion was reported in the Hel-
sinki Heart Study,33 which found equally
profound reductions in triglycerides with
gemfibrozil as in VA-HIT (on the order
of 30%) but no independent benefit of
triglyceride reduction. Furthermore, even
as a single lipid variable, VA-HIT triglyc-
eride concentrations during treatment
were not significantly related to the oc-
currence of a coronary event (Table 2).
Contrary to these trial results, we did
find that baseline triglycerides as a single

Gemfibrozil
22

19 Table 2. Plasma Lipids and Apolipoproteins as Predictors of Nonfatal MI and CHD Mortality*
16 Variable (Change) Relative Risk (95% CI) P Value
Baseline
13 HDL-C (5.0 mg/dL) 0.91 (0.83-0.99) .047
Triglycerides (50 mg/dL) 1.07 (1.00-1.15) .045
10
24 28 32 36 40 44 LDL-C (25 mg/dL) 1.07 (0.96-1.19) .22
Apolipoprotein B (10 mg/dL) 1.06 (1.01-1.11) .01
B Triglycerides
25 Apolipoprotein A-I (10 mg/dL) 0.92 (0.86-0.98) .01
5-Year CHD Event Rate, %

HDL2-C (2.0 mg/dL) 0.99 (0.91-1.08) .70

22
HDL3-C (5.0 mg/dL) 0.80 (0.71-0.90) ,.001
19 During treatment
HDL-C (5.0 mg/dL) 0.89 (0.81-0.97) .01
16 Triglycerides (50 mg/dL) 1.05 (0.98-1.14) .16
LDL-C (25 mg/dL) 1.07 (0.96-1.20) .19
13
Apolipoprotein B (10 mg/dL) 1.04 (0.99-1.10) .15
10 Apolipoprotein A-I (10 mg/dL) 0.95 (0.89-1.01) .07
50 95 140 185 230 275
HDL2-C (2.0 mg/dL) 0.99 (0.90-1.09) .84†
C LDL-C HDL3-C (5.0 mg/dL) 0.92 (0.84-1.01) .07
25
*MI indicates myocardial infarction; CHD, coronary heart disease; CI, confidence interval; HDL-C, high-density lipo-
5-Year CHD Event Rate, %

protein cholesterol; and LDL-C, low-density lipoprotein cholesterol. Individual lipid and apolipoprotein relative risk
22 values were adjusted for treatment and CHD risk factors of age, smoking, diabetes, hypertension, and body mass
index. Trial lipid values are the average of values at 4, 7, 12, and 18 months of follow-up or to the time of an index
19 end point; apolipoprotein and HDL-C subfractions are values at the 12-month visit. To convert HDL-C and its sub-
fractions and LDL-C to mmol/L, multiply values by 0.02586. To convert triglycerides to mmol/L, multiply by 0.01127.
†The interaction between treatment group and HDL2-C was significant (P = .02).
16

13
Table 3. Major Plasma Lipids as Multivariable Predictors of Nonfatal MI and CHD Mortality*
10 Variable (Change) Relative Risk (95% CI) P Value
80 95 110 125 140 155
Quintile Trial Concentrations, mg/dL Baseline
HDL-C (5.0 mg/dL) 0.93 (0.85-1.02) .13
Incidence rates were obtained from Kaplan-Meier sur- Triglycerides (50 mg/dL) 1.06 (0.99-1.14) .11
vival curves for each quintile of values. The median LDL-C (25 mg/dL) 1.06 (0.96-1.17) .26
event rates with interquartile ranges are plotted for
During treatment
the quintiles of the average of lipid values at 4- through
HDL-C (5.0 mg/dL) 0.89 (0.81-0.98) .02
18-month follow-up. To obtain near-even numbers
of subjects in each quintile, the placebo and gemfi- Triglycerides (50 mg/dL) 1.03 (0.95-1.11) .48
brozil groups were separately subdivided. The rela- LDL-C (25 mg/dL) 1.09 (0.98-1.21) .13
tion of the incidence of coronary heart disease (CHD)
events to trial values of high-density lipoprotein cho- *MI indicates myocardial infarction; CHD, coronary heart disease; CI, confidence interval; HDL-C, high-density lipo-
protein cholesterol; and LDL-C, low-density lipoprotein cholesterol. In this analysis, all the shown values were in-
lesterol (HDL-C), triglycerides, and low-density lipo- cluded in the multivariable model. Individual lipid relative risk values were adjusted for treatment and CHD risk factors
protein cholesterol (LDL-C) is shown by best-fit curves. of age, smoking, diabetes, hypertension, and body mass index. Trial lipid values are the average of values at 4, 7,
To convert mg/dL to mmol/L, multiply HDL-C and 12, and 18 months of follow-up or to the time of an index end point. To convert HDL-C and LDL-C to mmol/L, mul-
LDL-C values by 0.02586 and triglycerides by 0.01127. tiply values by 0.02586. To convert triglycerides to mmol/L, multiply values by 0.01127.

©2001 American Medical Association. All rights reserved. (Reprinted) JAMA, March 28, 2001—Vol 285, No. 12 1589

Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

 GEMFIBROZIL, LIPIDS, AND CORONARY EVENTS
lipid variable had a significant relation-
ship to the development of CHD events
(Table 2) and that the event rate was
highest at the highest tertile level of tri-
glycerides at baseline (Figure 2). More-
over, at this highest tertile level of base-
line triglycerides, gemfibrozil resulted in
a significant reduction in CHD events
(RR reduction of 28%). This observa-
tion is consistent with the Bezafibrate In-
farction Prevention Study18 that, like VA-
HIT, found a higher event rate and also
a significant benefit of fibrate therapy in
the subgroup of patients with the high-
est baseline values of triglycerides.
In VA-HIT we found that the subfrac-
tion HDL3-C, but not HDL2-C, at base-
line and as a percentage change with
therapy was significantly related to the
development of new CHD events. Fi-
brates selectively increase HDL3,34,35 the
subfraction of HDL that is smaller and
relatively poorer in free cholesterol than
HDL2 and, consequently, more likely to
initiate free cholesterol efflux from pe-
ripheral tissue sites than HDL2 in the
process of reverse cholesterol trans-
port.36,37 Although HDL3-C and total
HDL-C are strongly correlated, 4 angio-
graphic intervention trials have shown
HDL3-C to be a better predictor of coro-
nary lesion progression than total HDL-
C.38-41 Those results coincide with our
finding in VA-HIT that baseline con-
centrations of HDL 3 -C were more
strongly related to a CHD event than to-
tal HDL-C (Table 2).
We designed VA-HIT to exclude pa-
tients with a high concentration of
LDL-C. With gemfibrozil, concentra-
tions of LDL-C were not changed com-
pared with placebo nor were CHD events
significantly related to baseline or trial
LDL-C concentrations. Moreover, con-
centrations of apoB, the component of
LDL that has been found to be strongly
correlated with a reduction in CHD risk
when LDL-C is decreased with therapy,42
had no significant relation to the devel-
opment of a CHD event.
The absence of any discernible ef-
fect on LDL-C concentrations distin-
guishes VA-HIT from previous lipid in-
tervention trials that have demonstrated
a reduction in CHD end points or im-
1590 JAMA, March 28, 2001—Vol 285, No. 12 (Reprinted)
Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013
provement in angiographic end points
with drug therapy. In other trials in
which changes in HDL-C (or a major
subfraction of HDL) could be statisti-
cally shown to contribute to a reduc-
tion in CHD or angiographic events,
subjects have had higher than desir-
able LDL-C (or higher total choles-
terol) at entry and/or reductions in
LDL-C (or total cholesterol) with drug
therapy that clearly played a part in
CHD event reduction. These include the
Helsinki Heart Study,17 the only other
large clinical end point trial that has
shown a significant benefit of increas-
ing HDL-C, AFCAPS/TexCAPS,43 which
showed a significant benefit of increas-
ing apoA-I, and a number of trials in
which drug therapy improved angio-
graphic end points.38,40,41,44-46
Our analyses have several potential
limitations. First, our trial was con-
fined to men with CHD and a lipid pro-
file that excluded those with either a
high LDL-C (.140 mg/dL [3.6 mmol/
L]) or a high level of triglycerides
(.300 mg/dL [3.4 mmol/L]). We have
previously justified our exclusion of
women on the basis of finding rela-
tively few women with CHD and low
HDL-C levels in the VA health care sys-
tem when this trial was begun.24 We
also have justified our exclusion of per-
sons with a high LDL-C level to avoid
the possibility that these individuals
might be treated with another active
drug to lower those levels. Our exclu-
sion of persons with high triglyceride
levels was based on the relative infre-
quency of triglycerides higher than 300
mg/dL in a survey that we conducted
of 8500 men with known CHD,6 in
which only 12.8% had triglycerides
higher than 300 mg/dL and only 4.1%
had triglycerides higher than 300 mg/dL
together with a low level of LDL-C.
We believe that our analyses have
many strengths that prominently in-
clude our exclusive use of the primary
adjudicated end point of this trial, the
combined incidence of nonfatal MI and
CHD death, to define a new CHD event
and our adherence to analyses by in-
tention-to-treat, which preserved our
strict randomization scheme.
©2001 American Medical Association. All rights reserved.
An important practical issue relates
to the extent to which the results of VA-
HIT can be extrapolated to other kinds
of therapy that also have the property
of increasing HDL-C levels. For ex-
ample, would statins, niacin, or cer-
tain lifestyle changes such as weight
loss, which also increase HDL-C, be an-
ticipated to have the same beneficial ef-
fect as gemfibrozil in the clinical con-
text of VA-HIT? The data we have
presented would strongly argue that it
is not possible to assume that based on
only an HDL-C response, a clinical out-
come comparable to VA-HIT could be
achieved with another kind of therapy.
First, we have shown that at the same
trial levels of HDL-C and in a rela-
tively low range of HDL-C, there were
fewer CHD events with gemfibrozil than
with placebo. Second, we have shown
that trial lipid levels as variables in a
multivariable model can explain only
23% of the favorable effect of gemfi-
brozil. Although the results of VA-
HIT clearly show that clinical benefit
is correlated with higher values of
HDL-C, fibrates produce a variety of
other potentially favorable metabolic
changes mediated in part through the
activation of peroxisome proliferator–
activated receptors.23
In summary, VA-HIT, which was
conducted with the fibric acid deriva-
tive gemfibrozil, is the first lipid inter-
vention trial to show that raising HDL-C
concentrations in persons with estab-
lished CHD and both a low HDL-C and
a low LDL-C level will significantly re-
duce the incidence of major coronary
events. We have demonstrated that a re-
duction in new coronary events is at
least partly dependent on the concen-
tration of HDL-C achieved with gem-
fibrozil and that the benefit of this
therapy is independent of changes in the
concentration of triglycerides or LDL-C.
Author Affiliations: Department of Medicine, Bos-
ton University School of Medicine, Boston, Mass (Dr
Robins); VAMC Cooperative Studies Program Coor-
dinating Center, West Haven, Conn (Dr Collins); Sta-
tistics Collaborative, Washington, DC (Dr Wittes); Lipid
Research Laboratory, Tufts University School of Medi-
cine, Boston (Dr Schaefer and Ms McNamara); and
Departments of Medicine, Veterans Affairs Medical
Centers, Washington, DC (Dr Papademetriou), Fresno,
Calif (Dr Deedwania), Long Beach, Calif (Dr Kashyap),

Los Angeles, Calif (Dr Hershman), Cincinnati, Ohio (Dr
Wexler), and Minneapolis, Minn (Dr Rubins).
Financial Disclosures: Dr Robins serves as a consult-
ant to Fournier Pharma; Dr Schaefer serves on the
speaker’s bureau and as a consultant to Pfizer, is a con-
sultant to AstraZeneca, B. Braun of America, and
Schering-Plough, serves on the speaker’s bureau for
Kos Pharmaceuticals Inc, and has received grants from
Parke-Davis, Kos, and Pfizer; Dr Wittes serves as a con-
sultant to the Department of Veterans Affairs; Ms
McNamara serves or has served as a consultant to
Sigma Diagnostics, Otsuka America Pharmaceutical,
Glaxo Wellcome, and Schering-Plough; and Dr Kashyap
has received grants from Parke-Davis and is on the
speaker’s bureau for Kos. Drs Robins, Collins, and Ru-
bins have a US patent pending pertaining to the new
use of a fibrate in the treatment of low HDL-C levels
for the prevention of atherothrombotic events.
Author Contributions: Dr Robins, as coprincipal inves-
tigator of this study, had full access to all the data in
this study and takes full responsibility for the integrity
of the data and the accuracy of the data analysis.
Study concept and design: Robins, Collins, Wittes,
Schaefer, Rubins.
Acquisition of data: Robins, Collins, Papademetriou,
Deedwania, Schaefer, McNamara, Kashyap, Wexler.
Analysis and interpretation of data: Robins, Collins,
Wittes, Papademetriou, Deedwania, McNamara,
Kashyap, Hershman, Rubins.
Drafting of the manuscript: Robins, Collins, Rubins.
Critical revision of the manuscript for important in-
tellectual content: Robins, Collins, Wittes, Deedwa-
nia, Schaefer, McNamara, Kashyap, Hershman, Wex-
ler, Rubins.
Statistical expertise: Collins, Wittes.
Obtained funding: Robins, Rubins.
Administrative, technical, or material support: Rob-
ins, Collins, Papademetriou, Deedwania, Schaefer,
McNamara, Wexler, Rubins.
Study supervision: Robins, Collins, Rubins.
Funding/Support: The VA-HIT was supported by the
Cooperative Studies Program of the Department of
Veterans Affairs Office of Research and Develop-
ment and by a supplemental grant from Parke-Davis
Pharmaceuticals. Parke-Davis supplied the gemfibro-
zil and matching placebo for this study.
REFERENCES
1. Miller GJ, Miller NE. Plasma-high-density-
lipoprotein concentration and development of is-
chaemic heart-disease. Lancet. 1975;1:16-25.
2. Gordon DJ, Probstfield JL, Garrison RJ, et al. High-
density lipoprotein cholesterol and cardiovascular dis-
ease. Circulation. 1989;79:8-15.
3. Stampfer MJ, Sacks FM, Salvini S, et al. A prospec-
tive study of cholesterol, apoproteins and the risk of myo-
cardial infarction. N Engl J Med. 1991;325:373-381.
4. Goldbourt U, Yaari S, Medalie JH. Isolated low HDL
cholesterol as a risk factor for coronary heart disease
mortality. Arterioscler Thromb Vasc Biol. 1997;17:
107-113.
5. Kannel WB. Range of serum cholesterol values in
the population developing coronary artery disease. Am
J Cardiol. 1995;76:69C-77C.
6. Rubins HB, Robins SJ, Collins D, et al. Distribution
of lipids in 8,500 men with coronary artery disease.
Am J Cardiol. 1995;75:1196-1201.
7. Genest J Jr, McNamara JR, Ordovas JM, et al. Li-
poprotein cholesterol, apolipoprotein A-I and B and
lipoprotein (a) abnormalities in men with premature
coronary artery disease. J Am Coll Cardiol. 1992;19:
792-802.
8. Lamarche B, Despres J-P, Moorjani S, et al. Triglyc-
erides and HDL-cholesterol as risk factors for ischemic
heart disease. Atherosclerosis. 1996;119:235-245.
9. Rubins HB, Robins SJ, Collins D, et al. Gemfibrozil
for the secondary prevention of coronary heart dis-
©2001 American Medical Association. All rights reserved.
DownloadedViewFrom:
publicationhttp://jama.jamanetwork.com/
stats on 02/25/2013
GEMFIBROZIL, LIPIDS, AND CORONARY EVENTS
ease in men with low levels of high-density lipopro-
tein cholesterol. N Engl J Med. 1999;341:410-418.
10. Sacks FM, Pfeffer MA, Moye LA, et al. The ef-
fect of pravastatin on coronary events after myocar-
dial infarction in patients with average cholesterol lev-
els. N Engl J Med. 1996;335:1001-1009.
11. The Long-Term Intervention with Pravastatin in
Ischaemic Disease (LIPID) Study Group. Prevention of
cardiovascular events and death with pravastatin in
patients with coronary heart disease and a broad range
of initial cholesterol levels. N Engl J Med. 1998;339:
1349-1357.
12. Group of Physicians of the Newcastle upon Tyne
Region. Trial of clofibrate in the treatment of is-
chaemic heart disease. BMJ. 1971;4:767-775.
13. Research Committee of the Scottish Society of Phy-
sicians. Ischaemic heart disease. BMJ. 1971;4:775-
784.
14. A co-operative trial in the primary prevention of
ischaemic heart disease using clofibrate. Br Heart J.
1978;40:1069-1118.
15. Coronary drug Project Research Group. Clofi-
brate and niacin in coronary heart disease. JAMA. 1975;
231:360-381.
16. Carlson LA, Rosenhamer G. Reduction of mor-
tality in the Stockholm Ischaemic Heart Disease Sec-
ondary Prevention Study by combined treatment with
clofibrate and nicotinic acid. Acta Med Scand. 1988;
223:405-418.
17. Frick MH, Elo O, Haapa K, et al. Helsinki Heart
Study: primary-prevention trial with gemfibrozil in
middle-aged men with dyslipidemia. N Engl J Med.
1987;317:1237-1245.
18. The BIP Study Group. Secondary prevention by
raising HDL-cholesterol and reducing triglycerides in
patients with coronary artery disease: the Bezafi-
brate Infarction Prevention (BIP) Study. Circulation.
2000;102:21-27.
19. Staels B, Dallongeville J, Auwerx J, et al. Mecha-
nism of action of fibrates on lipid and lipoprotein me-
tabolism. Circulation. 1998;98:2088-2093.
20. Ponsin G, Girardot G, Berthezene F. Mechanism
of the gemfibrozil-induced decrease in the transfer of
cholesterol esters from high density lipoproteins to very
low and low density lipoproteins. Biochem Med Metab
Biol. 1994;52:58-64.
21. Franceschini G, Lovati MR, Manzoni C, et al. Effect
of gemfibrozil treatment in hypercholesterolemia on low
density lipoprotein (LDL) subclass distribution and LDL-
cell interaction. Atherosclerosis. 1995;114:61-71.
22. Saku K, Gartside PS, Hynd BA, Kashyap ML.
Mechanism of action of gemfibrozil on lipoprotein me-
tabolism. J Clin Invest. 1985;75:1702-1712.
23. Fruchart J-C, Duriez P, Staels B. Peroxisome pro-
liferator-activated receptor-alpha activators regulate
genes governing lipoprotein metabolism, vascular in-
flammation and atherosclerosis. Curr Opin Lipidol.
1999;10:245-257.
24. Rubins HB, Robins SJ, Iwane MK, et al. Rationale
and design of the Department of Veterans Affairs High
Density Lipoprotein Cholesterol Intervention Trial (HIT)
for secondary prevention of coronary artery disease
in men with low high-density lipoprotein cholesterol
and desirable low-density lipoprotein cholesterol. Am
J Cardiol. 1993;71:45-52.
25. McNamara JR, Schaefer EJ. Automated enzy-
matic standardized lipid analyses for plasma and li-
poprotein fractions. Clin Chim Acta. 1987;166:1-8.
26. Friedewald WT, Levy RI, Frederickson DS. Esti-
mation of the concentration of low-density lipopro-
tein cholesterol in plasma, without use of the prepara-
tive ultracentrifuge. Clin Chem. 1972;18:499-502.
27. McNamara JR, Huang C, Massov T, et al. Modi-
fication of the dextran-Mg2+ HDL cholesterol pre-
cipitation method for use with previously frozen plasma.
Clin Chem. 1994;40:233-239.
28. Contois JH, McNamara JR, Lammi-Keefe CJ, et al.
Reference intervals for plasma apolipoprotein A-I as de-
(Reprinted) JAMA, March 28, 2001—Vol 285, No. 12
termined with a commercially available immunotur-
bidometric assay. Clin Chem. 1996;42:507-514.
29. Contois JH, McNamara JR, Lammi-Keefe CJ, et
al. Reference intervals for plasma apolipoprotein B as
determined with a commercially available immuno-
turbidometric assay. Clin Chem. 1996;42:515-523.
30. Nguyen T, Warnick GR. Improved methods for
separation of total HDL and subclasses [abstract]. Clin
Chem. 1989;35:1086.
31. Cox DR. Regression models and life-tables. J R
Stat Soc B. 1972;34:187-202.
32. Rubins HB, Robins SJ, Collins D, for the Veterans
Affairs Cooperative Studies Program High-Density Li-
poprotein Intervention Trial Study Group. Baseline char-
acteristics of normocholesterolemic men with coro-
nary artery disease and low levels of high-density
lipoprotein cholesterol. Am J Cardiol. 1996;78:572-
575.
33. Manninen V, Elo O, Frick H, et al. Lipid alter-
ations and decline in the incidence of coronary heart
disease in the Helsinki Heart Study. JAMA. 1988;260:
641-651.
34. Sorisky A, Ooi TC, Simo IE, et al. Change in com-
position of high density lipoprotein during gemfibro-
zil therapy. Atherosclerosis. 1987;67:181-189.
35. Manttari M, Koskinen P, Manninen V, et al. Effect
of gemfibrozil on the concentration and composition of
serum lipoproteins. Atherosclerosis. 1990;81:11-17.
36. Fielding CJ. Early events in the transfer of cell-
derived cholesterol to human plasma. In: Miller NE, ed.
High Density Lipoprotein and Atherosclerosis II. Am-
sterdam, the Netherlands: Elsevier; 1989:149-158.
37. Castro Cabezas M, van Heusden GPH, de Bruin
TWA, et al. Reverse cholesterol transport: relation-
ship between free cholesterol uptake and HDL3 in nor-
molipidaemic and hyperlipidaemic subjects. Eur J Clin
Invest. 1993;23:122-129.
38. Watts GF, Mandalia S, Brunt JNH, et al. Indepen-
dent associations between plasma lipoprotein sub-
fraction levels and the course of coronary artery dis-
ease in the St Thomas’ Atherosclerosis Regression Study
(STARS). Metabolism. 1993;42:1461-1467.
39. Syvanne M, Nieminen MS, Frick MH, et al. Asso-
ciations between lipoproteins and the progression of
coronary and vein-graft atherosclerosis in a controlled
trial with gemfibrozil in men with low baseline levels of
HDL cholesterol. Circulation. 1998;98:1993-1999.
40. Ballantyne CM, Herd A, Ferlic LL, et al. Influence
of low HDL on progression of coronary artery dis-
ease and response to fluvastatin therapy. Circula-
tion. 1999;99:736-743.
41. Mack WJ, Krauss RM, Hodis HN. Lipoprotein sub-
classes in the Monitored Atherosclerosis Regression
Study (MARS): treatment effects and relation to coro-
nary angiographic progression. Thromb Vasc Biol.
1996;16:697-704.
42. Sehayek E, Eisenberg S. The role of native apo-
lipoprotein B-containing lipoproteins in atherosclero-
sis. Curr Opin Lipidol. 1994;5:350-353.
43. Gotto AM, Whitney E, Stein EA, et al. Relation
between baseline and on-treatment lipid parameters
and first acute major coronary events in AFCAPS/
TexCAPS. Circulation. 2000;101:477-484.
44. Blankenhorn DH, Alaupovic P, Wickham E, et al.
Prediction of angiographic change in native human
coronary arteries and aortocoronary bypass grafts. Cir-
culation. 1990;81:470-476.
45. Brown G, Albers JJ, Fisher LD, et al. Regression
of coronary artery disease as a result of intensive lipid-
lowering therapy in men with high levels of apolipo-
protein B. N Engl J Med. 1990;323:1289-1298.
46. Ruotolo G, Ericson CG, Tettamanti C, et al. Treat-
ment effects on serum lipoproteins, lipids, apolipo-
proteins and low density lipoprotein particle size and
relationships of lipoprotein variables to progression of
coronary artery disease in the Bezafibrate Coronary
Atherosclerosis Intervention Trial (BECAIT). J Am Coll
Cardiol. 1998;32:1648-1656.
1591