European Heart Journal Supplements (2003) 5 (Supplement D), D9—D16
Apolipoproteins: the new prognostic indicator?
C.J. Packard
University of Glasgow, Glasgow, Scotland, U.K.
KEYWORDS
Apolipoprotein;
apolipoprotein B:
apolipoprotein A-I
ratio;
coronary heart disease
prevention;
high-density
lipoprotein
cholesterol;
low-density lipoprotein
cholesterol;
statins
Accumulating evidence indicates that apolipoprotein (apo)B, apoA-I and,
particularly, the apoB:apoA-I ratio are strong predictors of risk for coronary heart
disease and of benefit in coronary heart disease prevention in lipid-lowering trials.
For example, the Apolipoprotein related MOrtality RISk study showed that the
apoB:apoA-I ratio was strongly related to increased risk for fatal myocardial
infarction, and that plasma apoB and apoA-I levels and apoB:apoA-I ratio were
better predictors of risk than levels of total cholesterol or triglycerides. Analysis of
lipid variables in the Air Force/Texas Coronary Atherosclerosis Prevention Study
population showed that baseline high-density lipoprotein cholesterol (HDL-C),
apoA-I, apoB, low-density lipoprotein cholesterol (LDL-C):HDL-C ratio, total
cholesterol:HDL-C ratio and apoB:apoA-I ratio significantly predicted first major
acute event, whereas LDL-C and total cholesterol did not. For 1-year on-treatment
values, only apoA-I, apoB and apoB:apoA-I ratio were significant predictors of a
future coronary event. New statins can reduce substantially the apoB:apoA-I ratio.
This is particularly evident for rosuvastatin (the latest addition to the class), which
both decreases apoB and increases apoA-I. A number of conceptual and practical
hurdles need to be overcome before apolipoprotein measurements can be
incorporated into routine risk assessment.
© 2003 The European Society of Cardiology. Published by Elsevier Science Ltd. All
rights reserved
Introduction
Apolipoproteins bind with phospholipids to form a
surface monolayer in all mature lipoprotein
particles. Apolipoprotein (apo)B is found in all
atherogenic lipoproteins, including low-density
lipoprotein (LDL), small dense LDL, very-lowdensity lipoprotein (containing cholesterol and
triglycerides), remnant particles, intermediatedensity lipoprotein and lipoprotein(a). ApoA-I and
apoA-II are the major apolipoprotein constituents
of the antiatherogenic high-density lipoprotein
(HDL) and its subfractions. Whereas early
Correspondence: Chris Packard, DSc, FRCPath, FRCP(Gla),
Professor of Pathological Biochemistry, Glasgow Royal
Infirmary, University Block, 10 Alexandra Parade, Glasgow G31
2ER, U.K.
epidemiological studies of risk factors for coronary
heart disease (CHD) focused on serum cholesterol
and identified high levels of LDL-cholesterol
(LDL-C) and low levels of HDL-cholesterol (HDL-C)
as important prognostic indicators for future CHD
events, more recent studies have suggested that
apoB and apoA-I may in fact be better markers of
CHD risk. Technical barriers to widespread use
have been surmounted, increasing the practical
utility of apolipoprotein assays, and they are now
widely available and satisfactorily standardized.1,2
The assays have excellent precision, usually better
than that for current assays for LDL-C and HDL-C.
Clinical applicability of the measures is also likely
to be relatively straightforward. As an initial step,
target levels could be set by extrapolation from
LDL-C and HDL-C targets and on the basis of
epidemiological data on apolipoprotein levels.
01520-765X/03/0D0009 + 08 $35.00/0 © 2003 The European Society of Cardiology, Published by Elsevier Science Ltd. All rights reserved.
D10
C.J. Packard
Apolipoproteins as predictors of risk
A considerable amount of data has accumulated
that indicates that plasma apoB is a strong
predictor of CHD risk. Numerous case—control
studies3—5 have reported comparable LDL-C levels
but significantly elevated apoB levels in patients
with coronary disease as compared with
individuals without disease. Indeed, the near
normal levels of LDL-C in many CHD patients was a
primary motivation in designing the Cholesterol
and Recurrent Events (CARE) trial, which showed a
significant effect of statin therapy in preventing
CHD in individuals with average cholesterol
levels.6 By way of example, in a recently reported
case—control study,5 we found that LDL-C levels
(in male patients with angiographic coronary
artery disease or post-myocardial infarction
patients) did not differ significantly from those in
healthy age-matched control individuals. Levels of
apoB were significantly higher, however, and
HDL-C levels were significantly lower in those with
disease (Table 1).
Findings supporting the predictive value of apoB
have also been reported in prospective studies.
The Quebec Cardiovascular Study7 evaluated
baseline lipid and apolipoprotein levels in 2155
men who were followed for 5 years. It was found
that each standard deviation increase in baseline
apoB level (i.e. an increment equal to the
population standard deviation) was associated
with a significant increase in relative risk (RR) for
CHD of 1.44 after controlling for age, smoking,
diabetes, and systolic blood pressure. Multiple
proportional hazards analysis showed that this
significant effect was maintained, after
adjustment for levels of triglycerides (RR for
apoB = 1.48), HDL-C (RR 1.40), and high or low
total cholesterol:HDL-C ratio (RR 1.29). Stepwise
proportional hazards analysis confirmed the
importance of both the total cholesterol:HDL-C
ratio and apoB in the prediction of disease.
Analysis of the interaction of these two variables
by categorizing individuals into those with high or
low apoB levels and high or low total
cholesterol:HDL-C ratio (median apoB 116 mg/dl
and total cholesterol:HDL-C ratio 5.65) showed
that the highest ischaemic heart disease rate was
among those with high apoB and high total
cholesterol:HDL-C ratio values (significant 2.6-fold
increase in rate as compared with low apoB/low
total cholesterol:HDL-C).
Similar findings have been made with regard to
the predictive ability of apo A-I.8,9 In one case—
control study of 184 patients with angiographic
coronary disease and 191 age- and sex-matched
control individuals,9 HDL-C (39 vs 43 mg/dl,
respectively), apoA-I (117 vs 137 mg/dl), HDL
particles containing only apoA-I (LpA-I; 37 vs
45 mg/dl) and HDL particles containing both
apoA-I and apoA-II (LpA-I:A-II; 80 vs 92 mg/dl)
were all significantly lower (P<0.001 for all except
HDL-C, which was P=0.03) in patients with disease.
Total cholesterol (201 vs 205 mg/dl) and
triglyceride (142 vs 133 mg/dl) levels did not
differ significantly between the two groups.
Among the HDL measures, total plasma apoA-I was
the single best predictor of disease, with neither
LpA-I nor LpA-I:A-II adding to predictive strength.
The recently reported Prospective Epidemiological Study of Myocardial Infarction (PRIME)
study10 assessed the predictive ability of HDL
constituents in more than 9000 men without CHD
who were followed for CHD events over 5 years.
CHD events occurred in 289 men during follow-up,
and HDL-C, apoA-I, LpA-I and LpA-I:A-II were all
significantly lower in those with events than in
those without. Division of the entire cohort into
quintiles according to the level of each HDL
variable showed that there was a significant linear
increase (P<0.0001) in RR across quintiles for all
variables. However, logistic regression analysis
Table 1 Lipid variables in patients with coronary artery disease or post-myocardial infarction patients, and in matched
control individuals
Variable
Cholesterol (mmol/l)
Triglycerides (mmol/l)
VLDL-C (mmol/l)
LDL-C (mmol/l)
HDL-C (mmol/l)
ApoA-I (g/l)
ApoB (g/l)
Control (n=52)
5.61
1.58
0.725
3.80
1.13
1.23
1.08
CAD (n=43)
5.92
2.18
1.00
3.96
0.99*
1.20
1.25*
PMI (n=43)
5.46
1.88
0.85
3.65
1.00*
1.17
1.20
P
0.084
0.053
0.052
0.292
0.031
0.710
0.042
Values are means; P values from analysis of variance. *P<0.05, by Fisher’s pairwise comparison. apo=apolipoprotein;
CAD=coronary artery disease; HDL-C=high-density lipoprotein cholesterol; LDL-C=low=density lipoprotein cholesterol; PMI= postmyocardial infarction; VLDL-C=very-low-density lipoprotein cholesterol. (Adapted with permission.5)
Apolipoproteins: the new prognostic indicator?
D11
5.5 years; 864 men and 395 women had fatal
myocardial infarction, with risk being 2.5-fold
greater in men. Univariate analysis showed that
risk increased from lowest to highest quartile of
plasma apoB by approximately 2.7-fold in both
men and women, and by about 3.6-fold in
individuals aged younger than 70 years. The risk
ratio for myocardial infarction decreased to 0.5 or
less in the highest compared with the lowest
quartile of apoA-I for both men and women.
Figure 1 shows apoB and apoA-I quartiles in
individuals younger than 70 years adjusted for age
and total cholesterol and triglycerides. There was
a stepwise increase in risk ratio, with both males
and females in the highest apoB/lowest apoA-I
quartiles having the greatest risk. There were
similar findings in individuals aged 70 years or
older, in whom cholesterol loses power as a risk
marker. Overall, the greatest increase in RR on
univariate analysis was observed for the
apoB:apoA-I ratio (risk ratios of 4.8 for men and
4.0 for women; Fig. 2). In stepwise regression
models, apoB, apoA-I or both, or the apoB:apoA-I
ratio were confirmed as stronger predictors of risk
than total cholesterol or triglyceride levels.
Receiver operating characteristics analysis showed
that apoB had a higher sensitivity and specificity
than LDL-C, especially in men and women whose
LDL-C was below the population median (Fig. 3).
Testing for interaction between LDL-C and apoB
indicated that both variables contribute information on risk, with the nature of the interaction
indicating that apoB is a better predictor than
LDL-C at low LDL-C or apoB concentrations. The
reason for the superior predictive ability of apoB
may be that the apolipoprotein serves as a marker
for the number of atherogenic particles because
one apoB peptide is present on each particle.
Thus, whereas LDL-C may be a good marker of the
Table 2 Multivariate comparison of incident coronary
heart disease patients and individuals without coronary
artery disease by logistic regression
Standardized
parameter (β)
Model 1
HDL-C
ApoA-I
Model 2
HDL-C
ApoA-I
LpA-I
Model 3
HDL-C
ApoA-I
LpA-I:apoA-I
Wald χ2
P
+0.187
—0.510
2.7
21.7
NS
0.0001
+0.185
—0.511
+0.015
2.5
19.4
0.0
NS
0.0001
NS
+0.185
—0.510
—0.002
2.5
20.9
0.0
NS
0.0001
NS
Data from the Prospective Epidemiological Study of
Myocardial Infarction (PRIME) study. Model 1: high-density
lipoprotein cholesterol (HDL-C) and apolipoprotein (apo)A-I
included after adjustment for lipid (LDL-C and triglycerides)
and non-lipid parameters (study center, diabetes, smoking,
high blood pressure). Model 2: HDL-C, apoA-I and HDL
particles containing only apoA-I (LpA-I) were included. Model
3: HDL-C, apoA-I and LpA-I:apoA-I ratio were included.
NS=non-significant. (Reproduced with permission.10)
showed that apoA-I was the strongest predictor
(Table 2). Neither measurements of LpA-I or those
of LPA-I:A-II added to predictive ability. Such
findings suggest that the best predictor of risk
among the HDL parameters is the total amount of
apoA-I rather than the amount found in any
particular form of HDL.
Perhaps the strongest evidence of the
predictive power of the apolipoproteins comes
from the recently reported Apolipoprotein related
MOrtality RISk (AMORIS) study,11 in which
apolipoproteins and other lipid measures were
related to fatal myocardial infarction in 175,553
individuals. This cohort of 98,722 men and 76,831
women was followed up for approximately
Men
Women
(<70 years of age)
4.0
5
4.1
4
2.3
3
2.3
2
1
1.4
0.8
1.7
1.8
1.6
1.1
1.1
1.7
1.2
1.42
Apo A-I (g/l)
1.65
7
6
5
4
3
1.39
2
l)
(g/
oB
p
A
1
0
1.10
0.90
1.3
1.80
1.0
0
1.12
6.3
8
3.1
Risk ratio
Risk ratio
10
9
3.6
6.1
5.5
4.8 3.9
3.1
2.2
2.3
2.3 1.7
0.0
2.8
2.4
1.43
1.5
1.84
1.7
1.72
1.29
1.0
l)
(g/
oB
p
A
1.06
0.80
1.2
1.9
Apo A-I (g/l)
Fig. 1 Relative risk ratios for fatal myocardial infarction by quartiles of apolipoprotein (apo)B and apoA-I in subjects in the
Apolipoprotein related MOrtality RISk (AMORIS) study aged under 70 years. (Reproduced with permission.11)
D12
C.J. Packard
Men
Women
‡
‡
4
4
‡
3
3
*
†
2
2
1
1
0
0
0.64
0.86
1.04
0.52
1.39
0.70
1.87
1.21
Apolipoprotein B:Apolipoprotein A-I
*P<0.05
†P<0.001
‡P<0.0001
Fig. 2 Age-adjusted relative risk ratios by apolipoprotein (apo)B:apoA-I ratio quartiles in men and women in the Apolipoprotein
related MOrtality RISk (AMORIS) study (triglycerides <4.5 mmol/l). (Reproduced with permission.11)
1.0 Men (n=263)
0.8
Apo B
LDL-C 3.2 mmol/l,
Triglyceride 1.3 mmol/l
LDL-C
Apo B 0.8 g/l
0.6
Area
0.60
0.65
Sensitivity
0.4
0.2
P<0.0001
0
1.0 Women (n=83)
0.8
LDL-C 3.2 mmol/l,
Triglyceride 3.3 mmol/l
LDL-C
Apo B
Apo B 1.2 g/l
0.6
Area
0.60
0.69
0.4
0.2
P<0.0001
0
0
0.2
0.4
0.6
0.8
1.0
Specificity
Fig. 3 Receiver-operating characteristics curves for low-density lipoprotein cholesterol (LDL-C) and apolipoprotein (apo)B in men
and women with LDL-C levels below the population median in the Apolipoprotein related MOrtality RISk (AMORIS) study. Also shown
in diagrammatic form is the concept that in hypertriglyceridemic subjects apoB, rather than LDL-C, is a truer index of the number
of atherogenic particles in the circulation. (Reproduced with permission.11)
number of atherogenic particles in a normotriglyceridaemic individual, in a hypertriglyceridaemic individual LDL particles have reduced
cholesterol content (i.e. LDL is small and dense)
and so more atherogenic particles will be present
in the circulation than are revealed by measurement of LDL-C. The number of atherogenic
particles in these circumstances will be more
accurately reflected by the plasma apoB level.
Despite the strength of the accumulating
evidence supporting predictive power of the
apolipoproteins, it needs to be noted that not all
studies have found apoB and apoA-I to be stronger
predictors of risk than conventional lipid
measures. For example, in the Quebec Cardiovascular Study,7 adjustment for other lipid
measures eliminated the association between
apoA-I and disease risk. Likewise, the large
Atherosclerosis Risk in Communities (ARIC) study12
followed 12,339 individuals without CHD at
baseline for 10 years, with 725 coronary events
occurring during follow-up. LDL-C, HDL-C and
triglyceride levels (in women) were independent
predictors of risk. When variables were assessed
by risk decile, the addition of apoB and apoA-I to
a model consisting of age, race, LDL-C, HDL-C and
triglycerides produced no improvement in
prediction.
Apolipoproteins: the new prognostic indicator?
LDL-C
7
6
5
4
3
2
1
0
2.0 2.4
2.8
3.2
Apo B
9
8
3.6 4.0
4.4
4.8
5.2
Estimated % of patients with endpoint
Estimated % of patients with endpoint
9
D13
8
7
6
5
4
3
2
1
0
0.7
0.9
Estimated % of patients with endpoint
9
1.1
1.3
1.5
Apo B value at year 1 (g/l)
LDL-C value at year 1 (mmol/l)
Apo B:Apo A-I
8
7
6
5
4
3
2
1
0
0.5
0.7
0.9
1.1
1.3
Apo B:Apo A-I value at year 1
Fig. 4 Logistic regression models of the relationship between first acute major coronary event and 1-year on-treatment values
for low-density lipoprotein cholesterol (LDL-C), apolipoprotein (apo)B and apoB:apoA-I ratio, adjusted for age, sex, marital status,
hypertension, smoking and family history, in the Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS) population. Solid line indicates lovastatin treatment; dashed line indicates placebo. (Reproduced with permission.13)
Apolipoproteins as predictors of
treatment benefit
Apolipoproteins have also been found to be strong
predictors of CHD prevention benefit in lipidlowering trials. The Air Force/Texas Coronary
Atherosclerosis Prevention Study (AFCAPS/
TexCAPS)13 was a primary prevention study of
lovastatin conducted in 6605 patients with
average total and LDL-C and below average HDL-C
levels. Lovastatin treatment was associated with a
37% reduction in risk for first major acute coronary
event in the setting of a 25% reduction in LDL-C
and a 6% increase in HDL-C. In analyses of baseline
and on-treatment (1-year) lipid levels, a Cox
backward stepwise regression model was used to
identify predictors of outcome, and logistic
regression models were used to examine the
relationship between lipid variables and CHD
event risk. A model without apolipoprotein data
showed that, among baseline lipid factors, LDL-C
(P=0.029) and HDL-C (P=0.010) levels were
significant predictors of risk. When baseline
apolipoproteins were included, the apoB/apoA-I
ratio (P<0.001) replaced LDL-C and HDL-C in the
model, with the latter two factors no longer being
significant; each 0.25 increment in baseline ratio
was associated with a 36% incremental risk for a
CHD event. For 1-year on-treatment values, only
apoA-I (P=0.013), apoB (P<0.001) and apoB:apoA-I
ratio (P<0.001) values were significant predictors
of an event. The ability of apoB and, particularly,
the apoB:apoA-I ratio to predict risk is illustrated
by the logistic regression lines shown for ontreatment values in Fig. 4. The separation and
difference in slope between lovastatin treatment
and placebo curves for LDL-C values indicate an
absence of a continuous relationship between
values and risk in the two study groups (e.g. there
are differences in risk levels associated with
identical LDL-C values in the two groups during
treatment). The regression lines for apoB for the
two groups are nearly overlapping and continuous,
indicating a better absolute relationship between
apoB levels and risk (i.e. particular values are
associated with nearly identical risk, irrespective
of active or placebo treatment, and the gradient
of risk is nearly continuous from lower values [in
the lovastatin group] to higher values [in the
placebo group]). The curves for apoB:apoA-I ratio
are virtually superimposed, indicating a
continuous gradient of risk according to ratio value
from low to high.
Analysis of the predictive ability of baseline and
on-treatment lipid levels was also performed in
the secondary prevention Long-term Intervention
with Pravastatin in Ischemic Disease (LIPID) trial
population.14 In that trial, pravastatin treatment
D14
C.J. Packard
Rosuvastatin 10 mg
n=45
10
Rosuvastatin 40 mg
n=44
+6
+5
% Change from baseline
0
–10
Apo A-1
–20
Apo B
–30
–40
–39
–47
–50
Fig. 5 Effects of rosuvastatin 10 mg and 40 mg on apolipoprotein (apo)B and apoA-I levels at 6 weeks in hypercholesterolaemic
patients (low-density lipoprotein cholesterol 160 to <250 mg/dl, triglycerides <400 mg/dl at baseline). (Data from Schneck et al.22)
was associated with a 24% reduction in risk for
fatal CHD or non-fatal myocardial infarction
among 9014 hypercholesterolaemic patients with
CHD. Adjusted analysis of baseline lipids showed
significant associations of levels of total
cholesterol, LDL-C, HDL-C, apoA-I, apoB and total
cholesterol:HDL-C with risk in the placebo group
(n=4502). Adjusted analysis of lipid values at
1 year on-treatment showed significant associations for total cholesterol, LDL-C, triglycerides,
apoA-I, apoB and total cholesterol:HDL-C ratio in
the pravastatin group (n=4386). Analysis of the
proportion of treatment effect (PTE; the total
treatment effect is the 24% reduction in risk)
explained by on-treatment lipid levels showed
that apoB was the most important single measure
accounting for the PTE. On-treatment apoB levels
accounted for 67% of the treatment effect. This
matched the PTE explained by the combination of
total cholesterol and HDL-C, and was greater than
the PTE for levels of LDL-C (52%), total
cholesterol (48%), HDL-C (11%), apoA-I (11%) and
triglycerides (9%).
Effects of statins on apolipoproteins
Newer representatives of the statin class,
including atorvastatin and rosuvastatin, produce
remarkably large decreases in LDL-C levels. In
light of the clinical trial findings noted above, it is
of interest to determine how these agents affect
apoA-I and apoB levels and the apoB:apoA-I ratio.
Rosuvastatin has been reported to reduce LDL-C
significantly more than other statins and to
increase HDL-C to a greater degree in individual
comparative trials and in pooled analyses of
clinical trial data.15—21
In a 6-week trial conducted in hypercholesterolaemic patients,22 rosuvastatin 10 mg
and 40 mg reduced apoB levels by 39% and 47%,
respectively, and increased apoA-I levels by 5% and
6%, respectively (Fig. 5). When the apoB:apoA-I
ratios for rosuvastatin were compared across the
dose range with those for atorvastatin, the former
achieved a significantly lower reduction from
baseline (—7.8%, P<0.001).
Likewise, in a 12-week trial comparing
rosuvastatin 10 mg (n=129) and atorvastatin 10 mg
(n=127) in hypercholesterolaemic patients,15
rosuvastatin treatment significantly reduced apoB
(33% vs 26%; P<0.001), significantly increased
apoA-I (7% vs 3%; P<0.05) and significantly reduced
apoB:apoA-I ratio (37% vs 28%; P<0.001) as
compared with atorvastatin. Data have also been
reported comparing rosuvastatin 10 mg (n=111)
with pravastatin 20 mg (n=136) and simvastatin
20 mg (n=129).16 Rosuvastatin significantly
reduced apoB (40% vs 21% and 30%, respectively;
P<0.001 for both comparisons) and apoB:apoA-I
ratio (42% vs 23% and 32%; P<0.001 for both) as
compared with the other two statins, and
increased apoA-I to a comparable degree (5% vs 4%
and 4%, respectively). Similar results were
reported in a pooled analysis of 12-week
treatment periods in comparative trials with
rosuvastatin and other statins (Fig. 6).23
Conclusion
There is increasing evidence that apoB, apoA-I
and, particularly, the apoB:apoA-I ratio may be
more sensitive than LDL-C and HDL-C in predicting
CHD risk. It is arguable that use of these measures
is a next natural step in assessing patient risk, and
Apolipoproteins: the new prognostic indicator?
D15
Rosuvastatin
10 mg
Atorvastatin
10 mg
Rosuvastatin
10 mg
Simvastatin
20 mg
Rosuvastatin
10 mg
Pravastatin
20 mg
n=374
n=382
n=225
n=240
n=225
n=250
0
% Change from baseline
–10
10
–20
20
–23
–30
30
–31
–31
–40
40
–40
*
–40
*
–40
*
–50
50
–60
60
*P<0.001
rosuvastatin vs any of the comparator drugs.
Fig. 6 Pooled data analysis of effects on apoB/apoA-I ratio of rosuvastatin 10 mg vs atorvastatin 10 mg, pravastatin 20 mg and
simvastatin 20 mg in 12-week, fixed-dose phases of phase III trials. Patients had low-density lipoprotein ≥160 and <250 mg/dl and
triglycerides <400 mg/dl. (Data on file, AstraZeneca.)
would represent an alternative to use of nonHDL-C, for example, as a risk measure. It is also
clear that statin treatment produces sizable
reductions in apoB along with moderate increases
in apoA-I, thereby improving the apoB:apoA-I
ratio, which is consistent with the observed
effects of statins in reducing CHD risk. It remains
to be determined whether apolipoproteins should
replace or complement LDL-C and HDL-C as risk
markers and whether apolipoproteins should be
measured in all patients or, if not, whether
treatment decisions in specific subgroups of
patients would be facilitated by these measurements. It also needs to be determined whether
apolipoprotein goals should be incorporated into
future lipid-lowering guidelines and, if so,
precisely what these goals should be.
References
1. Marcovina SM, Albers JJ, Henderson LO, Hannon WH.
International Federation of Clinical Chemistry standardisation
project for measurements of apolipoproteins A-I and B:III.
Comparability of apolipoprotein A-I values by use of international reference material. Clin Chem 1993;39:773—81.
2. Marcovina SM, Albers JJ, Kennedy H, Mei JV, Henderson LO,
Hannon WH. International Federation of Clinical Chemistry
standardisation project for measurements of apolipoprotein A-I and B:IV. Comparability of apolipoprotein B value
by use of international reference material. Clin Chem
1994;40:586—92.
3. Genest JJ, McNamara JR, Ordovas JM et al. Lipoprotein
cholesterol, apolipoprotein A-I and B and lipoprotein(a)
abnormality in men with premature coronary heart disease.
J Am Coll Cardiol 1992;19:792—802.
4. Kwiterovich PO Jr, Coresh J, Smith H et al. Comparison of
the plasma levels of apolipoproteins B and AI, and other
risk factors in men and women with premature coronary
heart disease. Am J Cardiol 1992;69:1015—21.
5. Caslake MJ, Packard CJ, Suckling KE, Holmes SD,
Chamberlain P, Macphee CH. Lipoprotein-associated phospholipase A(2), platelet-activating factor acetylhydrolase:
a potential new risk factor for coronary artery disease.
Atherosclerosis 2000;150:413—9.
6. Sacks FM, Pfeffer MA, Moye LA et al. The effect of pravastatin on coronary events after myocardial infarction in
patients with average cholesterol levels. Cholesterol and
Recurrent Events Trial investigators. N Engl J Med 1996;
335:1001—9.
7. Lamarche B, Moorjani S, Lupien PJ et al. Apolipoprotein AI and B levels and the risk of ischemic heart disease during
a five-year follow-up of men in the Quebec Cardiovascular
Study. Circulation 1996;94:273—8.
8. Garfagnini A, Devoto G, Rosselli P et al. Relationship
between HDL-C and apolipoprotein A-I and the severity of
coronary artery disease. Eur Heart J 1995;16:465—70.
9. O’Brien T, Nguyen TT, Hallaway BJ et al. The role of
lipoprotein A-I and lipoprotein A-I/A-II in predicting coronary artery disease. Arterioscler Thromb Vasc Biol 1995;
15:228—31.
10. Luc G, Bard J-M, Ferrières J et al. Value of HDL-C,
apolipoprotein A-I, lipoprotein A-I, and lipoprotein A-I/A-II
in prediction of coronary heart disease. The PRIME Study.
Prospective Epidemiological Study of Myocardial Infarction.
Arterioscler Thromb Vasc Biol 2002;22:1155—61.
11. Walldius G, Jungner I, Holme I et al. High apolipoprotein B,
low apolipoprotein A-I, and improvement in the prediction
of fatal myocardial infarction (AMORIS study): a prospective study. Lancet 2001;358:2026—33.
12. Sharrett AR, Ballantyne CM, Coady SA et al. Coronary heart
disease prediction from lipoprotein cholesterol levels,
triglycerides, lipoprotein(a), apolipoproteins A-I and B, and
HDL density subfractions. The Atherosclerosis Risk in
Communities (ARIC) Study. Circulation 2001;104:1108—13.
13. Gotto AM Jr, Whitney E, Stein EA et al. Relation between
baseline and on treatment lipid parameters and first acute
major coronary events in the Air Force/Texas Coronary
Atherosclerosis Prevention Study (AFCAPS/TexCAPS).
Circulation 2000;101:77—84.
14. Simes RJ, Marschner IC, Hunt D et al. Relationship between
lipid levels and clinical outcomes in the Long-term
Intervention With Pravastatin in Ischemic Disease (LIPID)
D16
15.
16.
17.
18.
19.
20.
21.
22.
23.
Trial. To what extent is the reduction in coronary events
with pravastatin explained by on-study lipid levels?
Circulation 2002;105:1162—9.
Davidson M, Ma P, Stein EA et al. Comparison of effects on
low-density lipoprotein cholesterol and high-density
lipoprotein cholesterol with rosuvastatin versus atorvastatin in patients with type IIa or IIb hypercholesterolemia.
Am J Cardiol 2002;89:268—75.
Paoletti R, Fahmy M, Mahla G, Mizan J, Southworth H.
Rosuvastatin demonstrates greater reduction of low-density lipoprotein cholesterol compared with pravastatin and
simvastatin in hypercholesterolaemic patients: a randomized, double-blind study. J Cardiovasc Risk 2001;8:383—90.
Brown WV, Bays HE, Hassman DR et al., for the Rosuvastatin
Study Group. Efficacy and safety of rosuvastatin compared
with pravastatin and simvastatin in patients with hypercholesterolemia: a randomized, double-blind, 52-week trial.
Am Heart J 2002;144:1036—43.
Olsson AG, Pears J, McKellar J, Mizan J, Raza A. Effect of
rosuvastation on low-density lipoprotein cholesterol in
patients with hypercholesterolemia. Am J Cardiol 2001;88:
504—8.
Wiklund O, Davidson M, Chitra R, Hutchinson H, Raza A.
Rosuvastatin is more effective than atorvastatin in modifying lipid profiles and achieving National Cholesterol
Education Program Adult Treatment Panel III LDL-cholesterol goals [abstract 789]. Atherosclerosis 2002;3:237.
Barter P, Shepherd J, Brown WV, Bays H, Southworth H,
Strutt K. Rosuvastatin significantly improves lipid parameters and ability to achieve low-density lipoprotein cholesterol goals compared with pravastatin [abstract]. J Am Coll
Cardiol 2002; 39(suppl B):142B.
Hamilton-Craig I, Schaefer E, Farnier M et al. Rosuvastatin
produces significantly greater reductions in low-densitylipoprotein cholesterol and more patients achieve treatment goals compared with simvastatin [abstract 2811]. J
Am Coll Cardiol 2002; 39(suppl B):146B.
Schneck DW, Knopp RH, Ballantyne CM, McPherson R, Chitra
RR, Simonson SG. Comparative effects of rosuvastatin and
atorvastatin across their dose ranges in patients with
hypercholesterolemia and without active arterial disease.
Am J Cardiol 2003;91:33—41.
Hassman D, Ballantyne CM, Chitra R, Blasetto JW.
Rosuvastatin lowers non-high-density lipoprotein cholesterol and apolipoprotein B values compared with atorvastatin, simvastatin, and pravastatin. Circulation 2002; 105
(suppl):10.
Discussion
Dr Pedersen: As a cardiologist, I have to say that it
has taken us a long time to get comfortable with
measuring total cholesterol and HDL-C, and now
you want to introduce us to something new. In the
way of possible analogy, I think the reason that use
of troponins in diagnosis of myocardial infarction
was adopted so quickly is that labs simply stopped
C.J. Packard
giving out the old enzyme values and started
giving out just the troponin values. Do you think
there is a case for stopping measurement of
cholesterol and just measuring apoB and apoA-I?
Dr Packard: I think that apoB and apoA-I are
actually easier and better to measure than lipid
fractions. Whether you could get the audience
here, for example, to stop measuring cholesterol
… well, you can see the horror on some faces. But,
there are extremely good reasons to adopt such an
approach. I don’t think it will happen today or
tomorrow, but it will be interesting to see what
happens over the next 3 to 5 years in this regard.
I think the problem with cardiologists and
lipoproteins is that cardiologists didn’t invent
them. I’m wondering, Mr Chairman, whether
cardiologists invented troponins?
Dr Pedersen: What would the target levels be
for apoB and apoA-I?
Dr Packard: Well, an extremely simple starting
point would be 1 and 1: if apo B is above 1 g/l, do
something about it; if apoA-I is below 1 g/l, do
something about it.
Dr Pedersen: [Reads a question from the
audience.] Given the limited resources for health
care, is it realistic to consider spending more
money to measure apolipoproteins rather than
just continuing to measure total cholesterol?
Dr Packard: Most people measure total
cholesterol, and now we do direct measurement of
HDL-C in the lab, which is actually more expensive
than measuring apolipoproteins. So, if I were to
suggest a replacement, it might be to measure
total cholesterol, apoB and apoA-I. That might be
sufficient and would probably be less expensive
than getting a full lipid profile, and would give
provide as much information — if not more. So, I
don’t think cost is a central issue.
Dr Stein: [Reads a question from the audience.]
The proportion of treatment effect analysis in the
LIPID trial that you discussed showed that ontreatment apoB accounted for two-thirds of the
treatment effect and apoA-I accounted for 11%.
Did that analysis also look at the apoB:apoA-I
ratio?
Dr Packard: As far as I know the authors have
not looked at the ratio; it’s not in the Simes
article[14]. But, it would be nice to have the
analysis done to see if it agrees with the AFCAPS/
TexCAPS analysis.