|
|
||||||||
Lipids and Lipoproteins |
a Author for correspondence. Fax 207-883-1527; e-mail wcraig{at}fbr.org.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
Retrospective studies, in which subjects with existing disease were compared with matched controls, have consistently shown an important association between increased concentrations of Lp(a) and IHD. Subjects in these studies have included survivors of myocardial infarction (4)(5), patients with symptoms of angina (6), and those with atherosclerosis confirmed by B-mode ultrasonography (7) or coronary angiography (8)(9). An association has also been demonstrated between Lp(a) concentrations and the severity of atherosclerosis, as quantitated by angiography (10).
Retrospective studies cannot distinguish whether Lp(a) is in the causative pathway or is increased for reasons secondary to the presence of disease. Since 1990, several prospective studies have addressed this issue, but with inconsistent results. Consequently, there remains uncertainty concerning both the relationship between Lp(a) and IHD and the utility of Lp(a) measurement in clinical practice (11)(12). We therefore performed a metaanalysis of the published prospective studies in an effort to resolve the uncertainty.
| Materials and Methods |
|---|
|
|
|---|
Data collected from identified studies included the age, sex, ethnicity, and clinical history of the participants. Studies were excluded if the study group included subjects with a history of IHD on entry, if subjects were preselected for a particular disease state that was itself associated with increased risk for IHD (15)(16), or if quantitative data were not available (17)(18)(19). Quantitative data on the number of subjects and respective serum or plasma Lp(a) concentrations for cases and controls were collected. Estimates of the mean, geometric mean, or median Lp(a) concentration for each subgroup were accepted for analysis; information on centiles or SD for the Lp(a) distribution in each group were also collected. Finally, data related to potential sources of heterogeneity were collected, including information on the type of Lp(a) assay used, the temperature of sample storage, and the length of sample storage.
data analysis
Lp(a) concentrations were expressed in mg/L; where necessary,
original data were converted from mg/dL, IU/L, or nmol/L. Conversion
from nmol/L to mg/L was performed using a molecular weight estimate for
the apo(a) assay calibrator of 299.376 kDa (20). IU/L
(21) was directly equivalent to mg/L.
The difference between cases and controls in each study was expressed as the ratio of Lp(a) concentration in cases divided by the Lp(a) concentration among controls [Lp(a) data were not transformed]. This approach provides a means of normalizing data between studies that use different Lp(a) assays, with different approaches to assay standardization. The case:control ratio is a robust estimate of the discriminatory power of Lp(a), assuming that the SD of the two population subgroups are similar between studies. Use of the Lp(a) case:control ratio also allows arithmetic mean Lp(a) data [not a satisfactory representation of the center of the highly skewed Lp(a) distribution] to be included, assuming that the shift in arithmetic means between cases and controls is proportional. The approach of pooling Lp(a) case:control ratio data makes the assumption that differences in Lp(a) concentrations between cases and controls are proportional between studies; proportional differences are typically seen when differences between assays are caused by problems with standardization. The Lp(a) case:control ratio should not be confused with odds or risk ratios; it simply expresses the ratio of Lp(a) concentrations between cases and controls before the development of IHD.
Log transformations were used where necessary. If not provided, log SD
was calculated using either centiles of the Lp(a) distribution provided
directly in the study (22)(23) or histograms
provided in the study (24)(25)(26)(27)(29). In four
studies
(21)(28)(30)(32), it was
not possible to estimate SD; therefore, a pooled estimate, weighted by
the number of cases, was used. Because original data were not
available, the log SD of the Lp(a) ratio was estimated from the central
estimates and log SD of cases and controls, assuming certain
conditions, using the method of Armitage and Berry (33),
where:
![]() |
Summary estimates of the effect [as Lp(a) case:control ratio] were calculated from the individual Lp(a) case:control ratios and the log SDs of their distributions, using published methodology (34) that included a formal test of heterogeneity. This method weights the estimate from each study by the inverse of a combination of within- and among-study variances. Analyses were performed using a statistical software package from BMDP Statistical Software, Inc.
| Results |
|---|
|
|
|---|
|
Quantitative Lp(a) data are shown for cases and controls in Table 2
. Data from 1617 cases (190 females and 1427 males) and 10 035
controls (357 females and 9678 males) were available for analysis.
There was marked variation in Lp(a) concentrations between studies,
even after conversion to the same units [median Lp(a) concentrations
in controls ranged from 64 to 288 mg/L]. This variation [45-fold for
all studies and 6-fold when the two most discrepant studies
(27, 29) are excluded] highlights the need for
an additional normalization step before combining data between studies.
|
Fig. 1
shows the Lp(a) case:control ratios and 95% confidence
intervals for each study. There was a positive association of the
combined estimate with IHD risk [Lp(a) case:control ratio 1.40 (95%
confidence interval, 1.221.57)]. This relationship was similar
between males [1.42 (1.211.63)] and females [1.32 (1.191.45)].
There was considerable variability in Lp(a) case:control ratios (range,
0.682.00); the heterogeneity observed was highly significant
(
= 901, P <0.001). The age of study
subjects, the length of follow-up, the length of sample storage, and
sample storage temperature were investigated as potential sources of
the observed heterogeneity. The storage temperature demonstrated the
strongest relationship with Lp(a) case:control ratio (r
= -0.44, n = 13, P = 0.13). Thus, lower freezer
temperatures correlated with a higher point estimate for the Lp(a)
case:control ratio.
|
Of the 12 studies included in the analysis, 9 provided a
doseresponse analysis of the relationship between Lp(a) and IHD
risk. Among these studies, six showed evidence of a positive
doseresponse effect
(21)(23)(26)(28)(31)(32),
one was equivocal (30), and two found no doseresponse
effect (22)(25) (Table 3
). Among studies that found Lp(a) to be a positive prospective
risk factor for IHD, seven tested the independence of this effect by
adjustment of data for coexisting risk factors
(21)(23)(26)(27)(29)(30)(31).
Although the specific covariates examined differed between studies, in
no case did adjustment for age, blood pressure, body mass index,
smoking, or lipoprotein-related variables eliminate the significant
relationship between Lp(a) and IHD risk
|
| Discussion |
|---|
|
|
|---|
There was marked heterogeneity in the extent of the relationship between Lp(a) and IHD between studies that was too great to be explained by chance alone. Our data suggest that this may be attributable in part to differences in sample storage temperature between studies. Although the inverse relationship between the Lp(a) case:control ratio and sample storage temperature was not statistically significant, it was large enough to suggest a trend and, moreover, was consistent with the results of formal storage studies. We have shown previously that sample storage may have a significant effect on Lp(a) concentration, with the extent differing according to temperature of storage (35). Kronenberg et al. (13) noted a proportional effect of sample storage on the results of Lp(a) measurement, in that greater loss of mass was noted among samples with low apo(a) molecular weight, even at -80 °C. Because there is an inverse association between apolipoprotein(a) size and Lp(a) mass (36), these data indicate that samples with higher Lp(a) concentrations had quantitatively lower recovery of Lp(a) concentration after storage; such an effect would tend to minimize observed differences in Lp(a) concentrations between cases and controls.
This sensitivity to storage temperature indicates that sample handling is a critical issue for studies using banked samples, and future studies should provide careful documentation of sample handling protocols. Regarding the present analysis, consideration of sample storage temperature effects suggests that the estimate we give for the Lp(a) case:control ratio is likely an underestimate of the actual relationship between Lp(a) and IHD.
At present there are no guidelines for the use of Lp(a) measurement in clinical practice, either as a screening test for IHD or to direct or monitor therapy in subjects with identified risk factors or disease. Even when a factor contributes to the development of a condition, this does not necessarily mean that its measurement is useful for screening. In the case of Lp(a), there is insufficient separation between concentrations from cases and controls to support the use of Lp(a) measurement as a screening test for IHD risk in the general population.
Although Lp(a) cannot be used as a screening test for IHD in the general population, its measurement may have a role in other clinical situations. For example, Lp(a) testing may be useful to monitor risk reduction efforts in subjects with IHD or high IHD risk. In addition, evidence from one study [40 subjects, 9 events ((37))] for interactions between Lp(a) and LDL in IHD risk has led to the suggestion that Lp(a) testing be used to identify individuals who might benefit from more aggressive LDL reduction therapy (38)(39).
Several elements must be considered in developing an objective strategy for clinical Lp(a) testing. First, Lp(a) assays must be standardized. Second, additional therapies must be developed to reduce Lp(a) concentrations. The presently available approach to reducing Lp(a) concentrations is essentially limited to the use of niacin (38), which is not universally tolerated. Third, intervention trials need to be performed to establish the benefit of Lp(a) reduction efforts. Finally, additional studies are needed to clarify the role of Lp(a) measurement in subjects undergoing therapy for other IHD risk factors.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
The following articles in journals at HighWire Press have cited this article:
![]() |
K.-L. Chien, H.-C. Hsu, T.-C. Su, F.-C. Sung, M.-F. Chen, and Y.-T. Lee Lipoprotein(a) and Cardiovascular Disease in Ethnic Chinese: The Chin-Shan Community Cardiovascular Cohort Study Clin. Chem., February 1, 2008; 54(2): 285 - 291. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. R. Kamstrup, M. Benn, A. Tybjaerg-Hansen, and B. G. Nordestgaard Extreme Lipoprotein(a) Levels and Risk of Myocardial Infarction in the General Population: The Copenhagen City Heart Study Circulation, January 15, 2008; 117(2): 176 - 184. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Yologlu, A. T. Sezgin, N. Sezgin, E. Yetkin, and E. Olmez Comparison of Different Regression Analyses for Identifying Risk Factors in Obese and Nonobese Patients With Coronary Artery Disease Angiology, November 1, 2007; 58(5): 543 - 549. [Abstract] [PDF] |
||||
![]() |
B. Smolders, R. Lemmens, and V. Thijs Lipoprotein (a) and Stroke: A Meta-Analysis of Observational Studies Stroke, June 1, 2007; 38(6): 1959 - 1966. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. C.H. Kwan, F. Kronenberg, S. Beddhu, and A. K. Cheung Lipoprotein Metabolism and Lipid Management in Chronic Kidney Disease J. Am. Soc. Nephrol., April 1, 2007; 18(4): 1246 - 1261. [Full Text] [PDF] |
||||
![]() |
J. Suk Danik, N. Rifai, J. E. Buring, and P. M Ridker Lipoprotein(a), measured with an assay independent of apolipoprotein(a) isoform size, and risk of future cardiovascular events among initially healthy women. JAMA, September 20, 2006; 296(11): 1363 - 1370. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Kollerits, M. Auinger, V. Reisig, T. Kastenbauer, A. Lingenhel, K. Irsigler, R. Prager, and F. Kronenberg Lipoprotein(a) as a Predictor of Cardiovascular Disease in a Prospectively Followed Cohort of Patients With Type 1 Diabetes Diabetes Care, July 1, 2006; 29(7): 1661 - 1663. [Full Text] [PDF] |
||||
![]() |
O. Zschenker, T. Illies, and D. Ameis Overexpression of lysosomal Acid lipase and other proteins in atherosclerosis. J. Biochem., July 1, 2006; 140(1): 23 - 38. [Abstract] [Full Text] [PDF] |
||||
![]() |
I. Shai, E. B. Rimm, S. E. Hankinson, C. Cannuscio, G. Curhan, J. E. Manson, N. Rifai, M. J. Stampfer, and J. Ma Lipoprotein (a) and coronary heart disease among women: beyond a cholesterol carrier? Eur. Heart J., August 2, 2005; 26(16): 1633 - 1639. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Moffat Braving New Worlds: To Conquer, to Endure Physical Therapy, November 1, 2004; 84(11): 1056 - 1086. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Carmena, P. Duriez, and J.-C. Fruchart Atherogenic Lipoprotein Particles in Atherosclerosis Circulation, June 15, 2004; 109(23_suppl_1): III-2 - III-7. [Abstract] [Full Text] |
||||
![]() |
J.-C. Fruchart, M. C. Nierman, E. S. G. Stroes, J. J. P. Kastelein, and P. Duriez New Risk Factors for Atherosclerosis and Patient Risk Assessment Circulation, June 15, 2004; 109(23_suppl_1): III-15 - III-19. [Abstract] [Full Text] |
||||
![]() |
A. A. Ariyo, C. Thach, R. Tracy, and the Cardiovascular Health Study Investigators Lp(a) Lipoprotein, Vascular Disease, and Mortality in the Elderly N. Engl. J. Med., November 27, 2003; 349(22): 2108 - 2115. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Yologlu, A. T. Sezgin, R. Ozdemir, N. Sezgin, C. Colak, E. Topal, and I. Barutcu Identifying Risk Factors in a Mostly Overweight Patient Population with Coronary Artery Disease Angiology, March 1, 2003; 54(2): 181 - 186. [Abstract] [PDF] |
||||
![]() |
V. Stangl, G. Baumann, and K. Stangl Coronary atherogenic risk factors in women Eur. Heart J., November 2, 2002; 23(22): 1738 - 1752. [Full Text] [PDF] |
||||
![]() |
J. Danesh, R. Collins, and R. Peto Lipoprotein(a) and Coronary Heart Disease : Meta-Analysis of Prospective Studies Circulation, September 5, 2000; 102(10): 1082 - 1085. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.G. Isles and J.R. Paterson Identifying patients at risk for coronary heart disease: implications from trials of lipid-lowering drug therapy QJM, September 1, 2000; 93(9): 567 - 574. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. M. Kostner, A. Ibovnik, H. Holzer, and H. Grillhofer Preparation of a stable fresh frozen primary lipoprotein[a] (Lp[a]) standard J. Lipid Res., December 1, 1999; 40(12): 2255 - 2263. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Boffelli, D. A. Zajchowski, Z. Yang, and R. M. Lawn Estrogen Modulation of Apolipoprotein(a) Expression. IDENTIFICATION OF A REGULATORY ELEMENT J. Biol. Chem., May 28, 1999; 274(22): 15569 - 15574. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |