| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters |
1
Core Laboratory for Clinical Studies, Washington University, School of Medicine, St. Louis, MO 63110
2
Esperion Therapeutics, Inc., Ann Arbor, MI 48108
aAuthor for correspondence. Fax 314-362-4782; e-mail thom{at}im.wustl.edu.
To the Editor:
Warnick et al. (1) recently reviewed the status of the measurement of HDL-cholesterol (HDL-C) and cautioned against the use of the new homogeneous methods, when atypical specimens may be present, without additional validation of the accuracy of the methods. We recently had the opportunity to analyze specimens from a clinical trial, which will serve as an excellent example of the warning given by these authors.
In a phase I clinical trial, 32 healthy volunteers were infused with a synthetic HDL comprising recombinant apolipoprotein (Apo) A-IMilano and phospholipid [1-palmitoyl-2-oleoyl-sn-3-glycerophosphocholine (POPC)]. This cholesterol-free, HDL-like complex is referred to as ETC-216. The intent of the trial was to determine safety/tolerability before initiating a phase II study in patients with acute coronary syndromes. It has been suggested that ApoA-IMilano/phospholipid infusions may rapidly stabilize atherosclerotic lesions by mobilizing vessel wall cholesterol (2).
Initial measurements of HDL-C concentrations before and 30 min after the start of the infusion gave divergent results when we compared results obtained with a polyethylene glycol/sulfated 
-cyclodextrin-based method (Roche HDL Plus) to those obtained with separation of HDL by size-exclusion chromatography by fast protein liquid chromatography (FPLC) with online determination of cholesterol concentration (3). To further investigate this discrepancy, sera from a subset of seven patients (fasted males; doses of 50–100 mg ApoA-IMilano protein/kg; infusion time, 1–1.3 h) were referred to the Core Laboratory for Clinical Studies for measurement of HDL-C by several homogeneous and traditional precipitation methods (see Table 1
). It is important to note that before treatment with ETC-216, each method reported similar HDL-C concentrations, suggesting agreement under normal conditions. Serum samples were obtained at 0.5 and 2 h after the start of the infusion and pooled for each patient because HDL-C values by FPLC were not different at these time points.
|
After treatment with ETC-216, the two precipitation methods based on dextran sulfate (Mr 50 000) and heparin manganese gave similar results, with a decrease in HDL-C of 21–25%. The HDL-C concentrations obtained by the homogeneous methods were different from the precipitation methods and also quite different from each other. The antibody-based Sigma EZ method showed an increase in HDL-C similar to the increase in total cholesterol (
7%), whereas the polyethylene glycol/sulfated -cyclodextrin-based Roche HDL Plus method showed a substantial loss in HDL-C (82%). FPLC showed the greatest increase in HDL-C (40%), consistent with animal studies in which a single infusion of ETC-216 was administered and HDL-C was measured by FPLC (2). In both animals (2) and humans, the acute increase in HDL-C is primarily cholesterol in the unesterified form. Both nuclear magnetic resonance spectroscopy and polyacrylamide gel electrophoresis indicated an increase in HDL size attributable to ETC-216 infusion, i.e., an increase in "large" HDL and a decrease in "small" HDL (data not shown).
ApoA-IMilano is a molecular variant of wild-type ApoA-I that contains an Arg173
Cys substitution, which leads to the formation of disulfide-linked dimers. These dimers may alter the physicochemical (4) and functional (5) properties of ApoA-I. In our study, the association of dimeric ApoA-IMilano with the endogenous HDL apparently changed the properties of HDL particles. One possible explanation could be the enhanced association of ApoE with ETC-216-modified HDL, and perhaps endogenous HDL, leading to differential interactions with the various reagents, each of which determines the HDL-C concentration by a different chemical mechanism. We did observe reductions in ApoE concentrations in supernatants of the chemical precipitation methods in the presence of ETC-216, as well as decreases in supernatant ApoA-I concentrations vs whole serum values. It appears that the marked differences among the HDL methods shown in Table 1
with ApoA-IMilano infusion are not observed in humans infused with wild-type ApoA-I because an increase in HDL-C was observed with either polyethylene glycol precipitation or FPLC (6)(7). Moreover, to our knowledge, HDL-C method discrepancies have not been reported in ApoA-IMilano individuals, although because all individuals are heterozygous, the concentrations of the mutant protein may be much lower. We in fact saw no difference between FPLC and the Roche HDL Plus methods at low doses of ETC-216.
The purpose of describing this experiment is not to suggest that one method is superior to another in all cases; some other perturbation to HDL structure may be detected very differently by the same reagents. In clinical practice, one would not expect to see such radical diversity in HDL as was observed in this clinical trial; however, these data indicate the extent to which different methods for measuring HDL-C can be affected, leading to very different interpretations of clinical values. Such differences would not be readily apparent to a clinician unless measurements were made on the same specimen by various methods, which in reality never occurs. The resolution of this dilemma will require time as the idiosyncrasies of each method are discovered and reported in the literature.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
