Apolipoprotein-defined and NMR lipoprotein subclasses in the Veterans Affairs Diabetes Trial

Abstract

Aims

The VADT was a randomized clinical trial designed to assess the effect of intensive vs. standard glucose management on cardiovascular events in Type 2 diabetes. At the end of the study, intensive management failed to improve outcomes. We performed plasma lipoprotein subclass analyses to yield new information on the effects of study randomization on cardiovascular risk.

Methods

This is a cross-sectional study of a subset of the VADT (740 men: 368 intensive; 372 standard), conducted at least six months (mean ± SD: 2.1 ± 0.8 years) post-randomization. Conventional lipids, apolipoprotein-defined (ADLS) lipoprotein subclasses, ApoCIII, ApoE, and Nuclear Magnetic Resonance (NMR) lipoprotein subclasses were determined.

Results

In intensive vs. standard groups, conventional lipids and ADLS did not differ significantly. However, with intensive treatment, NMR-determined large and medium VLDL subclasses and VLDL diameter were lower, LDL diameter was higher, medium HDL was higher, and small HDL was lower (all p < 0.05). Also, ApoCIII levels were lower (p < 0.01).

Conclusions

In a subset of diabetic men from the VADT, intensive glucose management did not affect conventional lipids or ADLS, but had some beneficial effects on particle characteristics as defined by NMR and on ApoCIII.

Introduction

The Veterans Affairs Diabetes Trial (VADT) was a randomized clinical trial designed to determine whether intensive glucose management would improve cardiovascular (CV) outcomes. The study cohort comprised US veterans (mean ± SD; age 60.4 ± 9.0 years) with long-standing Type 2 diabetes (T2DM) (time since diagnosis: 11.5 ± 7.0 years). At baseline, 40% had established cardiovascular disease (CVD) and all others were at high risk (Duckworth et al., 2009). The primary endpoint was time to first occurrence of a composite of major CV events. A total of 1791 participants (97% men) were randomized to intensive (target HbA1c < 6%) vs. standard control, aiming for an absolute separation of 1.5% in HbA1c levels between randomization groups (Duckworth et al., 2009). The study concluded that intensive glucose control yielded only a non-significant 11.9% reduction in the primary endpoint (Duckworth et al., 2009). In accordance with the study protocol, blood pressure and conventional lipid profiles were maintained at similar levels between randomization groups throughout (Duckworth et al., 2009, Meyers et al., 2006).

Several hypotheses have been advanced to explain the absence of CVD benefit with intensive glucose management, a finding consistent with those of other large, recent cohort studies (Gerstein et al., 2008, Gerstein et al., 2011, Patel et al., 2008, Zoungas et al., 2010). Intensive management may have been initiated too late: there was a significant burden of disease at study entry, and an earlier, sustained intervention might have been more effective, as suggested by UKPDS (Holman, Paul, Bethel, Matthews, & Neil, 2008). Additionally, in the VADT, rosiglitazone was the first-line oral hypoglycemic agent, and was taken by a majority of participants in both randomization groups (Abraira, Duckworth, & Moritz, 2009). During the VADT, concerns developed that rosiglitazone may be implicated in CVD (Ajjan and Grant, 2008, Nissen, 2007a, Nissen, 2007b, Nissen and Wolski, 2007, Nissen and Wolski, 2010). While the randomization groups were comparable in terms of conventional lipid profiles, detailed studies of lipoprotein subclasses have not yet been reported, and have the potential to yield useful insights.

In this study, we employed two complementary methods to measure lipoprotein subclass profiles. First, apolipoprotein-defined lipoprotein subclasses (ADLS) define particles according to their qualitative apolipoprotein complement (Alaupovic, 1991, Alaupovic, 1996, Alaupovic, 2003). Lipoproteins belong to one of two ‘families’, containing either ApoAI or ApoB. The ApoAI family is broadly equivalent to HDL, and includes two subclasses (LpAI and LpAI:AII). The ApoB family is equivalent to VLDL, IDL, and LDL, and includes five subclasses (LpB, LpB:E, LpB:C, LpB:C:E and LpAII:B:C:D:E). In each case, the particle name reflects its qualitative apolipoprotein complement. Since apolipoproteins are critically important determinants of lipoprotein interactions with cells and enzymes, ADLS measures are believed to reflect particle metabolism (Alaupovic, McConathy, Fesmire, Tavella, & Bard, 1988). We also measured total apolipoprotein C-III (ApoCIII), which is involved in the transport and catabolism of triacylglycerols (Alaupovic, 1981, Ginsberg et al., 1986), as well as ApoCIII bound to ApoAI-containing lipoproteins (ApoCIII-HS, heparin soluble) and ApoCIII bound to ApoB-containing lipoproteins (Apo CIII-HP, heparin precipitate).

Second, Nuclear Magnetic Resonance (NMR) subclass analyses categorize lipoproteins according to particle diameter. The method does not involve physical separation of particles, and, unlike ADLS, is rapid and clinically applicable. It yields measures of three lipoprotein subclasses in the VLDL range, two in LDL, and three in HDL. It also provides estimates of molar particle concentrations and mean particle diameter for VLDL, LDL, and HDL. In contrast to ADLS, NMR provides structural, not functional measures of lipoprotein subclasses.

We report cross-sectional, post-randomization measures of conventional, ADLS, and NMR lipoprotein analyses in a subgroup of the VADT cohort, analyzing the data according to randomization group (intensive vs. standard glucose management).