Ezetimibe and low density lipoprotein subfractions: an ongoing debate

Abstract

There is evidence that small dense low density lipoprotein (sdLDL) subfractions are more atherogenic than buoyant LDL. Therefore, it is relevant to assess the effect of lipid lowering drugs on LDL subfractions. In this editorial we discuss the effect of ezetimibe (EZE; a cholesterol transport inhibitor) alone or in combination with a statin, on sdLDL levels. We conclude that, based on currently available evidence, more studies are needed before a definitive answer can be provided regarding the effect of EZE on sdLDL levels.

Key words::

• Cholesterol
• Ezetimibe
• Simvastatin
• Small dense low density lipoprotein

In this issue of Current Medical Research and Opinion Florentin et al.1 report the effect of simvastatin (SIMVA; 40 mg/day) monotherapy vs SIMVA (10 mg/day) + ezetimibe (EZE; 10 mg/day) combination therapy on small dense low density lipoprotein (sdLDL) subfractions. Producing similar changes in the routinely measured lipid fractions was achieved for high density lipoprotein cholesterol (HDL-C) levels (an increase of 0.3% in both groups; p = ns vs baseline) and triglycerides (TG; −17% for SIMVA and −19% for SIMVA + EZE; p < 0.0001 vs baseline). For LDL-cholesterol (LDL-C) the fall was significantly greater with combination therapy (−43% for SIMVA and −49% for SIMVA + EZE; p < 0.05). The sdLDL decreased to a similar extent in both groups (−42% for SIMVA and −46% for SIMVA + EZE; both p < 0.0001 compared with baseline). Therefore, broadly comparable changes in LDL-C, HDL-C and TG achieved by a statin alone or a lower dose of a statin + EZE resulted in similar decreases in sdLDL levels.

Florentin et al.1 also consider the contradictory literature regarding the effect of EZE on LDL subfractions. These discrepancies may be attributed to the small numbers of patients included and differences in their selection criteria as well as variations in the methodology used to measure sdLDL. Baseline lipid levels may also contribute to the different results. For example, Florentin et al.1 report that baseline sdLDL-C and TG levels independently and significantly correlated with the alterations in sdLDL-C concentrations in both groups. In another study2, baseline LDL-C levels predicted the response to treatment, whereas others reported that a beneficial effect of EZE on sdLDL was only observed at higher TG levels3. In agreement with the findings of Florentin et al.1, no differences were found in LDL subfraction responses when comparing SIMVA + EZE vs atorvastatin4. Furthermore, in the same study4 TG levels tended to influence the LDL subfraction results for both treatment options.

Clearly there is a need for large appropriately designed studies to resolve the controversies discussed above. Such a study should consider the baseline lipid profile. Furthermore, the treatment-induced alterations in lipid profile should be similar in the groups that include or exclude EZE. In this context, it is of interest that the changes in LDL-C levels from baseline in one study2 were: 22.1, 40.7 and 59.6% for EZE alone (10 mg/day), SIMVA alone (40 mg/day) and SIMVA + EZE (40 + 10 mg/day), respectively. Corresponding changes for TG levels were: +27, −8.9 and −11.8%, respectively2. Since LDL-C and TG levels may influence the amount of sdLDL1–5, these differences in responses may have affected the result, i.e. that EZE had an adverse effect on sdLDL fractions2. However, LDL size was not significantly affected by all treatments2.

A key issue for future studies is to establish or refute the principle of ‘equivalence’. In other words, whether similar reductions in routinely measured lipid variables (i.e. LDL-C, HDL-C and TG) achieved by EZE alone, statin alone or statin + EZE lead to comparable changes in sdLDL levels. These studies should be powered to show differences between groups at the end of the treatment period unlike in other studies2. Power calculations should be feasible now that smaller studies have been published1–4. In addition, it has been argued that subjects with normal cholesterol levels are a better model to assess the effectiveness of EZE2. Although this suggestion is acceptable, we also need to consider that this is not the population that will be treated with EZE. Furthermore, there is some evidence that the TG lowering effect of EZE may be greater in patients with higher baseline TG levels6–9. In turn, TG levels may influence the change in sdLDL, as mentioned above1,3–5.

Linking any EZE-related effect on sdLDL to the outcomes of EZE trials is premature. Several factors lead us to this conclusion:

1. The EZE-sdLDL studies published to date have limitations, as briefly discussed above and by others1,2,5,6,10.

2. The evidence that sdLDL is more atherogenic than buoyant LDL is substantial but not definitive, especially in terms of a lack of large randomised double blind intervention trials with clinical endpoints5,10. Moreover, any effect on sdLDL must be considered against the background of several lipid and non-lipid changes that may follow treatment with a statin, EZE or both drugs5,6. In particular, a meta-analysis (n = 5039) showed that the addition of EZE to a statin decreased LDL-C levels by 23.6%11 along with favourable changes in HDL-C and TG concentrations. Florentin et al.1 also showed that the EZE + SIMVA combination produced a significantly (p < 0.05) greater fall in apolipoprotein B levels than monotherapy with SIMVA. This may be relevant as apolipoprotein B concentrations represent the number of LDL particles2 and have been considered as a better indicator of cardiovascular risk than LDL-C with the added advantage of being measured in the non-fasting state12. Of note, Florentin et al.1 also reported that high sensitivity C-reactive protein, a predictor of vascular risk, was equally decreased by 23% (p < 0.05 vs baseline) in both groups.

3. The methods used to measure sdLDL are not equivalent1,2,3,5,10. Which one best represents any cardiovascular risk while also being amenable to ‘routine’ use remains to be established.

4. The EZE trials are the subject of substantial criticism regarding their design and interpretation1,2,6,10. In addition, the Simvastatin and Ezetimibe in Aortic Stenosis (SEAS) trial findings can be re-interpreted in terms of aortic valve events (AVEs) and ischaemic cardiovascular events (ICEs)13,14. SEAS compared placebo vs EZE + SIMVA (10 + 40 mg/day) in patients with aortic stenosis. The reduction in ICEs (47 and 36%) in the two lowest tertiles of aortic jet velocities in SEAS was as expected from meta-analyses of statin treatment13,14. Treatment with EZE + SIMVA did not influence AVEs13,14; probably statin monotherapy also does not improve AVEs13,14. Furthermore, the evidence that EZE may reduce vascular events has been supported by the recently presented Study of Heart and Renal Protection (SHARP) trial in patients (n = 9500 approximately) with impaired renal function (pre-dialysis and on dialysis)15. Treatment with EZE + SIMVA (10 + 20 mg/day) compared with placebo resulted in a reduction in cardiovascular events to an extent expected from meta-analyses15. However, we should await the publication of the full paper before making more detailed comments.

In conclusion, more studies are needed before a definitive answer can be provided regarding the effect of EZE on sdLDL16. In the wider context, we need to consider that if LDL-C goals are not achieved with a well tolerated statin dose, adding EZE (or another lipid lowering agent) remains an option, even in high-risk patients such as those with diabetes mellitus17.