Among patients on high-doses of statin therapy who achieve low levels of LDL cholesterol, HDL cholesterol concentrations are not predictive of cardiovascular (CV) risk, according to a sub-study of the JUPITER trial published July 21 in the Lancet.
“Randomized trials of statin therapy consistently report large, significant reductions in MI, stroke and vascular death in both primary and secondary prevention,” the authors wrote. While statin therapy is often a first-line treatment for patients with or at risk for CVD, there is often evidence of residual CV risk for these patients that may perhaps be due to HDL cholesterol concentrations.
To investigate whether low LDL levels via high-dose statins can eradicate this outlying risk, Paul M. Ridker, MD, of Brigham and Women’s Hospital and Harvard Medical School in Boston, and colleagues used the 17,802 patient cohort from the JUPITER trial, which assessed whether rosuvastatin (Crestor, AstraZeneca) decreased first-time CV events in patients compared to placebo.
The JUPITER trial results showed that when patients with high levels of C-reactive protein without diabetes or CVD were administered 20 mg of rosuvastatin per day, major adverse cardiovascular events (MACE) were reduced by 44 percent.
During the current sub-study, using the JUPITER trial's primary endpoint of non-fatal MI, non-fatal stroke, hospitalization for angina, arterial revascularization or CV death, Ridker et al divided patients into sex-specific quartiles of baseline and on-treatment HDL cholesterol and calculated CV events by HDL levels using Cox proportional hazard regression models.
Because women tend to have higher concentrations of HDL cholesterol than men and less of a risk for CV events, the researchers used sex-specific quartiles to avoid confounding data.
The sub-study showed that patients with heightened HDL cholesterol concentrations showed high levels of apolipoprotein A1, but low levels of triglycerides. In all four quartiles, CV event rates were significantly lower when rosuvastatin was administered.
Results showed that in the highest quartile of baseline HDL cholesterol concentration in the placebo group there was a reported 46 percent lower risk for CV events at study entry than the lowest reference quartile. Additionally, HDL cholesterol concentrations were inversely related to CV risk in patients administered placebo who had median on-treatment LDL-cholesterol concentrations of 2.8 mmol/L both at baseline and on-treatment.
While HDL cholesterol and baseline and on-treatment Apo A1 concentrations showed a strong link to the risk of CV events in the placebo group, there was no significant link in patients administered rosuvastatin who had low median LDL cholesterol concentrations equal to 1.4 mmol/L.
“This analysis provides little evidence that residual risk after aggressive use of statin therapy is related to HDL cholesterol concentration,” the authors wrote. “Although concentrations of HDL cholesterol inversely correlate with cardiovascular risk in the general population, whether HDL cholesterol concentrations remain an important biomarker of risk after LDL cholesterol reduction with high-dose statin therapy is not fully understood.”
Limitations of the sub-study were based on the fact that diabetics were excluded from evaluation during the JUPITER trial therefore, these results may not be generalizable to other patient populations.
“Our data should not reduce enthusiasm for measurement of HDL cholesterol concentration as part of an initial CV risk assessment; as shown here among those allocated to placebo, HDL cholesterol was a powerful inverse risk predictor,” the authors concluded.
“However, these primary prevention data and recent secondary prevention data from the TNT and PROVE IT trials provide little evidence to support the hypotheses that HDL cholesterol levels predict risk of vascular events in the setting of high-dose statin therapy.”
In an accompanying editorial, Derek J. Hausenloy, MD, of the University College London Hospital, and colleagues wrote: “Why HDL cholesterol concentrations did not predict CV risk at very low concentrations of LDL cholesterol in today’s JUPITER analysis is unclear.”
In addition, the authors said that additional lipid measures such as comparing Apo B to Apo A1 may more accurately predict CV risk better than HDL concentration.
“The findings from today’s study should not detract from the fact that raising HDL cholesterol remains a major