CardioExchange Archive

In a new analysis of data from the JUPITER trial, researchers found that high-density lipoprotein particles (HDL-P) may be a better marker of residual risk than chemically measured high-density lipoprotein cholesterol (HDL-C) or apolipoprotein A-1 (apoA-1). CardioExchange’s editor-in-chief Harlan Krumholz interviewed lead investigator Samia Mora about the analysis, which was recently published in Circulation.

Harlan Krumholz: You previously showed in the JUPITER trial that on-treatment HDL-C was not predictive of residual risk among statin-treated individuals, while HDL-C was predictive among those taking placebo. Should we have risk-stratification tools that are specific to whether patients are taking statins?

Mora: It is not yet clear whether specific risk-stratification tools should be used for patients taking statins compared with those not taking statins, and whether such a strategy would improve outcomes among statin-treated patients. What is clear is that vascular events among statin-treated patients remain common, especially among patients with known vascular disease, and that there may be other risk factors or determinants of residual risk among statin-treated patients.  For example, in the TNT trial of patients with known coronary disease, standard risk factors and other known clinical variables explained a moderate amount of the residual risk. However, the c-index of 0.7 suggested that additional factors may contribute to residual risk.

Our earlier JUPITER results provocatively suggested that on-statin HDL-C levels were not associated with residual vascular risk, but we were unable to rule out a possible association for HDL-C with residual risk in JUPITER because of the relatively small number of events in the rosuvastatin arm.  In a recent meta-analysis of eight statin trials (including JUPITER) among 38,153 statin-allocated patients with a total of 5387 vascular events, on-treatment HDL-C levels were predictive of residual risk.

Harlan Krumholz: In your new article you suggest that other indices of HDL might be more useful in patients on statins. How did you come up with that hypothesis?

Mora: Our hypothesis was driven by our earlier intriguing findings from the MESA study.  Among statin-naïve individuals, the number of HDL-P was a better predictor of subclinical atherosclerosis and clinical events compared with HDL-C.

Harlan Krumholz: You find that HDL-P is predictive in patients on rosuvastatin but not in those on placebo. Given this finding, what should practicing doctors do now?

Mora: In our current study from JUPITER, we found that the number of HDL-P was associated with incident vascular events among both the placebo- and rosuvastatin-allocated groups, and was a better predictor of residual risk among the rosuvastatin-allocated group than HDL-C or apoA-1.  Our results should be viewed as hypothesis generating and will require further evaluation in other studies before they can be applied clinically. However, the current results from JUPITER along with the earlier results from MESA highlight that HDL-C is not the same measure as HDL-P number, and that other properties of HDL may be significant.  We just don’t know what we should do about them. Practicing doctors can reduce residual risk by applying and reinforcing multi-factorial risk reduction strategies that have been demonstrated to be efficacious and life-saving.

Harlan Krumholz:  Also, if you are a patient on statins, can you now ignore your HDL level?

Mora: HDL levels can be measured with various tests, such as HDL-C, HDL-P number,  HDL size, HDL function, apoA-1, and many other emerging properties of HDL. Our current study suggests that patients with a low HDL-P number on statin therapy are more likely to suffer from vascular events compared with individuals who have a low HDL-Con statin therapy. We just don’t know yet what is the best way to reduce this “HDL-related” risk.

Harlan Krumholz: We have seen patients whose HDL levels declined on statins, do you have any evidence about what that does to risk?

Mora: The best observational evidence to-date addressing this question comes from the large meta-analysis of 38,153 patients, which had enough events by strata of HDL-C change.  In that study conducted among statin-allocated patients, neither an increase nor a decrease in HDL-C levels related to subsequent vascular risk. By contrast, an increase in apoA-1after statin allocation was associated with a reduction in subsequent vascular events. Interestingly, apoA-1 is more correlated with HDL-P than it is with HDL-C, but we don’t have enough data to address whether an increase in HDL-P number after statin therapy would result in better outcomes.