This site is intended for health professionals only

At the heart of general practice since 1960

HDL: after the unexpected

When a promising treatment failed, hopes that HDL therapy would rival statins fell with it. Chemical pathologist Dr Anthony Wierzbicki discusses the reasons why increasing HDL should be a good thing – but isn’t necessarily

When a promising treatment failed, hopes that HDL therapy would rival statins fell with it. Chemical pathologist Dr Anthony Wierzbicki discusses the reasons why increasing HDL should be a good thing – but isn't necessarily

The controversy

It has been universally agreed based on epidemiological studies that raising HDL-C is a good thing. Prospective studies show a consistent association between higher HDL-C and lower cardiovascular events. Indeed the evidence has been strong enough for target levels of >1mmol/l and >1.2mmol/l to be included in consensus statements from expert panels1.

Optimism about the relationship between HDL-C and cardiovascular risk drove the research behind finding agents that could raise HDL-C by as much as statins reduce LDL-C (25-50%). Thus there were great hopes attached to torcetrapib – the first CETP inhibitor – which raised HDL-C by 70 to 130%.

A trial programme was designed combining looking at markers of progression of atherosclerosis – carotid intima media thickness (RADIANCE 1 & 2) and coronary atheroma burden by intravascular ultrasound (ILLUSTRATE). A 15,000-patient, five-year endpoint-driven study (ILLUMINATE) also started.

Then the roof fell in. Routine safety monitoring of ILLUMINATE showed a 43% excess rate of deaths. Torcetrapib development was halted. By this time the RADIANCE and ILLUSTRATE studies had completed but showed no effect on atherosclerosis progression despite HDL-C being raised 50 to 60% and LDL-C reduced by 20%. The HDL field dissolved into the post hoc rationalisation that the human mind seems to require when something totally unexpected occurs.

The fall-out

Most lipid management is based on lowering LDL-C with statins and abundant evidence exists for the efficacy of statins, and managing HDL-C is not part of the primary targets in the guidelines, and has remained a peripheral activity. A few clinicians have attempted to raise HDL-C, for example in Indian Asians, patients with diabetes, insulin resistance or mixed hyperlipidaemia.

Reflecting on the evidence

The HDL controversy persists and will not be answered until at least 2010. All the previous clinical trials that raised HDL-C used either fibrates or niacin (nicotinic acid). In 1975, the Coronary Drug Project with 3g niacin showed a reduction in cardiovascular events at six years that in turn became a reduction in mortality at 15 years' follow-up.

Post hoc analyses seem to show benefit in patients with diabetes and those with metabolic syndrome including low HDL-C. In contrast the clofibrate arm of CDP showed no benefit2. The WHO study showed a reduction in coronary events with clofibrate but excess mortality.

Other fibrates have shown conflicting results. Studies with gemfibrozil in primary prevention (Helsinki Heart) and secondary prevention with low HDL-C (VA-HIT) were positive; bezafibrate in secondary prevention showed no benefit in the BIP study3. Fenofibrate reduced the progression of coronary atherosclerosis in DAIS yet the results of the FIELD study in type 2 diabetes are mired in controversy. So HDL researchers are awaiting modern endpoint studies.

Practical implications for GPs

Interpreting HDL levels

Measuring HDL-C forms part of standard risk assessment for cardiovascular disease and should be performed in all patients at risk, especially those with likely low levels – Indian Asians, Turks, West Africans. Modern methods for measuring HDL-C read slightly higher but are more reliable than those used in Framingham study so risk tends to be slightly underestimated in general. Patients with consistently ultra-low (<0.75mmol/l) and ultra-high HDL-C levels (>3.5mmol/l) may benefit from specialist advice or assessment.

What HDL level/cardiovascular risk level is required to treat HDL?

The general guidance that a HDL-C <1mmol/l in men and <1.2mmol/l in women remains valid. These cut-offs identify high-risk groups in all the epidemiology and fit one outcome trial (VA-HIT where the recruitment criterion was HDL <0.96mmol/l in men). Yet HDL-C is not a primary target of therapy. Total and LDL cholesterol levels are critical.

This is where TC:HDL ratios may mislead as they underestimate LDL-C driven risk in patients with high HDL-C and may overestimate it in patients with low levels due to a denominator effect. Identifying absolute levels of HDL-C and LDL-C is a much better guide to treatment decisions. So the need for treatment should be assessed on the presence of significant atheromatous disease or calculated cardiovascular risk >20% per decade.

Primary therapy should be directed to reducing LDL-C with lifestyle and then statin therapy with the aim of achieving at least a 30% reduction and preferably levels <3mmol/l and better closer to 2mmol/l, especially in patients with established CVD.

Statins have modest effects in raising HDL-C. In general statins raise HDL-C more in patients starting with lower levels and have shown benefit in a low HDL-C primary prevention population in the AF/TexCAPS study.

In patients with low HDL-C it is worth repeating HDL-C on statin therapy and determining the effect. If low levels persist then consideration should be given to therapies to raise HDL-C. These are mostly prescribed in secondary care or by lipid specialists in primary care as combination therapy requires extra input to be performed successfully.

Reviewing lifestyle or prescribed medications that may reduce HDL

Smoking reduces HDL-C. Smoking cessation therapy will raise HDL-C levels and deliver additional benefits on other causes of morbidity and mortality. Relatively few drugs lower HDL-C. The concerns usually attach to hormone replacement therapy where progestagens lower HDL-C.

Similarly tibolone reduces HDL-C but there is controversy over whether this represents a hyper-functional effect. There is little unequivocal evidence for HRT in cardiovascular prevention and so this can be discontinued if it is being used for cardiovascular purposes.

Prescribed medications that may increase HDL

The actions of so-called HDL-C raising drugs are actually complex and few raise HDL-C significantly (>20%). All the agents currently used raise HDL-C as a corollary of reducing triglycerides, and increase lipoprotein particle sizes across all classes though niacin also has direct effects.

• Fibrates are PPAR-agonists which reduce triglycerides and raise HDL-C by 10 to 15%. The most successful in trial terms, gemfibrozil, has least effect on HDL-C levels, no effect on LDL-C, but has significant drug interactions with statins. Bezafibrate and fenofibrate show lesser interactions with statins though the risk of rhabdomyolysis is increased twofold (compared with fivefold for gemfibrozil). Yet the results of FIELD using fenofibrate in diabetes were controversial, especially in combination therapy3.

• Nicotinic acid (niacin) acts through an adipocyte receptor to reduce free fatty acid release. However, as yet the exact mechanism of HDL-C raising by niacin remains unclear. Niacin increases HDL-C by up to 25% and also decreases triglycerides, LDL-C and lipoprotein (a) levels. Multiple studies have shown benefits with niacin in monotherapy, in combination with fibrates or statins in reducing progression of atherosclerosis2.

The Coronary Drug Project demonstrated in 1975 that niacin can reduce cardiovascular events and eventually mortality. But its use is limited by flushing which is not totally aspirin suppressible.

More modern formulations show lower rates of flushing and newer versions are in development. In addition co-formulation with a PGD1 receptor antagonist seems to reduce flushing dramatically. These novel formulations will be tested on top of statin therapy in cardiovascular disease in the AIM-HIGH and THRIVE/HPS-2 trials.

• Other methods of raising HDL-C include treatments that improve insulin resistance or reduce weight. On average a 1kg weight loss induces a 1-2% rise in HDL-C in insulin-resistant patients irrespective of whether this is achieved by diet and exercise or through pharmacotherapy4.

Conclusion

Raising HDL-C is probably a good idea but remains to be proven in outcome studies.

Dr Anthony Wierzbicki is consultant chemical pathologist and director of the lipid unit at St Thomas' and Guy's Hospitals

• Low HDL-C is common in patients with insulin resistance/metabolic syndrome

• Patients with low HDL-C (<1mmol in="" men;=""><1.2mmol )="" in="" women)="" may="" require="" specific="">

• Fibrates reduce triglycerides and raise HDL-C by 10-15% and may reduce cardio• Low HDL-C is common in patients with insulin resistance/metabolic syndrome

• Patients with low HDL-C (<1mmol in="" men;=""><1.2mmol )="" in="" women)="" may="" require="" specific="">

• Fibrates reduce triglycerides and raise HDL-C by 10-15% and may reduce cardioKey points

• Low HDL-C is common in patients with insulin resistance/metabolic syndrome
• Patients with low HDL-C (<1mmol in="" men;=""><1.2mmol )="" in="" women)="" may="" require="" specific="">
• Fibrates reduce triglycerides and raise HDL-C by 10-15% and may reduce cardiovascular events
• Nicotinic acid raises HDL-C by up to 25% and reduces progression of atherosclerosis and cardiovascular events
• It is possible to raise HDL-C by other methods, such as diet and exercise, and weight loss

Rate this article 

Click to rate

  • 1 star out of 5
  • 2 stars out of 5
  • 3 stars out of 5
  • 4 stars out of 5
  • 5 stars out of 5

0 out of 5 stars

Have your say