Dr Dan Rader (Source: University of Pennsylvania)

HDL is thought to bring about its beneficial effects mainly by reverse cholesterol transportmoving cholesterol from the artery wall back to the liver for excretion. However, it also has anti-inflammatory, antioxidant, antithrombotic, and nitric-oxide-inducing properties. It is not known exactly how much these other properties contribute to the benefits associated with HDL, or whether the cardioprotective properties of HDL are specific to one or more of the many HDL subfractions, and these issues are the subject of much investigation at present. HDL researcher Dr Dan Rader (University of Pennsylvania, Philadelphia) says: "I think mechanisms other than reverse cholesterol transport are potentially important but of unproven significance in vivo." Another basic researcher in the field, Prof Monty Krieger (Massachusetts Institute of Technology, Cambridge), believes the nitric-oxide-inducing properties are probably quite important but says the quantitative significance of these alternative actions are unknown. "We just don't know for sure what role they play," he commented.

Cholesterol is transported to the artery wall by LDL and transported from the artery wall back to the liver by HDL.

Although it has long been suspected that HDL is protective, raising HDL levels as an approach to reducing heart disease risk has not really been fully explored as yet, but that is now changing. And the field received a tremendous boost last year with the ApoA-1 Milano study.

Dr Prediman K Shah

ApoA-1 Milano is an unusual variant of ApoA-1, the main constituent of HDL. The Milano variant was discovered by Drs Cesare Sirtori and Guido Francheschini (University of Milan, Italy) in a group of related Italians who all had unusually low HDL levels but no sign of cardiovascular disease.1 Suspecting that this ApoA-1 Milano may be more protective than normal ApoA-1, researchers, led by Dr Prediman K Shah (Cedars-Sinai Medical Center, Los Angeles, CA), showed that injection of ApoA-1 Milano reduced the lipid content and inflammation of atherosclerotic lesions and halted the progression and induced regression of atherosclerosis in animal models.2

Dr Steven Nissen

Then, last year, a group led by Dr Steven Nissen (Cleveland Clinic, OH) replicated these results in humans. Using the relatively new technique of intravascular ultrasound (IVUS), they reported that just five weekly infusions of a recombinant version of ApoA-1 Milano produced a modest but significant 4% regression of coronary atherosclerosis in 47 ACS patients.3 As the first compelling evidence of atherosclerosis regression in humans, this study really set the whole HDL field alight.

"Exciting is not the word to describe my ApoA-1 study. Astonishing is more suitable," Nissen says. "Nothing I will ever do will have so much impact as that." He adds: "We thought HDL was protective, but no one had shown a convincing change in risk with HDL. Now we have shown that. I know it was only in 47 patients, but this was the only 47-patient trial that shook the world. Now everyone is accelerating their HDL programs." Nissen's study was indeed well received. It attracted extensive media coverage, in which the Milano variant was described as "liquid Drano for the coronary arteries," and served as a major stimulant for further HDL research.

This was the only 47-patient trial that shook the world.

This study is particularly exciting in that it suggests that raising HDL may be the key to atherosclerosis regression, which has remained elusive with interventions that lower LDL. In another IVUS study recently conducted by Nissen (REVERSAL), the highest doses of statins that reduced LDL the most halted the progression of atherosclerosis but did not induce regression.4 Dr Christie Ballantyne (Baylor College of Medicine, Houston, TX) says: "REVERSAL should have been called HALT. I think we can say from those results that lowering LDL will not cause regression."

Dr Christie Ballantyne

Nissen says that this isn't really all that surprising considering the roles of LDL and HDL. "LDL is the culprit behind the accumulation of atheroma, but it does not have any role in removing cholesterol. That is the job of HDL. We have to target both LDL and HDL. We're not going to abandon LDL and the statins, but we need to add in HDL-raising therapy. With LDL lowering we have been stuck with a 35% reduction in event rates. That is wonderful if you are one of those 35%, but it's not so great if you are one of the other 65%. They need something else. And that something else is probably HDL-raising treatment."

Kastelein agrees: "If you can get a 50% to 60% reduction in LDL and a 50% to 60% increase in HDL, we are no longer talking about the 20% to 30% reduction in heart disease seen with the statins, but a 70% reduction. This would be hugely significant." And Shah has a similar view. "It is very exciting for our patients. There is real potential to reduce coronary events by 70%, 80%, or 90% by combining potent LDL-lowering and HDL-raising therapies. The past two decades were dominated by LDL. The next decade will belong to HDL," he comments.

REVERSAL should have been called HALT.

Other research published recently suggests that high levels of HDL are also associated with long life and higher cognitive scores in the elderly.5 Kastelein notes: "The beauty of HDL is that more is better and less is worse in almost all situations. This is wonderful for drug development. It is the same for ApoA-1, the main component of HDL. ApoA-1 is good for you in almost all situations. I can't think of one clinical situation where it is not beneficial to raise ApoA-1 or HDL."

Dr Virgil Brown

But Dr Virgil Brown (Emory University, Atlanta, GA) is a little more cautious and thinks the situation is rather more complicated. "The key issue is to understand what we're doing when we raise HDL. Some HDL-raising strategies may be good; others may be bad or nonproductive," he commented.

The ApoA-1 Milano results have catapulted drug companies and cholesterol researchers worldwide into a frenzy of renewed enthusiasm over the development of new drugs that raise HDL. But even the most advanced of these is a few years away from the market at best, so in the meantime, some clinicians are making the best of what's already available. Lifestyle changes, such as exercising, losing weight, drinking modestly, and quitting smoking can have a modest effect on raising HDL, and there are also a few available drugs that can be tried.

The beauty of HDL is that more is better and less is worse in almost all situations.

Niacin: Worth trying despite side effects?

The main drug already marketed that is known to raise HDL is niacin, but it comes with side effects that put many people off trying it. Ballantyne believes it is still worth pursuing, as 2000 mg of niacin can produce a 30% increase in HDL. "I work in a referral center with very motivated patients, and 85% can take at least 1000 mg. But in nonreferral centers the acceptance of niacin will be worse. The main problem is flushing. And it is not a simple drug to use. You have to give 500 mg first, and then increase to 1000 mg, then 2000 mg, and it can cause total body flushing, glucose intolerance, increased uric acid levels, and liver-function abnormalities."

He believes that using niacin and a statin together is a good strategy at the present time, when options for HDL raising are limited. He says the effects of niacin and statins should be additive, and a preliminary study (HATS by Dr B Greg Brown) has suggested that this combination can cause modest regression of atherosclerosis.6 "This was not the world's biggest study, but it did show a little regression," he comments.

We are using a lot of the niacin-plus-statin combination with very good results.

Shah also believes this combination has merit. "Niacin is a reasonably effective drug for raising HDL. It is very cheap, so no drug company is pushing it, and many people are therefore unaware of its benefits. One third of patients can't tolerate it, but that means two thirds can," he says. "Based on the HATS study, we are using a lot of the niacin-plus-statin combination with very good results. Physicians should be aggressively using this combination," he adds.

But Kastelein is not a fan of niacin, which he says is not frequently used in Europe because of its side effects. However, he is eagerly awaiting the introduction of the long-acting formulation, Niaspan^® (Kos Pharmaceuticals Inc, Miami, FL), in Europe later this year, which he says would be a good choice to use with a statin. But he believes a better option would be the development of new molecules in this class that may lack the side-effect profile of niacin. "It is better to develop new drugs with better HDL-raising action and fewer side effects."

And a "niacin without the side effects" is one possibility that is being investigated for new HDL drugs. This relies on identification of the pathways through which niacin works to develop drugs with similar but maybe more specific actions. Niacin acts on nicotinic acid receptors, and this receptor has been cloned and new nicotinic acid agonists are in development. "This is very early, with no clinical trials yet, but there is a lot of good basic work going on," Ballantyne notes.

It is better to develop new drugs with better HDL-raising action and fewer side effects [than niacin].

The fibrates also seem to have a modest effect on raising HDL, and the VA HIT study with gemfibrozil showed a significant reduction in cardiac events.7 But, as pointed out by Dr Richard Havel (University of California, San Francisco), who was involved in some of the original work during the 1950s characterizing HDL, fibrates also lower triglycerides, "so we don't know which of these effects was responsible for the reduction in events in VA HIT."

So there are at least two available drugs that have some effect on raising HDL, but new agents with more potent effects are needed. Ballantyne explains, "If we consider that the optimal LDL is below 100 mg/dL, that is the bottom quintile in the US. If we use the same logic for HDL, we should be aiming at getting people into the top quintileabout 60 to 80 mg/dL. But we can't get anywhere near that with current therapeutic interventions. There is, however, lots of promise and room for improvement."

Several new classes of drugs are now in development that have much more potent HDL-raising effects. The most advanced are the cholesteryl ester transfer protein (CETP) inhibitors.

CETP inhibitors. CETP is a plasma glycoprotein that has several functions, one of which is to mediate the transfer of cholesterol from HDL to LDL particles. CETP inhibition may thus be expected to lead to higher HDL concentrations. Two CETP inhibitors are in late-stage clinical developmentone from Pfizer (torcetrapib) and one from Japan Tobacco (JJT-705). The Pfizer drug has just started phase 3 trials, and the JT compound has just finished phase 2. Pfizer is developing one pill that would contain both atorvastatin and torcetrapib that would give both significant LDL lowering and HDL raising. "That's very exciting," says Ballantyne.

Phase 2 results with torcetrapib have recently been published, showing significant increases in HDL (of up to 100%).8,9 Nissen is now conducting a phase 3 IVUS study measuring plaque burden. The trial involves 886 patients, half of whom are given atorvastatin alone and half of whom are given atorvastatin plus torcetrapib. The study has just started, and results will be available in two years.

If this IVUS study is positive and if the FDA accepts IVUS as a legitimate surrogate end point, torcetrapib could be approved in the next two to three years. Pfizer is also conducting a clinical-end-point trial with torcetrapib, but results from this trial will not be available for several years. Nissen does not believe that it should be necessary to wait for these results for approval of torcetrapib. "Do we really have to wait that long for this drug? I am optimistic that we can convince the FDA to approve new drugs on the basis of a reduction in plaque burden, especially now that IVUS has been validated by the REVERSAL and PROVE-IT studies," he commented. Other experts have mixed views on this issue. One who believes that IVUS studies may not be sufficient for approval is Dr Michael Miller (University of Maryland Medical Center, Baltimore). "The FDA has become more conservative in this regard, especially in the wake of the negative clinical trials with HRT and antioxidants," he points out.

Dr Michael Miller (Source: University of Maryland Medical Center)

Kastelein is also involved in the CETP-inhibitor program, running trials of both the Pfizer and JT compounds. He noted that phase 2 trials have shown that the JT compound raises HDL by about 35% or 40%, whereas the Pfizer compound seems to raise HDL slightly more. But he says this does not necessarily mean that torcetrapib will be better in terms of reducing CHD events. "We believe that raising HDL is good, but we don't know that raising it by an enormous amount will be better than just raising it moderately," he points out.

There are many uncertainties that remain with the CETP inhibitors. Brown points out that animal models have shown conflicting results with CETP inhibition, with some actually suggesting that this could be a detrimental approach. "I am a bit concerned about CETP inhibitorsthey prevent the transfer of cholesterol from HDL to LDL. If we give statins and so lower LDL, then also block transfer of cholesterol to LDL, we may block the transfer of cholesterol to the liver," he explains. "In some animal models more atherosclerosis is seen with CETP inhibition. It is true that rodents do not have CETP, so these studies are far from conclusive, but CETP has evolved for some reason in humansit may have some advantage."

What is really sexy about IVUS is that if we see the loss of plaque we can say that the drug is working.

Shah also notes that not all people who have a mutation in the CETP gene that causes high HDL are protected from heart disease. "The type of HDL formed under these circumstances is thought not to be as effective as natural HDL. That is a concern," he says. And Havel is similarly cautious: "People with CETP deficiency have high HDL, but it is not at all clear whether they are protected from CHD. It is certainly more complicated than saying that high HDL is the be-all and end-all of risk."

Nissen accepts these concerns. "The important issue is whether the HDL we are forming with CETP inhibitors has the same benefits as normal HDL. What is really sexy about IVUS is that if we see the loss of plaque we can say that the drug is working." Kastelein agrees: "We have to wait for the human studies before we know what the real effects of blocking this enzyme will be. Nissen's IVUS study will be a great help in this regard," he comments. And both Brown and Shah say their concerns will be alleviated if the IVUS studies with CETP inhibitors show atheroma regression.

Rader says CETP inhibition should not be regarded as representative of other HDL-raising approaches if it fails. "I think it might workthere is a reasonable chance, but it's not guaranteed. It is the most immediate strategy in development, but if it doesn't work we shouldn't be put off the whole field of raising HDL, as there are mechanistic issues that may explain why it may not work," he explains.

As Ballantyne points out: "There is a lot of discussion about which is the right way to raise HDL. We just don't know yet. But there is an aggressive development program under way with the CETP inhibitors. We should know if they work in about two years."

PPARs. Another class of drugs in development to raise HDL are the peroxisome proliferation activated receptor (PPAR) agonists. PPARs are nuclear receptors that activate a gene (ABC-A1) that in turn stimulates the first step of reverse cholesterol transportthe efflux of cholesterol out of cells onto HDL particles.

There are three types of PPARsalpha, gamma, and deltaand all are targets for agonists as drugs. Fibrates work by binding to PPAR alpha, causing both an increase in HDL and a decrease in triglycerides. PPAR gamma raises HDL only modestly, and its major effect seems to be a reduction in insulin resistance. The glitazone diabetes drugs work as PPAR gamma agonists. And PPAR delta seems to have all three effects.

Rader notes that PPAR research is a very active field. "There is lots of interest in PPARs as they are nuclear receptors, which make good targets for small molecules. Also, there are two drug classes already available that target PPARs, so this strategy has been shown to be plausible," he explains.

He says that PPAR gamma agonists with improved HDL-raising properties are a particular target, and there are some early candidates. "Dual alpha and gamma agonists would be even better. They are thought to have the most potential as they affect two different targets. There are some in early clinical trials." Rader explains that these dual agonists are being developed for both diabetes and lipid disorders, which are of course complementary disease states. "We are studying glitazones in metabolic-syndrome patients with low HDL, which could be a possible new indication for these drugs," he notes.

Other experts also believe PPARs have potential but point out that there are side-effect concerns with this class. Indeed, Merck has recently dropped a PPAR agonist from quite late-stage clinical research because of tumors in animal studies.

Kastelein comments: "Yes, the PPARs are exciting, but there are concerns. If we turn on the PPARs we may turn on many other genesso they may have side effects. We don't really know what to expect." He believes that the fact that these drugs are agonists makes it more difficult for them to have one specific action. "There are very few examples of successful pharmaceuticals that are agonists, as they need to stimulate only one target to avoid side effects. This is quite difficult to achieve. It is much easier to selectively block one target. That is why most drugs are antagonists. Agonists tend to be more general and to activate more than one receptor or enzyme pathway, which can give side effects. PPARS are all agonists. I am working with them and of course would like them to be successful, but we have to wait and see."

Brown says he is concerned that the idea of jointly stimulating gamma and alpha PPARs has not been explored enough before new dual agonists are developed. "We have drugs available that act as PPAR-alpha agonists (fibrates) and others that act as PPAR-gamma agonists (glitazones). We should be doing studies of these two drug types together to see whether they are synergistic before spending millions developing new drugs. I understand that much more money can be made from developing a new patentable product. But I cannot understand why such a huge investment would be made before the scientific question has been settled of whether a dual PPAR-alpha and -gamma agonist would be a good idea."

I cannot understand why such a huge investment would be made before the scientific question has been settled.

Brown says he believes the PPARs are "gimmicky drugs," and the fact that the Merck product ran into side-effect problems should make researchers extra vigilant for similar problems with other drugs in this class. "We need to be very careful with clinical trials of these drugs now. PPAR agonists turn on a lot of different genes. We don't know what effect that will have."

HDL mimetics. Of all the approaches to raising HDL, the most obvious is just to give HDL or its main constituent, Apo-A1, or agents that mimic these substances. This is the basis for the Apo-A1 Milano studies, with Nissen's IVUS clinical trial actually performed with a synthetic Apo-A1 Milano mimic, Esperion's ETC-216.

The Milano mimic itself has several limitationsit is too large to be orally available so would not be suitable for chronic long-term treatment, and it is expensive and laborious to make. But it is still being developed as a short-term treatment. Shah notes that the Milano peptide could be given intravenously in the short term after an acute event when the patient is in the hospital and then treatment with a different drug could be given on discharge.

Orally available peptides. However, Esperion, along with several other groups, is also working on smaller peptides that are claimed to be orally active. This is where the real hope lies. Kastelein says that if the orally available mimetics work, that would be "very significant," but he remains a little skeptical of the oral-availability claims until he sees the human data. "If you infuse HDL or ApoA-1 Milano it is amazing how much cholesterol you can get out of the artery wall, but these will never make a chronic therapy. Pfizer bought Esperion because of the great potential of ApoA-1 Milano and the ApoA-1 mimetics. Certainly this is a very exciting development, but the problem is that orally available in rabbits may not be the same as orally available in humans. I would like to see human data before making any judgments."

Virgil Brown has very similar views: "I'd like to see a study of the smaller peptides themselves rather than Milano. Milano is too big. The cost of making it will be huge, and it will be impractical. But smaller peptides are very feasible. There is some suggestion that the dextro-amino-acid versions of these peptides do have some intestinal absorption. But this is very early."

One company developing such a peptide is Bruin Pharmaceuticals of Los Angeles, a small startup founded by UCLA cardiologist Dr Alan Fogelman. Its peptide, known as D4F, is claimed to be orally active and has shown a reduction of atherosclerosis in mice. Shah, who is also conducting studies with D4F, says: "We find it is absorbed orally and shows pretty significant effects on atherosclerosis in vein grafts and has the potential to go into clinical trials. It would be very exciting if it worked in humans."

Another, possibly more controversial, approach to HDL manipulation is upregulation of the SR-BI receptor in the liver. One of the pioneers of this strategy is Krieger, who discovered that SR-BI was a receptor for HDL. He explains that SR-BI clears HDL from the plasma, thus mediating cholesterol transit from the blood to the bile. Because of this it results in a reduction in blood levels of HDL, but Krieger claims that this is not always a bad thing.

Dr Monty Krieger (Source: MIT)

He says: "Rather than concentrating on simply raising HDL, I believe a better way of looking at this field is helping HDL work better." He points out that in murine models, if SR-BI is absent, HDL increases and so does atherosclerosis, and if SR-BI is upregulated there is a reduction in atherosclerosis and a reduction in HDL, "as SR-BI is literally sucking it out of the blood." Krieger notes that if HDL is most effective in reducing atherosclerosis by removing cholesterol from the artery to the liver, upregulating SR-BI would increase the flux of cholesterol movement but result in a lower steady-state level of HDL in plasma.

Some experts have voiced skepticism about whether a therapy based on SR-BI would be viable, as it would lead to a lowering of HDL, which goes against the current obsession with raising HDL. But Krieger says this is not sensible. "Yes, I can see that there may be a barrier to overcome to convince people that in some cases lowering HDL could be beneficial, but surely if that is the best way of reducing risk, it should darned well be done. It seems rather Luddite to say we are not going to do it because people may not like the idea of lowering HDL if it is in fact the right thing to do. If it is the right thing to do, it is the right thing to do. Full stop." If what happens in mice happens in humans, SR-BI would be a very attractive target.

It seems rather Luddite to say we are not going to do it because people may not like the idea of lowering HDL if it is in fact the right thing to do.

He says the example of ApoA-1 Milano illustrates his point "The people in Italy with ApoA-1 Milano seem to be protected from heart disease, but they actually have low levels of HDL. The Milano mutation may actually be helping HDL to work better so that you do not need so much," he argues.

And then there is the possibility of gene therapy. Brown believes that HDL is a prime target for gene therapy. "This is a much more adventuresome strategy but could be the answer in the long term and would be relatively inexpensive. We need safe vectors for long-term administration, but these are coming now."

Shah and his team are working on a gene therapy based on the ApoA-1 Milano gene in an adeno-associated virus vector and have shown that such gene transfer is feasible and reduces atherosclerotic plaque build-up after a single injection. "We are introducing this into animals so that they start producing their own ApoA-1 Milano. We already have proof of concept, but it is not yet ready for humans. This is very promising for the future," he noted.

Several other approaches are also under investigation. These include the following:

Endothelial lipaseIncreasing this leads to a reduction in HDL, so inhibiting endothelial lipase would make sense. Rader is working on this approach.

LXR (another nuclear receptor) agonists and other small moleculesThese increase ABC-A1 and thus cause efflux of cholesterol from cells. Various molecules have been identified but side effects have so far prevented their development.

While raising HDL certainly seems to be a promising strategy, there are many issues that need to be clarified. For example, are different types of HDL important? Shah believes that the quality of an individual's HDL may be just as important as the quantity. "Certain forms of HDL are effective and others are not. Particles enriched with ApoA-1 are good, but those with ApoA-2 are not." While it is true that in general the higher the HDL, the fewer clinical events, there is no good functional assay to assess the quality of HDL, he notes. "ApoA-1 Milano is a great example of how quality, not just quantity, can be important. In individuals with this mutation, their HDL is low but seems to be turbocharged in some way. Similarly, mutations in CETP give high HDL levels, but the HDL produced may be dysfunctional and may not be protective. These are exceptions to the general rule that higher HDL levels are better."

Small, dense LDL is more harmful than fluffy, puffy LDL, but nobody has really focused on thateveryone is just concentrating on lowering LDL.

And then in some circumstancesfor example, when an acute event occursHDL can change from being protective to harmful.10 Shah explains: "You can have good-quality HDL, which is anti-inflammatory, but then when you have an acute event this turns proinflammatory but then later switches back. We don't really understand what is going on here, but it is clear that HDL is incredibly complex. There is lots we don't know about the detailsbut the first wave of research is just about raising it."

He points out that LDL is similar: "Small, dense LDL is more harmful than fluffy, puffy LDL, but nobody has really focused on thateveryone is just concentrating on lowering LDL full stop."

But Brown notes that HDL is much more complicated than LDL. "HDL is involved in at least nine pathways. LDL is a simple molecule with one or two receptors involved in uptake or degradation. We cannot apply the simple model of LDL to HDL." He also believes that different HDL strategies may be appropriate for different people: "We're not all one gene pool. CETP inhibition may be more successful in one group of patients and another strategy in another group."

So much more basic research in this field is needed and is under way. Rader believes the focus should be on discovering new targets for HDL drugs. "This is a huge issue," he says. "We have only a few targets at present, and this is not enough when we're looking at a whole new area of research." He adds that current work trying to understand which genes are involved in HDL action should provide new targets.

And the exact mechanisms behind the benefit also need to be explored. As Krieger notes: "We don't fully understand how HDL exerts its benefits. We know it is protective, but we don't know exactly how this is so. We've got a lot more to learn."