Too Much of a GOOD Thing? Dysfunctional HDL
Case Presentations:
Patient A:
A 54 year old woman is referred to the cardiology prevention clinic for a severely elevated coronary artery calcium score (CAC) of 310 on a CAC CT scan.
She reports that since her mid 20’s her total cholesterol has at times been over 300 mg/dL and her LDL-C has always been fairly high, around 150 mg/dL but she was told not to worry about it because her HDL-C is also “super high,” around 120 mg/dL, making her ratio ok.
She is otherwise very fit and healthy, exercising at a high level about 5-6 days a week. She eats a whole foods diet, rarely containing red meat. She has never smoked and drinks a couple of glasses of wine weekly.
Given that she is in her 50’s, she decided to get a CAC CT screening test along with all of her other screenings for cancer and is shocked by the results.
Flying Under the Radar
Common, but poor advice: “Yeah, your LDL-C is a bit high, but your HDL-C is also high and that makes up for it. Overall your ratio is great, so you have nothing to worry about. Besides, you’re super fit. Keep doing what you’re doing!”
This is an unfortunate and important cause of “residual risk” in women particularly, since women tend to have higher HDL-C levels on average than men. But certainly can be seen with men as well.
Ratios really serve no purpose in assessing risk in a lipid panel. Bonus points are no longer given for high HDL-C.
HDL-C, which stands for High Density Lipoprotein Cholesterol, is the mass sum of all the cholesterol contained within High Density Lipoproteins and is measured in milligrams per deciliter (mg/dL) of blood.
HDL-C, as a biomarker, tells us nothing about cardiac risk.
Why? Because the actual cholesterol concentration itself doesn’t tell us how HDL is functioning in its role in reverse cholesterol transport (bringing cholesterol from peripheral tissues back to the liver).
Scientists are trying hard to develop “functionality” assays but some argue it will never be accomplished.
High density lipoprotein (HDL) is a very complex lipoprotein that is not only involved in reverse cholesterol transport (RCT) but also has other important functions that are anti-inflammatory, antioxidant, antithrombotic and immune modulating.
Unfortunately, many clinicians continue to use outdated methods for analyzing lipid profiles, namely ratios and awarding “bonus points” for having high HDL-C levels.
In general, it is believed that having low HDL-C is bad and having high HDL-C is good. But this is not always the case.
There are examples of genetic variants (Apo A1 Milano), where HDL-C is in the teens, yet this population of Italians has one of the highest concentrations of centenarians in the world. It’s felt that their HDL is so efficient at RCT that the HDL particle does not need to carry a lot of cholesterol.
On the other end of the spectrum there are those with very high HDL-C, for instance >100 mg/dL, where the HDL is inefficient at RCT and the HDL particle becomes overloaded with cholesterol.
Read more in Deep Dive.
CardioAdvocate Checklist:
Review your lipid panel
Focus on LDL-C and nonHDL-C. These biomarkers are surrogates for ApoB-containing particles. This is what determines atherogenic risk.
Triglycerides: While not themselves the appropriate biomarker for determining ASCVD risk, they are important for recognizing other concomitant diseases, when elevated. For more on this see “Atherogenic triad.”
Same goes for HDL-C. Look at it. See if it’s low or really high, as it might alert you to other problems, but it is not, itself, to be used in ACVD risk assessment.
If TG’s are high, >150 mg/dL, ALWAYS look at nonHDL-C. If not reported, you must use math and calculate it. (Provider tip: all labs CAN report nonHDL-C. If they don’t, call the director, give them the equation TC-HDL-C and provide reference ranges, which are 30 points higher than the LDL-C reference range).
Deep Dive
Yes, every day in the preventive cardiology clinic we spend considerable amounts of time refuting and reeducating patients (and their referring providers) that there is no such thing as “good” cholesterol. It’s not their fault, it’s just that medical education often moves slowly and entrenched models, when inaccurate, are tough to undo.
For many years, describing HDL as GOOD cholesterol and LDL as BAD cholesterol, were convenient and simple ways to explain their respective roles in cholesterol metabolism to the general public in a Public Service Announcement-like model.
It was long observed in various trials that low HDL-C was associated with higher rates of cardiovascular disease and higher HDL-C concentrations seemed to offer protection against heart disease.
Naturally, this observation sparked investigation into drug therapies that could raise HDL-C
HDL Functionality vs HDL Cholesterol
As a biomarker,HDL-C tells us nothing about HDL functionality. When looking at HDL-C on the lipid panel, it’s tempting to react positively to a high level and negatively to a low level, but that doesn’t tell the full story.
This is not to say that one shouldn’t look at HDL-C on the lipid panel. It may offer clues to some sort of “phenotype” or other disease pathology. But one must not look at HDL-C or any of its ratios and make any assumptions about how HDL is behaving.
Dysfunctional HDL
In the case of very high HDL-C the danger in assuming it’s protective or may make up for a high LDL-C (nice ratio) is that risk may be inappropriately ignored or downplayed by both the patient and their provider, to the point where the patient fails to seek appropriate screening and/or fails to receive appropriate therapy and counseling. This is a very important cause of residual risk but what should be an obvious one. Meaning, they are “hiding in plain sight.”
Here is nice short paper about dysfunctional HDL on the NLA website: Lipid Luminations: Elevated and Dysfunctional HDL | National Lipid Association Online
Apo A1 Milano
On the other end of the spectrum and perhaps less consequential if “missed” is the story of Apo A1 Milano, where very low levels of HDL-C are protective due to highly functional HDL.
Apo A1 Milano is a heterozygous gene mutation with a single amino acid substitution, first discovered in a family in Northern Italy in 1980. The condition manifests as VERY low levels of HDL-C and moderate hypertriglyceridemia (Cardiovascular Status of Carriers of the Apolipoprotein AI Milano Mutant : The Limone sul Garda Study), but unlike others with a similar lipid phenotype, does not lead to an increased risk in ASCVD and is associated with increased longevity.
Naturally, this sparked interest in the development of therapies such as this trial studying the Effect of Recombinant ApoA-I Milano on Coronary Atherosclerosis in Patients With Acute Coronary Syndromes: A Randomized Controlled Trial using intravenous injections and showing plaque regression with intravascular ultrasound (IVUS). This was obviously a promising result however this was not observed in the MILANO-PILOT TRIAL which evaluated the use of this in those with acute coronary syndrome (ACS) treated with statins.
Enthusiasm has been tempered for now, after the 2024 publication of the large scale RCT (randomized controlled trial) AEGIS II in The New England Journal of Medicine showed no improvement in MACE with Apo A1 infusions after acute myocardial infarction (AMI).
The Pitfalls of Raising HDL-C
AIM HIGH Trial: Niacin and the HDL-C hypothesis
If one wanted to make all of the lipid panel metrics look pretty, Niacin would appear to be the perfect lipid lowering therapy. It lowers LDL-C, it lowers triglycerides and it raises HDL-C.
While statins have been King of lipid lowering therapies since 1987, it’s well known that those with low HDL-C (and high TG’s) identify a population of patients with significant residual risk (See “Atherogenic triad”). In its heyday Niaspan (extended release Niacin FDA approved in 1997) was heavily utilized as an adjunct to or in replace of statins due to its impact on those lipid panel metrics.
Then AIM-HIGH was published. This was an National Institutes of Health (NIH)/National Heart, Lung, and Blood Institute (NHLBI) secondary prevention RCT of ~3400 patients designed to assess cardiovascular outcomes when using Niacin on top of simvastatin + ezetimibe, if needed, for those already at LDL-C goal but with low HDL-C. It was halted after 3 years due to futility, despite raising HDL-C from a median of 35 mg/dL to 42 mg/dL, reducing TG from 164 mg/dL to 122 mg/dL and lowering LDL-C from 74 mg/dL to 62 mg/dL.
It was explained that the lack of incremental benefit of Niacin was likely due to the fact that, while secondary prevention patients are at high risk, this group had been on statins for a while and were very stable. Furthermore, at that time it was felt that the evolution of statins and consequently more aggressive LDL-C goals, had reached a point where additional nonstatin therapies such as Niacin, Fibrates and CTEP inhibitors (in development) were simply not likely to impart much additional risk reduction. Meaning, we were at a point where perhaps that was about as good as we could do. We, of course, now know that’s not true.
Off-Target Effects: Niacin not so harmless
HPS2-THRIVE
The HPS2-THRIVE trial essentially killed Niacin. This was a massive RCT studying ~25,000 patients over 4 years, with established vascular disease on statin therapy (secondary prevention) on either extended release Niacin-laropiprant (antagonist of the prostaglandin D2 receptor DP1 that reduces flushing) or placebo.
There was no significant risk reduction. There were significant adverse events to include worsening diabetes control, new onset diabetes, gastrointestinal (peptic ulcers, diarrhea, indigestion), musculoskeletal (myopathy, or muscle weakness/dysfunction) and skin-related serious adverse events. There was also an unexpected increase in serious infection and in serious bleeding.
Importantly, the musculoskeletal side effects of myopathy were 10 times worse in China than in Europe.
This is a good example of why we need large scale randomized trials assessing outcomes. It doesn’t do anybody any good to improve biomarkers or risk factors, if the off target effects cause significant harm.
This is why utilizing supplements as if they were therapeutic drugs is risky, when they have not been appropriately studied in large scale RCTs. It’s the wild west. One cannot know the risk to benefit ratio.
Niaspan, or extended release Niacin, has been discontinued.
CETP Inhibitors: How lipid biomarkers continue to fool us
The history of the development of CETP inhibitors is a fascinating story and teaches us many lessons about biomarkers, drug development and understanding biology.
Cholesteryl Ester Transfer Protein (CETP) is a protein that transfers cholesterol esters from HDL to other ApoB containing lipoproteins such as LDL, in exchange for triglycerides.
CETP inhibitors were developed initially in an effort to raise HDL-C, as we chased that HDL-C theory.
Initially this failed, but not for reasons convincing enough to dismiss HDL-C.
Below, we march through each trial and their respective downfall.
Nonetheless, after many failed iterations of this novel pathway, there is a glimmer of hope, as yet another CETP inhibitor appears to have risen from the dead, offering new insight into this mechanism. CETP inhibition may yet have a viable and bright future in lipid therapeutics and beyond.
Read on and stay tuned, as this story continues to unfold.
Torcetrapib: Off-target effects (2007)
Torcetrapib is a potent CETP inhibitor and the first to be studied in the ILLUMINATE Trial (NEJM), which was a large RCT looking at 15,000 patients with high cardiovascular risk on atorvastatin + torcetrapib or atorvastatin alone.
Despite raising HDL-C 72% and lowering LDL-C 25%, the trial was terminated early due to an increase in death and cardiac events.
The failure of Torcetrapib was attributed to “off-target” effects, namely an average increase in systolic blood pressure of 5.4 mmHg.This may seem small to those not familiar with clinical trials, but as an average, this is considerable.
Dalcetrapib: Futility (2012)
The dal-OUTCOMES (NEJM) trial was a large scale RCT of ~15,000 high risk secondary prevention (recent ACS) patients.
This study was terminated by the data and safety monitoring board due to futility. Despite raising HDL-C by 40%, there was no significant risk reduction.
Importantly, there was minimal change in LDL-C.
Futility was therefore felt to be due to inadequate increase in HDL-C and minimal lowering in LDL-C.
Evacetrapib: Trial too short, not enough ApoB reduction (May 2017)
A third attempt was undertaken with Evacetrapib in the ACCELERATE (NEJM), where it was felt that perhaps a CETP inhibitor that both raises HDL-C and lowers LDL-C might work?
Again, another large RCT of ~12,000 high risk patients failed to show a benefit.
This was a bit perplexing at the time, since there was a whopping 133% increase in HDL-C! And a 31% decrease in LDL-C.
Why did Evacetrapib fail, then?
HDL Functionality:
Epidemiologic studies showing benefit to high HDL-C were mostly observed in healthy subjects. Some felt that CETP inhibition in less healthy subjects (those with underlying cardiovascular disease) may produce dysfunctional HDL, but genetic studies of CETP loss-of-function does not seem to support this.
Mechanism of LDL-C reduction:
Even if HDL-C is thrown out, there was enough LDL-C lowering to see a 15% reduction in cardiovascular risk, as seen in statin trials of similar magnitude LDL-C lowering.
However the mechanism of LDL-C lowering may be different. With statins, and virtually all other LDL-C/ApoB lowering therapies, besides the newest ANGPTL3 pathway for with Homozygous Familial Hypercholesterolemia Treatment | EVKEEZA® (evinacumab-dgnb), the mechanism is through upregulation of LDL receptors and clearance of ApoB (which are predominantly LDL) particles from the plasma.
With Evacetrapib, CETP inhibition causes significant reduction in LDL-C, with a notable decrease in small dense LDL particles (60-70%), but not as significant a change in overall LDL particle count (22%), nor ApoB (20%).
The modest reduction in total LDL particle count, combined with a relatively short 2 year period of investigation, may not have been long enough to see a statistically significant change. Though, there was no separation in curves and no signal that a change was occurring, as seen in other lipid lowering trials.
Other potential off-target signals:
While much less, there was a 1.4 mmHg average increase in systolic blood pressure.
CRP was also noted to be slightly elevated, which may signal some sort of inflammatory process.
Anacetrapib: Worked, but not worth it to pursue (September 2017)
The HPS3-TIMI55-REVEAL study with Anacetrapib was published soon after the Evacetrabip study. This was a very large RCT of ~30,000 patients with ASCVD who were already very well controlled, with mean LDL-C of 61 mg/dL, nonHDL-C of 92 mmHg, HDL-C 40 mg/dL, studied over 4 years.
HDL-C increased by 43 mg/dL (relative difference of 104%) compared to placebo and nonHDL-C dropped 17% mg/dL.
There was a small (11%), but statistically significant reduction in cardiovascular outcomes and consistent with the expected risk reduction one should observe with a 17% lowering of nonHDL-C, when compared with other statin trials (10%).
Thus, the positive results were felt to be due to anacetrapib’s ability to lower LDL-C, rather than any other mechanism. But in a population of patients already very well managed, with low baseline LDL-C, the impact was small.
Other observations from this trial were a lower incidence of new onset diabetes.
Despite positive results, its manufacturer, Merck, opted not to seek regulatory approval due to weak evidence.
So, What Has the CETP Inhibitor Story Reinforced?
HDL is a complex lipoprotein, involved in many functions besides reverse cholesterol transport
HDL-C, as a biomarker, does not tell us how HDL is functioning
Favorable changes in lipid biomarkers by a drug, may be offset by deleterious off target effects (just like the case with Niacin)
Intensity matters (not all drugs in a class are equivalent)
Mechanisms matter (the only benefit thus far appears to be seen with its ability to upregulate LDL receptors and thus plasma clearance)
LDL biomarkers matter [lowering LDL-C without commensurate lowering of total ApoB/total LDL-P (discordance in biomarkers) may not produce the intended clinical result]
Persistence pays off? (Are we getting closer to an effective agent? See below)
Obicetrapib (In Phase 3 development)
Just when we thought the CETP class was dead, a couple of brilliant lipidologists, Dr. John Kastelein (@JohnKastelein) / X and Dr. Michael H. Davidson - UChicago Medicine, resurrected another CETP inhibitor off the cutting room floor. Abandoned by a large pharmaceutical company, these 2 believe they may have found an agent which might finally crack the code and founded a small pharmaceutical company called NewAmsterdam Pharma Corporation (@NewAmsPharma) / X.
Obicetrapib has shown positive results in Phase 3 clinical trials:
BROOKLYN: The first of 4 trials, looking at LDL-C reduction in those with heterozygous familial hypercholesterolemia (HeFH) not at goal on statin therapy.
Achieved its primary endpoint, lowering LDL-C 41%.
BROADWAY: Evaluated patients with ASCVD or HeFH, not at goal on maximum existing lipid lowering therapies.
21% reduction in MACE at 1 year (exploratory)
Lowers Lp(a)
Lowers small LDL-P
Improves glycemic metrics
Perhaps the reduction in cardiovascular risk goes beyond LDL
PREVAIL: Ongoing large scale RCT looking at Major Adverse Cardiovascular Events (MACE) outcomes in those with ASCVD (secondary prevention) on top of maximum tolerated lipid lowering therapy.
For more information: Learn About Obicetrapib From NewAmsterdam Pharma
