A Natural Human Experiment: PCSK9 and Cholesterol
Seth Bilazarian, MD: This is Seth Bilazarian with theheart.org on Medscape here at the American College of Cardiology meeting in San Diego with an opportunity to sit down with Dr Mark Sabatine, chair of the TIMI Study Group and senior physician at the Brigham & Women's Hospital. Dr Sabatine has just presented the OSLER study[1,2] data—a group of trials that have been pulled together to provide data on the [proprotein convertase subtilisin-kexin type 9] PCSK9 inhibitor evolocumab (Repatha, Amgen). For the purposes of disclosure, I am an investigator on the FOURIER trial and a former OSLER house officer, to the extent that that may represent a conflict.
I'm excited to sit down with you, who taught me a lot through a variety of channels about PCSK9 from discovering a genetic mutation—a natural human experiment—to potentially a drug therapy, which we are very close to. Give us a brief review on that.
Marc S Sabatine, MD: It is an amazing story. It's one of the great stories in genetics and cardiovascular disease. A little over a decade ago, they found a family that looked like they had familial hypercholesterolemia. They had high [low-density-lipoprotein] LDL levels and an increased risk for myocardial infarction [MI], and they looked to see whether they had a mutation in the LDL receptor, which is the most common cause for familial hypercholesterolemia, but they didn't. They found a mutation in a new locus for this gene—PCSK9. No one knew exactly what it did.
They did some very elegant biology, in which they figured out that PCSK9 was a protein that chaperones the LDL receptor to destruction. And the LDL receptor on the surface of the hepatocyte binds the LDL and brings it into the liver. It then has one of two fates. It can release the LDL and then go back to the surface of the liver cell and take more LDL cholesterol out of the circulation, or it could be destroyed. And PCSK9 favors the latter.
So then people thought, if we have found individuals with a gain-of-function mutation, who have too much PCSK9 activity and therefore have fewer LDL receptors, more LDL in their circulation, and increased risk for MI, why don't we start looking for the opposite? Can we find people who have loss-of-function mutations? Jonathan Cohen and Helen Hobbs found those individuals and found that they had lower levels of LDL and a decreased risk for MI. At that point, it became clear that this would, potentially, be a great drug target. Now, several companies are working to develop inhibitors that are in various stages of development.
Dr Bilazarian: The PCSK9 is opposite of LDL, in simple terms—the higher the PCSK9, the lower the LDL; the lower the PCSK9, the higher the LDL. By inhibiting PCSK9, we can drive LDL-C levels down.
Dr Sabatine: The PCSK9 is basically the opposite of the LDL receptor. If you have high levels of PCSK9 and low levels of the LDL receptor, you will have higher levels of LDL cholesterol in the circulation. If you inhibit PCSK9, you have more LDL receptors and therefore less circulating LDL.
Monoclonal Antibodies: New to Cardiology
Dr Bilazarian: There are many ways to attack it that are being developed. The ways that seem closest to approval are the monoclonal antibodies. Cardiologists don't know a lot about these. These drugs are being used for autoimmune therapy or in oncology. There are 27 that have been approved in the United States. I didn't know that until I prepared for this talk.
Monoclonal antibodies are outside of our usual comfort zone. The only thing that we have a familiarity with as interventional cardiologists is the drug ReoPro (abciximab, Eli Lilly)). The "–mab" suffix is what distinguishes this group. These monoclonal antibodies have been targeted as a strategy to try to address this problem. Could you tell us what we know about monoclonal antibodies?
Dr Sabatine: In cardiology, we don't use monoclonal antibodies very frequently, but they are certainly used very frequently in other fields, in gastroenterology, rheumatology, and oncology. The good news for the monoclonal antibodies is that they are very specific, like the antibodies in our bodies. So they target the protein of interest. They need to be given as subcutaneous injections in this case. They should have minimal adverse effects and they are not dependent on renal or liver function, so it makes therapy simpler.
Dr Bilazarian: These monoclonal antibodies are in development. You presented the OSLER study today, and the OSLER study composite has shown the dramatic LDL-C lowering that we know these PCSK9 inhibitors are capable of. They dramatically lower LDL-C levels, but the study that you presented today was the first to show a clinical end point benefit—a composite end point of death and myocardial infarction was reduced in this trial. Can you comment on that for us?
Dr Sabatine: Vis à vis evolocumab, there have been a series of so-called "parent studies," which were phase 2 and phase 3 trials that were done for dose ranging and dose confirmation in a variety of populations—low-risk patients in whom to test the drug as monotherapy, patients on statins, patients with familial hypercholesterolemia, and patients who were statin-intolerant. Those were typically short-term (12-week) studies, as are most lipid trials.
After completing one of those parent studies, patients could then enroll into the OSLER extension study, where they were rerandomized to receive either evolocumab or standard of care. We followed them for about 1 year. We enrolled 4465 patients. This is a good number of patients, not followed just for 12 weeks, but for a year. That gave us accrual of enough cardiovascular outcomes that we could start to look at differences.
Cutting Cardiac Events in Half
Dr Bilazarian: You showed a more than 1% reduction in major adverse cardiac events at 1 year. The curves seem to begin to diverge around 90 days. As a clinician, it's just amazing to me. I was a participant in a study run by TIMI called IMPROVE-IT that went on for a very long time. Does a very small LDL-C lowering effect take longer to show a benefit, and does a dramatic LDL-C lowering effect show an earlier benefit? Is that a way to think about this?
Dr Sabatine: That's a great observation. Certainly the clinical benefit appears to be linked to the absolute LDL-C reduction, and that is why these agents are so promising, because they can reduce LDL cholesterol by about 60%. In OSLER, we reduced LDL cholesterol 61%. That was an absolute difference of 73 mg/dL. That is on the order of what we have seen with simvastatin in the 4S trial, where we had patients with an MI, and the question we were asking back then was if their LDL-C levels were close to 190 mg/dL, should we bother to reduce it further?
You make the correct observation in terms of the difference in event rates. It was reduced from about 2% to 1%. Keep in mind that this was done in all-comers. It included a variety of different populations. We saw the same halving of the event rates when we looked at the subset of patients who had known vascular disease. Those were higher-risk patients. In that group, the event rate went from about 5% to 2.5%.
In terms of the time at which the curves diverged, it was after 2 to 3 months. We have seen different times to divergence in different lipid-lowering trials. Some have diverged for 1 to 2 years. Others—PROVE IT-TIMI 22, the MIRACL trial, and TNT—tended to diverge quite early in the course. I agree with you that in this case, the reason that the curves diverged so early was because we achieved a very large reduction in LDL-C and got down to very low levels.
Getting the Drugs to the Pharmacy
Dr Bilazarian: It is incredibly exciting that we are going to have a very powerful tool that will achieve a dramatic reduction in events for our patients. I may not have to get up for a STEMI call in the middle of the night. This is a dramatic change. This may be the last chance I have to sit with you, an expert, before these drugs are released, because three drugs are now in phase 3 trials, and two of these drugs have [Prescription Drug User Fee Act] PDUFA dates at the US Food and Drug Administration [FDA]. As a reminder for those who may not be familiar with the regulatory terms, that means that the FDA has to give an answer for Praluent [alirocumab, Sanofi/Regeneron] on July 24. So we will have an answer—yes or no, whether that drug is available to go to the pharmacy. One month later, on August 27, we have Repatha. So by Labor Day, we may have two new drugs and whole new class of drugs that can achieve dramatic LDL lowering.
I would love to have your insights about these three trials—ODYSSEY Outcomes, PROFICIO, and SPIRE —which are very similar. They are enormous trials, with 20,000 patients. Thus far, they all seem to show very good safety. They all seem to show something that surprised me, as a clinician. The injections are very well tolerated, and the patients seem to prefer them to taking a pill every day. There seem to be no muscle effects that are not attributable to a coincident statin. Is that a pretty good summary of where we are?
Dr Sabatine: For the two compounds that are furthest along—evolocumab and alirocumab—you are right. The FDA will be commenting on them. We expect it at the end of the summer and so we may very well have these drugs available to us as clinicians soon thereafter. In parallel, there are dedicated cardiovascular-outcomes trials. For evolocumab, it is the FOURIER trial. PROFICIO is the collection of the phase 2 and phase 3 trials that have been done to date. Then there are ODYSSEY Outcomes and the SPIRE trials.
We don't expect those trials to report until 2017, so we are going to have a gap between when some of these drugs will be available and when we have the definitive cardiovascular-outcomes data.
Human vs Mouse Antibody: Does it Matter?
Dr Bilazarian: These are going to be the hard questions. What will I need to know after Labor Day weekend with a patient who has a very high LDL-C level despite maximally tolerated statin therapy? How can I counsel this patient to make an informed decision? Admittedly, I'm not informed yet except in very general terms about monoclonal antibodies. What do I need to know about humanized vs chimeric vs mouse? The ReoPro drug that I mentioned earlier was a chimeric monoclonal antibody. Are these going to be class effects, or is there something else that we should be worried about? We are not going to have much long-term data, so should we just be very cautious about using them?
Dr Sabatine: These are great questions. There are some subtle differences in the antibodies. Evolocumab and alirocumab are fully human antibodies. Bococizumab is not fully human. In theory, that could lead to differences for neutralizing antibodies, but we will need to see when those data emerge. I have no idea what the FDA is going to say and typically, for lipid-lowering drugs, they approve the drug for lipid lowering and then it falls on the clinicians to decide in which patients we want to use that drug.
We will need to risk-stratify our patients. Certain subgroups immediately come to mind, such as those patients who, despite being on maximally tolerated statin therapy, don't have LDL-C levels as low as you would like them to be. You will want to add another drug, and we can talk about targets in a minute.
Then there are patients who have familial hypercholesterolemia who have very high LDL-C levels. We know we can't get them into range and because they have had lifelong high LDL-C levels, they are at very high risk. That is another logical group. Then, there are the statin-intolerant patients who, despite our best attempts to get them on a statin dose, simply don't tolerate it. Those will be the three primary groups. The crux of the matter will be for patients on a statin. Are you or are you not happy with their LDL-C levels? This begs the question of our old friend, targets.
Where Do Targets Fit In?
Dr Bilazarian: Give me your thoughts. I drank the Kool-Aid on targets. I was pressing patients to get their LDL-C levels under 70 mg/dL for secondary prevention. I thought I was doing the right thing. We had a little hiccup with the current lipid-lowering guidelines. What should I do in September when a patient presents to me with an LDL-C of, say, 70 to 100 mg/dL or 100 to130 mg/dL for secondary prevention? Are these the go-to drugs?
Dr Sabatine: I miss the targets; it was some pretty good Kool-Aid. The US guidelines ask a very narrow question, focused on the data from the statin trials. They did a service in terms of advocating high-intensity statin therapy and that's good. But stepping back as a physician-scientist, there are some very compelling data that lower is better and that has appeared to be true across a whole range of trials, and now with different drugs.
Now we are seeing it with ezetimibe (Zetia, Merck/Schering-Plough) in the IMPROVE-IT trial. We are getting powerful hints of that with the PCSK9 inhibitors. In my mind, post-IMPROVE-IT, I want my patients with acute coronary syndrome (ACS) to be in the mid-50s, not 70 mg/dL. How will I get them there? I will start with a statin because they are generic and they are well-tolerated. There is an excellent evidence base. I might then add ezetimibe now that we have data from the IMPROVE-IT trial. That still leaves a lot of patients who aren't at goal, and we have seen with high-intensity statin, plus or minus ezetimibe, that anywhere from one-third to one-half of patients don't get to a goal that used to be 70 mg/dL.
It should probably be lower now for the post-ACS patients—down to the mid-50s. Keep in mind, from the cholesterol-treatment trial, that those individuals who had LDL-C levels at baseline of 80 mg/dL or less benefited just as much as those with higher LDL levels. In the JUPITER trial of relatively stable patients, there was the same risk reduction if the LDL started at less than 60 mg/dL, and that means going from less than 60 mg/dL to less than 30 mg/dL. Patients do benefit with having much lower levels, and now, we will have statins, we will have ezetimibe, and the PCSK9 inhibitors are a third very powerful tool to use, and I will use them when I need to get the patient's LDL level under control.
One, Two, or Three Drugs?
Dr Bilazarian: I agree with everything that you say, and I have very little to argue about except the concern, at the community-based physician level, about polypharmacy. Everything you say makes sense on a science level—statin, ezetimibe, PCSK9. But we potentially are using three drugs where one drug in this class seems to lower LDL-C by 50% to 70%. So for a patient whose LDL-C is 140 mg/dL, one drug could potentially do that. I'm not going to press you for an answer, but that will be the challenge for clinicians. I could give one drug, a once-a-month or a once-every-2-week injection, that will get the patient to a level of 50 mg/dL, without using the other two drugs—avoiding polypharmacy, avoiding statin myalgia, and potentially at a lower cost, if I'm permitted to do that by the pharmacy benefits manager.
Dr Sabatine: That is going to be the issue. They are probably going to be fairly keen on maximizing the statin first. But you raise a very interesting point. It doesn't appear to make a difference how we lower the LDL cholesterol, barring some particular side effect from the drug (and different drugs can lower LDL cholesterol through different means), but for both statins and PCSK9 inhibitors, the ultimate final common pathway is upregulation of the LDL receptor.
If that is true, then the question is, should you maximize the statin if that is going to cause side effects, or could you have a lower dose of statin and then couple it with an additional drug, either ezetimibe or a PCSK9 inhibitor? It certainly makes sense to use statins as the foundation, then add the other drugs as needed. That's how I envision that they are going to be used.
Dr Bilazarian: You showed today, as part of the OSLER data set, that patients with LDL-C levels less than 25 mg/dL seem to have no adverse effects. Trialists have been beating this drum for a long time with PROVE-IT and other trial data. Should we now believe that there is no lowest LDL-C level that we should worry about, that we should not be thinking about LDL-C as a healthy cholesterol but as dangerous remnant, and we should go beyond any concerns about arms and legs falling off?
When I was a fellow in fellow's clinic, I remember the attendings worried that lovastatin might cause problems down the road because of its impact on cholesterol. Is that where we are now?
Dr Sabatine: Your description of LDL is right—it is not used for cholesterol synthesis. It is sort of an end-stage molecule. So far, it looks very good, that as we get low, we haven't seen any problems. There are theoretical concerns for certain issues, for certain hormones that need cholesterol as the backbone. But so far, none of that has emerged as an issue. For neurocognitive side effects, the FDA has issued a notice about that. It is not based on very strong data, and from all the formal studies in which patients have been randomized, it hasn't emerged as a problem.
That being said, we did see an imbalance with neurocognitive events being reported more frequently [in OSLER]. However, we didn't see any link between that and low LDL-C levels. So far, it looks good. If you think about hunter-gatherer populations that have no coronary disease, their LDL levels are in the 40-mg/dL range. For us, as a society, we need to keep in mind that even though 120 mg/dL might be the average LDL-C level in this country, that doesn't mean it's normal. It just means it's average, and that is a tricky concept.
We will certainly need to see in the large outcomes trials, not only efficacy but safety, because we will have a much larger database of individuals with low LDL-cholesterol levels followed for a long time.
Any Role in Primary Prevention?
Dr Bilazarian: We have been talking about secondary prevention. As a clinical and interventional cardiologist, I want to push the LDL-C to get it as low as possible. Could you comment on primary prevention? It is not a core part of my practice as a cardiologist—it falls on the primary-care physicians. But you were recently the senior author of a paper in Lancet trying to help us understand strategies that will help evaluate and identify patients who are at higher risk in primary prevention using genetic tools. Can you comment on how that might be valuable to clinicians, and how it might be applicable to the PCSK9 story?
Dr Sabatine: As we think about our guidelines, in some ways, when we think about risk stratification, it's a little bit short-sighted to say that we don't need to have aggressive goals until the patient has had a heart attack. Then, we really need to work on the LDL-C. We should work on the LDL-C levels of those individuals, but wouldn't it be great if we could prevent that first heart attack? When you start extending therapy out to broader populations, there are issues of side effects and the trade-off of benefit and risk, and cost comes into play.
For that reason, we thought, with the different risk-stratification tools, could we use genetics to help with that? We had a composite genetic risk score that we used that predicted the risk for major adverse cardiovascular events. Once we did that, we could predict a gradient of events, so that even if the relative risk reduction is the same with statin therapy, the absolute risk reduction goes up and therefore, the number needed to treat goes down.
We will start using genetics in addition to clinical data and maybe biochemical data to identify patients in whom we need to think about lifelong risk and start them on therapy earlier.
Dr Bilazarian: What a fantastic review. Thank you for joining me here to try to put PCSK9 inhibitors in perspective. I hope that you will meet with me at the American Heart Association meeting when we will have had the PCSK9 inhibitors for 3 or 4 months, and I will have a whole new set of questions to ask you about how to use them at that time. For now, we are expecting the birth of a new class of lipid-lowering agents, potentially, in the coming months and trying to evaluate how to use them and how to talk to our patients about them is a significant challenge.
© 2015 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: PCSK9 Inhibitors: What to Know if They Come to Market - Medscape - Apr 28, 2015.