How Low Can You Go? Concerns Over Aggressive LDL Lowering Eased as PCSK9 Evidence Mounts

When the FOURIER trial was presented at the American College of Cardiology 2017 Scientific Session in March, some physicians said they’d hoped to see evolocumab (Repatha, Amgen) make a bigger dint in cardiovascular events. Others said the absolute difference with placebo—1.5%—would have an important effect on clinical outcomes.

Universally, however, everyone digesting the first large clinical outcomes trial for a PCSK9 inhibitor had their eyes on a different set of numbers: the massive reduction in LDL cholesterol achieved with this new class of drug.

For proponents of the premise that “lower is better,” the study reinforced a concept that began in earnest with the PROVE-IT trial, a 2004 study showing that treating ACS patients to lower LDL cholesterol levels improved clinical outcomes. Still, FOURIER didn’t so much as nudge the LDL threshold downward as blast a hole in floor, providing support for LDL cholesterol levels at depths previous generations might have considered unsafe.

In FOURIER, for example, the median LDL cholesterol achieved with treatment was 30 mg/dL. That number doesn’t tell the full story, however. In the trial, treatment with evolocumab reduced LDL cholesterol levels to 25 mg/dL or lower in 42% of the 13,784 patients who received the drug. Approximately one in four patients had an LDL cholesterol level of less than 20 mg/dL.

Not to be outdone, similar reductions in LDL cholesterol have also been observed with alirocumab (Praluent, Sanofi/Regeneron), although the clinical outcomes data is still not yet available. In pooled data from multiple clinical trials, researchers recently showed that among 3,440 patients treated with alirocumab, 25% achieved an LDL cholesterol level of less than 25 mg/dL and approximately 10% went lower than 15 mg/dL.

All of which raises the question: how low is too low?

Steven Nissen, MD (Cleveland Clinic, OH), for one, is not too concerned. “Personally, I’m quite comfortable when the LDL is very low,” he told TCTMD. “And FOURIER, of course, is very, very big. If there was something bad that was going to happen when you got to a really low LDL cholesterol, we would have seen it.”

In the GLAGOV trial, an intravascular ultrasound study led by Nissen and Stephen Nicholls, MBBS (University of Adelaide, Australia), patients with coronary artery disease who received evolocumab for 18 months also had dramatic reductions in LDL cholesterol levels, down from 93 mg/dL at baseline to 36.6 mg/dL. Some patients, such as those who started with an LDL cholesterol less than 70 mg/dL, achieved on-treatment LDL cholesterol levels as low as 24 mg/dL.

There’s never been any evidence from any clinical trials showing a hazard of very low LDL cholesterol levels. Steven Nissen

In GLAGOV, approximately two-thirds of treated patients had a regression of coronary atherosclerosis. Based on studies like FOURIER, and to a lesser extent GLAGOV, Nissen said, “it’s hard to argue that we should reduce the dose simply because the LDL is very low.”

Seth Martin, MD (Johns Hopkin University Medical Center, Baltimore, MD), is also reassured by the accumulating trials. “There is reasonable concern about safety, that we need to study this over the long-term, but there are a number of different lines of evidence to suggest it should be safe,” he said. “In terms of numbers, I have heard colleagues say that the goal should be less than 50 mg/dL instead of less than 70 mg/dL, or that we should be targeting between 30 and 50 mg/dL. Right before FOURIER, I was getting lots of emails from colleagues that had patients treated with lipid-lowering therapies in that range, down around 30 mg/dL, asking if they were too low.”

ACC/AHA Guidance Left Wiggle Room for Nervous Docs

In 2013, the American College of Cardiology/American Heart Association (ACC/AHA) published new guidelines for the treatment of elevated cholesterol. Unlike the older recommendations, the current advice abandons LDL treatment targets and instead recommend either a moderate- or high-intensity statin based on the patient’s 10-year risk of cardiovascular events. In those with atherosclerotic cardiovascular disease, a high-intensity statin—either rosuvastatin 20-40 mg or atorvastatin 40-80 mg—is recommended.

The ACC/AHA guidelines, however, also stipulate that in cases where on-treatment LDL cholesterol levels drop below 40 mg/dL, physicians could back off treatment by decreasing the dose.

Martin noted that the ACC/AHA stipulation has put a little bit of fear into some practicing physicians who have treated patients with PCSK9 inhibitors and said it’s one of the reasons why there are questions about the potential safety of going this low. Still, many who have used evolocumab and alirocumab are not seeing any red flags.

“The thinking shifted for most folks treating patients with PCSK9 inhibitors and seeing that patients were doing just fine,” said Martin. “Clearly we need to keep studying it. There’s been a lot of concern, a concern that’s also reflected in the protocols of clinical trials.” In the major statin trials, he noted, when LDL cholesterol have been reduced to very low levels, there’s typically been a dose adjustment. Even in the ODYSSEY clinical trial program with alirocumab, the goal was to keep LDL cholesterol levels around 25 to 50 mg/dL.

Of course, we’ve only had 2 years of follow-up, and I think that means we have to be honest with the patient. Sekar Kathiresan

Sekar Kathiresan, MD (Massachusetts General Hospital, Boston, MA), is also a proponent of “lower is better,” telling TCTMD that FOURIER should force physicians to work hard at reducing LDL cholesterol to low levels in patients who have atherosclerotic cardiovascular disease. That said, there are caveats to his support.

“I would like to see more information on the patients who got to very low LDL levels,” said Kathiresan. “But from all indications—half the patients in the trial got to less than 30 mg/dL, there were lots of patients who got to 10 mg/dL, 20 mg/dL—there doesn’t seem to be a problem. Of course, we’ve only had 2 years of follow-up, and I think that means we have to be honest with the patient. We can say, look, at 2 years, there wasn’t a problem but it doesn’t mean that with more prolonged follow-up something might show up. But I’m not alarmed.”

EBBINGHAUS and OSLER: No Neurocognitive Niggles

In 2012, the US Food and Drug Administration (FDA) issued a controversial safety communication and updated the label for statins, warning of the potential for “generally nonserious and reversible cognitive side effects” such as memory loss or confusion.

Two years later, the FDA asked Sanofi/Regeneron to monitor for adverse neurocognitive events in their long-term phase III clinical trials testing alirocumab, including ODYSSEY Outcomes. The recommendation to monitor for cognitive changes emerged with a signal showing numerically higher number of cognitive adverse events among evolocumab- and alirocumab-treated patients in case series and different studies, including ODYSSEY Long-Term and OSLER, respectively.

Fueled by concerns and the potential for harm, Amgen sponsored the EBBINGHAUS trial, a substudy of FOURIER that included 1,204 patients who underwent baseline and follow-up cognitive testing. As reported by TCTMD, the EBBINGHAUS researchers observed no differences between evolocumab- and placebo-treated patients in terms of performance on a battery of neuropsychological tests, and they saw no impairment of memory and executive function among patients who achieved LDL cholesterol levels less than 25 mg/dL with evolocumab.

Alirocumab has also been scrutinized for neurocognitive side effects. In a pooled analysis of the 3,240 patients who received alirocumab as part of the ODYSSEY development program, PCSK9-inhibitor treatment did not increase the risk of neurocognitive adverse events, even among patients who achieved LDL targets of less than 15 mg/dL. The pooled analysis included 14 studies, ranging in duration from 8 to 104 weeks.

James Underberg, MD (NYU Langone Medical Center, New York, NY), said the cumulative data show that very low LDL cholesterol levels do not result in cognitive decline.

“I wouldn’t say that it’s OK to give somebody a medication to lower their LDL cholesterol to single digits for 10 years—we just don’t have that safety data—but if someone requires a low LDL for cardiovascular risk reduction, I think the data supports that it’s OK to do if you have a good reason to do it,” he said. “I always tell people, ‘What I know is based on the duration of the data,’ and we know that over the time the patients were studied, it seems to be safe, which is why we always have to pay attention to accumulating data.”

Underberg stressed that PCSK9 inhibition, and the resulting reduction in LDL cholesterol, is reserved strictly for patients who are at high risk for recurrent events. “This is not a drug in all comers for primary prevention. I think that’s something to always remember.” He added that there are several lines of evidence suggesting individuals don’t need a lot of LDL cholesterol and low values are scarier in theory than in practice. Nissen made a similar argument, suggesting LDL cholesterol may simply be the “garbage” of lipid metabolism.

I always tell people, ‘What I know is based on the duration of the data.’ James Underberg

“There’s never been any evidence from any clinical trials showing a hazard of very low LDL cholesterol levels,” said Nissen. “We’ve seen people with single digits. This goes back to the high-dose rosuvastatin and atorvastatin trials, with all comers getting the top dose of a statin, and some of them had very low LDL cholesterol levels.”

Martin agreed, saying it’s possible that some patients might achieve LDL cholesterol levels close to zero and it doesn’t appear to pose a problem. He pointed out, though, that the Friedwald equation—the standard method for estimating LDL cholesterol levels—is not accurate in patients with very low levels (or high triglycerides). In 2013, Martin and colleagues published a study showing that nearly 25% of individuals with LDL cholesterol levels of less than 70 mg/dL had higher levels when measured directly with ultracentrifugation.

“When we talk about going too low, part of it might be that you’re not actually that low if you knew what the LDL actually was,” said Martin.

Diabetes and Cataracts

In addition to the neurocognitive concerns, there have been worries about the potential development of diabetes and cataracts with very low LDL cholesterol levels. Neither of these adverse effects, however, were observed in the FOURIER trial, with adjudicated cases similar between patients who received evolocumab or placebo.

Similarly, in the pooled Odyssey analysis evaluating alirocumab there was also no increased risk of new-onset diabetes, although the risk of cataracts was higher among patients who achieved LDL cholesterol levels less than 25 mg/dL compared with those with higher LDL levels (2.0% vs 0.6%; P = 0.0018).

Brendan Everett, MD (Brigham and Women’s Hospital, Boston, MA), who wrote an editorial accompanying the alirocumab analysis, told TCTMD that FOURIER largely assuaged concerns about safety. Like others, though, he stressed the trials were short duration for a treatment intended for lifetime use. He pointed out that the trial was careful in ascertaining the occurrence of new-onset diabetes (8.1% with evolocumab vs 7.7% with placebo; P = NS) and that EBBINGHAUS reassured him about the risk of neurocognitive events.

That said, while the size of FOURIER provides some confidence in the ability to detect adverse events, “you never know what might be detected when you move from a clinical trial population to the general population and treat patients for an extended period of time,” said Everett. “Once you leave a placebo-controlled trial population, adverse effects become more difficult to measure.”

Regarding the risk of cataracts with LDL lowering, Everett noted that the HOPE-3 study also showed a higher risk of cataract surgery amongst rosuvastatin-treated patients compared with placebo (3.8 vs 3.1%; P = 0.02).

The thinking shifted for most folks treating patients with PCSK9 inhibitors and seeing that patients were doing just fine. Seth Martin

“My general sense is that cataract is a real adverse effect that is seen with LDL cholesterol lowering regardless of mechanism,” said Everett. “How important it is probably depends upon the patient, in the sense of how the patient weighs his or her own health issues. Some are willing to take it no matter what the risks are because they think the benefit is so important. For others, you might talk to them about the benefit of LDL reduction until you’re blue in the face, but the possibility of an adverse effect looms so large for that particular patient that they elect not to go on therapy.”

To TCTMD, Kathiresan noted it took some time for the potential risk of adverse effects with statins to become clearer, so physicians and researchers need to stay vigilant. “We all have to be humble, in the sense that it took 25 years for statins being in practice before people realized there was a small increased risk of diabetes,” he said. “But with the PCSK9 inhibitors, there clearly isn’t anything dramatic going on in the first 2 years.”

Everett, for his part, said he has become much more comfortable with very low LDL cholesterol levels, pointing out that high-risk secondary prevention patients can derive real benefit in terms of lowering their risk of cardiovascular death/MI/stroke with further LDL lowering. This includes secondary prevention patients with already low LDL cholesterol. “Previously you might not have considered it, but now you have to think about [treatment] in patients with an LDL of 70 mg/dL or even the sixties,” he said.

Underberg said that the clinical question of importance is not whether patients should be treated to a specific low value, but rather when to intervene with a second agent, such as ezetimibe (Zetia, Merck/Schering-Plough) or a PCSK9 inhibitor, in someone already taking a statin. Some patients might be adequately controlled with an LDL cholesterol level around 70 mg/dL, while other secondary-prevention patients will continue to have events.

“The next step at that point would be to add a PCSK9 inhibitor and their LDL cholesterol will go down to a very low level,” said Underberg. “But I’m not sure that where that number ends up is as important as that you now have them on the PCSK9 inhibitor and further lowered their LDL cholesterol.”