FOURIER Surveys Reassure About Cognitive Safety of Evolocumab

Adding the PCSK9 inhibitor evolocumab (Repatha; Amgen) to statin therapy in patients with stable atherosclerotic cardiovascular disease does not have a detrimental impact on cognition, at least over the first 2 to 3 years, surveys conducted among FOURIER participants indicate.

Everyday Cognition (ECog) surveys completed at the end of the trial, after a median follow-up of 2.2 years, showed that the proportion of patients reporting cognitive decline did not differ between the placebo and evolocumab arms in terms of the total score (3.6% vs 3.7%; P = 0.62), memory (5.8% vs 6.0%; P =0.53), or executive function (3.6% vs 3.7%; P = 0.83). Nor were there any differences in various domains of executive function.

Of note, lead author Baris Gencer, MD (Brigham and Women’s Hospital, Boston, MA), and colleagues report, the percentage of patients with a decline in the total cognitive score at the end of the trial was not significantly different between those who achieved an LDL cholesterol level below 20 mg/dL within 4 weeks of starting treatment and those who continued to have levels of 100 mg/dL or greater (3.8% vs 4.5%; P = 0.57).

The findings are consistent with those from the EBBINGHAUS study, which involved serial assessments with the formal, objective, and computer-based Cambridge Neuropsychological Test Automated Battery (CANTAB) but consisted of just 1,204 FOURIER participants. This new analysis, published online ahead of the May 12, 2020, issue of the Journal of the American College of Cardiology, includes data on 22,655 trial participants.

Looking at the two studies, “there’s no signal of impaired cognition with PCSK9 inhibitors compared to placebo in this patient population,” senior author Robert Giugliano, MD (Brigham and Women’s Hospital), told TCTMD, acknowledging the relatively short follow-up.

Cognitive Concerns With Low LDL Cholesterol

Worries about cognitive problems associated with low LDL-cholesterol levels started long before the development of the PCSK9 inhibitors. Observational data had already shown a link between naturally low serum LDL cholesterol and poorer cognitive function when the US Food and Drug Administration, in 2012, issued a safety communication about the potential for “generally nonserious and reversible cognitive side effects” with statins. That was based on postmarketing adverse event reports, published observational data, and randomized trials.

Subsequent reviews and analyses did not find strong evidence to support that relationship. And as Gencer et al point out, “the 2014 National Lipid Association Task Force on Statin Safety concluded that statin therapy was not associated with adverse effects on cognition and did not recommend assessing cognitive function before initiating treatment with a statin.” A 2015 meta-analysis of randomized trials also failed to show an association between statins and cognitive impairment.

The PCSK9 inhibitors drive LDL-cholesterol levels much lower than can be achieved with statins alone, so there remains some question about whether the new agents have an impact on cognitive function.

In the FOURIER trial, patients with atherosclerotic CVD and LDL-cholesterol levels of at least 70 mg/dL or non-HDL-cholesterol levels of at least 100 mg/dL despite statin therapy, plus additional CV risk factors, were randomized to evolocumab or placebo. The main results showed that evolocumab significantly reduced composite CV events, including stroke, without a difference in clinical neurocognitive adverse events. Findings were similar for another PCSK9 inhibitor, alirocumab (Praluent: Sanofi/Regeneron), in the ODYSSEY OUTCOMES trial.

As part of the assessment of the cognitive safety of evolocumab in FOURIER, the investigators asked participants to complete a 23-item survey covering the memory and executive function domains of the ECog scale at the end of the trial; 82.2% ultimately filled it out. The survey required patients to compare everyday function at the end of the trial with how they were functioning at the beginning of the trial using the following scoring:

• 1: no change or improvement
• 2: occasionally worse
• 3: consistently a little worse
• 4: consistently much worse

The proportion of patients reporting cognitive decline, defined as a score of 2 or greater, was not significantly different between the two trial arms. Findings were consistent across various subgroups.

“These data confirm the neurocognitive safety of intensive LDL-cholesterol reduction with evolocumab while reducing recurrent CV events in high-risk patients, and suggest that very low achieved LDL-cholesterol levels may be safely targeted for high-risk patients,” the researchers conclude.

Reassurance, With Caveats

Jennifer Robinson, MD (University of Iowa, Iowa City), who wrote an accompanying editorial, told TCTMD that EBBINGHAUS provided some reassurance about the cognitive safety of PCSK9 inhibitors, but was limited by its small sample size (less than 5% of the total trial population) and the fact that follow-up was shorter than for the overall trial. This latest analysis provides further reassurance, said Robinson, an EBBINGHAUS investigator.

But, as she details in her editorial, there are three major caveats. First, the 18% of patients who did not complete the ECog survey had more risk factors for cognitive impairment than those who did, leaving open the question of whether these results can apply to higher-risk patients. Second, the participants were relatively young overall, with a mean age of 62.5. And third, most of the patients were included based on a history of MI (81%), with a minority having a history of nonhemorrhagic stroke (19%).

That said, “within the context of who the trial was performed in, it seems to be pretty safe over about a 3-year period,” Robinson commented.

Responding to some of Robinson’s critiques about the patients who took the ECog surveys, Giugliano acknowledged that the patients who took the survey were at lower risk. “Some of that is just the way the study is set up, which is to ask people at the end of the trial, so it’s a little bit of an unfair criticism, I think,” he said. “And also it’s not like we forced people to fill out this survey. It was voluntary, and you can imagine if you had some terrible event, cancer or a bad heart attack or a stroke, and now you’re debilitated, you’re living in a nursing home, it’s going to be tough to get those people to fill it out.” If, in fact, those patients had filled out the survey, he added, “the results could have looked favorable for the drug because the drug reduces stroke.”

As for whether the results would apply to higher-risk patients, Giugliano pointed out that even though the patient population overall was relatively young, there were roughly 2,000 patients 75 or older. And there also were substantial numbers of patients with, for instance, diabetes, prior cerebrovascular disease, or nonstroke neurological disease, with no indications that the findings differed in those subsets.

Robinson’s criticisms have some merits, Giugliano said. But, he added, “honestly, I think that the editorial was a little bit negative.”

Robinson said longer-term data, out to at least 5 years, would strengthen the cognitive safety message of evolocumab and Amgen is planning open-label follow-up of that length for EBBINGHAUS. “I think if there’s no signals there, that’s very reassuring,” she said, explaining that 5 years should be enough time to overcome any “cognitive reserve” that could be compensating for potential deficits.