WASHINGTON, DC — New research is weighing in on the "do statins impair cognition?" debate with a report of transient brain-activation changes on functional magnetic resonance imaging (fMRI) scans.
A substudy of the Effect of Statins on Muscle Performance (STOMP) study showed similar scores on several neuropsychological tests for statin-naïve patients (mean age 48 years) taking 80 mg/day of atorvastatin for 6 months and those taking matching placebo—echoing findings just released from the EBBINGHAUS trial of patients on the proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor evolocumab (Repatha, Amgen) plus statins vs placebo.
However, the STOMP analysis did find "small but significant" between-group differences in activation patterns in two brain areas. While on study treatment, the placebo group had greater activation in the right putamen/dorsal striatum during a verbal memory task than did the atorvastatin group. However, after the drug "washed out," the atorvastatin group had greater activation in that area.
While "on drug," the atorvastatin group had greater activation in the bilateral precuneus during a figural memory task vs the placebo group—but this pattern also reversed after the washout phase.
So what does this mean? "We aren't sure," admitted Dr Beth Taylor (University of Connecticut and Hartford Hospital) to heartwire from Medscape. "There was no convincing evidence of measurable verbal or nonverbal memory dysfunction due to statins."
And although they found small brain-pattern differences, "the clinical implications are unclear. We need larger and longer-term trials." Taylor presented the findings at the American College of Cardiology (ACC) 2017 Scientific Sessions.
"I love that you looked at fMRI data along with neurocognitive testing," session cochair Dr Douglas L Mann (Washington University School of Medicine, St Louis, MO) said after the presentation. "But as a cardiologist, I would question that the age group studied was relatively young. Most of the time, these people would not be on statins.
"A concern would be if there was a small biological effect of statins superimposed on a 70-year-old brain," said Mann, adding that plasticity in younger brains may be one factor in the differences seen.
First fMRI Investigation
The 2014 Statin Cognitive Safety Task Force stated that evidence was "weak to nonexistent" that statins are associated with cognitive side effects. Still, Taylor said, "Central nervous system complaints of statins are the second most frequently reported adverse effect."
"We're talking mild cognitive complaints: memory loss, fuzziness, and general reports that patients don't feel like they're performing as well as they used to. It plays a role in patients discontinuing or being noncompliant with medication."
She added that this is the first study to investigate direct effects of statins on the central nervous system using fMRI.
Most other studies have used only standard neuropsychological tests, "and those have very small effect sizes," said Taylor. "They're better for detecting things like stroke recovery or deficits as opposed to these mild cognitive complaints."
The original STOMP trial was designed to evaluate treatment-related muscle side effects in 420 healthy volunteers. The new analysis examined 150 participants who received high-dose atorvastatin (n=66, 59% women) or placebo (n=84, 43% women) for 6 months.
Measurements included the Hopkins Verbal Learning Test-Revised (HVLT-R) for auditory memory, the Brief Visuospatial Memory Test-Revised, the Stroop Color-Word and Trail-Making Tests, and the patient-reported Cognitive Failures Questionnaire (CFQ). In addition, fMRI was used to measure neuronal activation in 77 of the participants during an in-scanner verbal Steinberg working memory task and again during a visual figural memory task.
All of the tests were administered after 6 months on treatment and again 2 months after treatment was stopped (drug washout).
There were minimal treatment effects on the neuropsychological tests except for a few exceptions. HVLT-R scores on total and delayed recall improved significantly (P=0.01 and 0.04, respectively) at drug cessation, albeit in both groups.
Scores on the Stroop Color-Word increased at drug cessation for those receiving placebo (P<0.01), while scores on the 18-point Clock Test decreased (P=0.02). CFQ total scores and memory/distractibility/blunders/names scores did not differ significantly between the groups at any time.
As mentioned, however, fMRI showed "group-time interaction in bilateral paracentral lobule/precuneus during the encoding phase of the visual memory task," reported Taylor. Those receiving atorvastatin had a greater blood oxygen level–dependent (BOLD) response on treatment vs the placebo group (P=0.006) but less response during the drug washout phase (P=0.001).
During the Sternberg Task, the atorvastatin group had less BOLD response than the placebo group in the "right putamen extending into the right globus pallidus" during the treatment phase (P=0.005) and more BOLD response during washout (P<0.001).
"What's interesting about the putamen is that it's involved in things like sensory motor function and other rudimentary functions. But it's also very much affected by Parkinson's and Alzheimer's disease. And it has some implication in working memory," explained Taylor.
When the investigators teased out data on the older participants, "we didn't see any obvious age effects. But I'd caution that we were probably underpowered to look at age-specific outcomes."
When asked whether the study's findings will again bring up cognitive concerns from patients, she told heartwire that "in general, we found minimal differences between the treatment groups. And most regional activation patterns were not different in scanner test scores."
Regarding the small but different brain patterns, the most puzzling aspect was that the groups were different at both time points. "That makes the data hard to interpret. I think we have to conclude that there's no sign of obvious memory dysfunction with statins. And that should be really reassuring for patients, even when looking at fMRI," said Taylor.
Still, "the group-time interaction makes us wonder if this could be a mechanism underlying those infrequent case reports. It's very preliminary but does deserve further confirmation," she said.
More Research Needed?
"Congratulations on this deep dive," said panelist Dr Neil J Stone (Northwestern Medical Faculty Foundation, Chicago, IL) during the postpresentation question-and-answer session.
"The guidelines have looked at this subject and, based on a large meta-analysis, we couldn't find a clear-cut signal. And I think your data agree with our finding."
Stone added that cognitive impairment is common as patients age, especially for those older than 70 years. "As we begin to worry about our individual patients, we need to be mindful that cognitive difficulties become more frequent then."
He added that there also often other variables to take into account, including other medications used and ischemic and neuropsychiatric causes.
"I think your research says that we need to look further. But one concern I have is that when looking at lots of different types of domains, P values have to be adjusted. So you're looking and looking and looking and eventually you'll find something," said Mann.
"It's the old worry: what happens when you torture the data until it confesses?"
The STOMP parent and ancillary cognitive studies were funded by grants from the National Heart, Lung, and Blood Institute. Taylor has received research funding or grants from Regeneron. Mann and Stone have reported no relevant financial relationships.
Heartwire from Medscape © 2017
Cite this: 'Suggestion' of Brain Changes on Statins by Functional MRI in STOMP - Medscape - Mar 23, 2017.