A Phase 2b randomized clinical trial of the oral p38α inhibitor neflamapimod in dementia with Lewy bodies (DLB)

Imagine your brain is a bustling city. In this city, there are different types of traffic jams. One famous kind is Alzheimer's, where the roads get clogged with sticky, gooey trash (amyloid plaques) and tangled wires (tau tangles). But there's another, less famous traffic jam called Dementia with Lewy Bodies (DLB). In DLB, the problem isn't just the trash; it's that the city's power grid—the part that keeps the lights on and the traffic moving smoothly—starts to flicker and fail. This power grid is the "cholinergic system," a network of nerve cells that helps us pay attention, remember things, and move our bodies. When this system sputters, people with DLB experience confusion, hallucinations, and shaky movements. Scientists have been trying to build a "power surge" to fix this flickering grid, but so far, the tools they've tried only give a tiny, temporary spark. They need a better way to keep the lights on.

Enter a new tool called neflamapimod. Think of it as a specialized mechanic designed to repair the specific wires in the DLB power grid. In a smaller, earlier test, this mechanic seemed to work wonders, but only if the city didn't also have the sticky Alzheimer's trash clogging the streets. The big question was: If we find people who have only the DLB flickering (and no Alzheimer's trash), and we give them the mechanic at the right strength, can we fix the grid? A team of researchers set out to answer this with a massive, 16-week experiment involving 159 people. They wanted to see if the mechanic could stop the city from getting darker and more chaotic.

The story of this experiment, however, is a bit like a race where the runners started with the wrong shoes and the map was slightly blurry. The researchers recruited 159 participants with probable DLB. They tried to be very careful, screening out anyone who might have that sticky Alzheimer's trash in their system by checking a blood marker called plasma p-tau181. They set a rule: if the marker was too high (≥27.2 pg/mL), the person couldn't join. They hoped this would create a group of "pure" DLB patients, the ones most likely to benefit from the repair. They split the group in two: one half got the real mechanic (neflamapimod, 40mg three times a day), and the other half got a fake pill (placebo).

When the 16 weeks were up, the main result was a bit of a letdown. The researchers measured how much the participants' dementia severity changed using a score called the CDR-SB. In the big picture, the group taking the real mechanic did not do significantly better than the group taking the fake pill. The "power grid" didn't get a clear, measurable boost. The same was true for other tests, like how fast people could stand up and walk (the Timed Up and Go test) or how they felt about their overall condition. Even blood tests for general brain inflammation didn't show a difference between the two groups.

But, the story doesn't end with a flat "it didn't work." The researchers started digging into why it didn't work, and they found two sneaky culprits.

First, the map was blurry. The test they used to screen out the Alzheimer's trash had changed versions. The researchers thought they were excluding people with high trash levels, but because of a mix-up between the old and new test versions, they accidentally let in more people with Alzheimer's trash than they planned. It turns out that the mechanic only works well on the "pure" DLB grid, not on a grid that is also clogged with Alzheimer's trash. When the researchers looked back at the data and separated the "pure" DLB patients (those with very low trash levels, specifically below 21 pg/mL), the story changed. In this smaller, cleaner group, the mechanic did seem to help. The people taking the real drug got slightly better, while the placebo group got slightly worse. It was a hint that the mechanic might actually work, but only if you are very strict about who you let into the city.

Second, the runners had the wrong shoes. The researchers checked how much of the drug actually made it into the blood. They expected the "shoes" (the pill capsules) to deliver a specific amount of medicine, but the average amount in the blood was lower than expected. It was like trying to run a marathon with shoes that were a size too small; the runner (the drug) just couldn't get enough speed to finish the race. When they looked at the people who did get a higher dose of the drug in their blood, those people showed more improvement. This suggested that the specific batch of pills used in the main part of the study wasn't strong enough to do the job.

There was one other interesting clue. A small group of participants got brain scans (MRI) to look at the "power grid" itself. In this tiny group, the people taking the drug actually saw their brain's power grid (the basal forebrain) get slightly bigger or stop shrinking, while the placebo group's grid shrank. This matched what scientists hoped the drug would do: protect and repair the specific nerve cells that were failing.

So, what's the verdict? The main headline is that the big, 16-week test didn't prove the drug works for everyone. The primary goal was missed. However, the researchers are not throwing in the towel. They believe the failure wasn't because the drug is bad, but because the experiment had two glitches: too many people with the wrong type of dementia (Alzheimer's co-pathology) were included, and the pills used weren't strong enough to get the right dose into the blood.

The paper suggests that if they try again with a stricter rule to keep out the Alzheimer's trash and use a new, stronger version of the pill that guarantees the right dose, they might finally see the "power surge" they are looking for. It's a story of a promising mechanic who was sent to the wrong neighborhood with the wrong tools, but who might just be the hero this specific city needs if they get the logistics right next time.