$Cassava Sciences Inc(SAVA)$Independent researchers are expressing doubts about the purported benefits of an experimental treatment for Alzheimer’s from Cassava Sciences.

The company said Thursday that a preliminary analysis of a small clinical trial showed its drug, called simufilam, improved the cognition of patients with Alzheimer’s disease — a benefit that no other treatment for Alzheimer’s has ever shown.

Asked by STAT to examine the simufilam data, Alzheimer’ scientists not involved in Cassava’s study said the cognitive benefit claim was exaggerated and not supported by the design of the clinical trial.

The conclusions reached by Cassava about simufilam’s benefit for Alzheimer’s patients are “overblown, inappropriate, and naive,” said Rob Howard, an Alzheimer’s researcher and professor of psychiatry at University College London. “It’s the sort of presentation that one might expect an undergraduate to make.”

Lon Schneider, a physician and Alzheimer’s expert at the Keck School of Medicine of the University of Southern California, described Cassava’s study results as “uninterpretable.”

Instead of having an academic researcher present the simulfilam data at the Alzheimer’s Association International Conference, which is the norm for scientific conferences, Cassava used an employee — Lindsay Burns, senior vice president for neuroscience. Burns is also married to Cassava CEO Remi Barbier.

Cassava said 50 patients with Alzheimer’s treated with simufilam for nine months showed an average improvement of three points on a 70-point cognition scale known as ADAS-cog. To put the benefit of its drug in perspective, the company pointed to a previously published analysis showing a typical Alzheimer’s patient would lose more than four points on the ADAS-cog cognition scale over the same time period.

The simufilam data “suggests disease modification,” Nadav Friedmann, Cassava’s chief medical officer, said in a press release. “It appears the drug’s unique mechanism of action has potential to provide transformative treatment benefits following nine months of dosing.”

But Cassava conducted its clinical trial without a placebo control arm, meaning the company lacks any data to compare directly against simufilam. All of the patients know they’re receiving a drug that is designed to help them, which can artificially boost scores on cognitive tests like the ADAS-cog, particularly over a relatively short time period like nine months.

Data on 50 patients were presented Thursday, but Cassava has enrolled 150 patients into the study. The company said results from the remaining 100 patients aren’t mature enough to be disclosed yet. The company did not respond to a request for comment.

“These data are interesting, but nothing more at this point because without a placebo arm, there is a benefit to behavior just from being in the study,” said George Perry, a neuroscientist at the University of Texas at San Antonio. “The placebo effect is most noticeable early — nine months is probably borderline.”

Biogen’s controversial and recently approved drug Aduhelm showed a fractional slowing of cognitive decline over 18 months compared to a placebo in its only nominally positive Phase 3 clinical trial. No Alzheimer’s drug studied in large, placebo-controlled studies has ever shown the ability to improve cognitive function.

That placebo effect in the simulfilam study may already be waning. In addition to cognition, Cassava also assessed the Alzheimer’s patients for certain behaviors like anxiety, agitation and delusion. At nine months, the improvement in so-called “neuropsychiatric symptoms” fell compared to the same assessment done at six months.

“Because the trial was open label and uncontrolled, no convincing claims of efficacy can or should be made,” said Howard. A three-point improvement in the ADAS-cog scale for a study like this is “well within what could be attributed to a placebo effect.”

Likewise, Schneider cautioned against even believing Cassava’s statement characterizing the loss of cognition typically seen by the average patient with Alzheimer’s. That analysis, he said, comes with wide variances, meaning the “implausible improvement” of three points on the ADAS-cog scale for simulfilam is “not inconsistent” with a worsening of cognition.

“As we know with clinical measures, the numbers the company presents could easily reverse after the next 50 or 100 patients are added,” said Schneider.

On Reddit message boards, simulfilam is being promoted by “meme” traders as a cure for Alzheimer’s. Meanwhile, institutional investor ownership in the company is nearly zero. Cassava only pivoted to Alzheimer’s recently. For most of its existence under a different corporate name, Pain Therapeutics, the company failed four times to secure Food and Drug Administration approval for an opioid painkiller.

Cassava shares fell 24% to $103 on Thursday, but the stock is still up more than 3,000% over the past year. Cassava carries an astounding market value of more than $4 billion, floated almost entirely on the hype of an Alzheimer’s drug with very little supportive evidence.