Summary
A medRxiv preprint reports that laromestrocel reduced measures of brain inflammation and atrophy in 49 people with mild Alzheimer’s disease. The findings come from an early clinical study and have not yet been published as a peer-reviewed journal article.
A medRxiv preprint reports that laromestrocel, a mesenchymal stem cell therapy, was associated with lower brain inflammation, less brain atrophy and better clinical scores in people with mild Alzheimer’s disease. The analysis included 49 participants and compared two monthly dosing regimens with placebo over 39 weeks.
The findings, posted on September 17, 2026, come from the CLEAR MIND study, registered as NCT05233774. The report is a preprint, so its results have not yet been presented as a peer-reviewed journal article.
What the study reported
The strongest effects were reported in brain regions commonly affected by Alzheimer’s disease. In the left hippocampus, a region important for memory, measures of inflammation were lower in both the 25-million-cell dose given four times and the 100-million-cell dose given four times compared with placebo. The reported p-value was below 0.001 for both groups; the relevant analyses included 11 participants in the 25-million-cell group and 10 in the 100-million-cell group. The placebo comparison included nine participants.
The 25-million-cell regimen was also associated with lower inflammation in the left temporal cortex, with p=0.007, and in the left parietal cortex at week 26, with p=0.042. Both treatment groups showed an effect in the medulla, with p=0.007 for the 25-million-cell group and p=0.007 for the 100-million-cell group.
The authors report that reduced neuroinflammation was associated with reduced brain atrophy in the hippocampus. Across 40 participants, the correlation was R=-0.326 with p=0.040. Reduced inflammation was also associated with better clinical scores, particularly cognitive function measured using the Montreal Cognitive Assessment, or MoCA. That correlation was R=-0.401 with p=0.011 among the same 40 participants.
A blood analysis found lower plasma neurogranin in the 25-million-cell group, with p=0.023 and N=11. Neurogranin is a protein used in research as an indicator related to synaptic injury; the authors interpret the reduction as consistent with less synaptic loss. Additional blood-biomarker results supported what the paper describes as a multimodal mechanism involving peripheral immune activity.
Why the mechanism matters
Alzheimer’s disease involves progressive loss of brain tissue and cognitive function. Neuroinflammation refers to immune activity in the brain and surrounding nervous system. Persistent inflammatory signaling can interact with processes involved in neuronal and synaptic injury, making it a target for experimental treatments.
Laromestrocel is being investigated as a way to alter that inflammatory environment rather than simply treating a symptom such as memory impairment. The pattern reported in the preprint—regional changes in brain inflammation alongside measures of atrophy, cognition and blood biomarkers—supports the authors’ proposed mechanism. The associations are clinically relevant because they connect a biological process with structural and functional measures in the same study.
The evidence is still early. This is a preprint based on a relatively small clinical study, and the supplied report does not provide the full trial design or detailed endpoint methods in its abstract. The results therefore need confirmation in larger, peer-reviewed studies with clearly described methods and longer follow-up. The report also includes several Longeveron-affiliated authors; its conflict-of-interest statement identifies Joshua M. Hare as Longeveron’s chief scientific officer, a compensated consultant and board member, and an equity holder. The study was funded by the Alzheimer’s Association. The record describes clinical research but does not report regulatory approval for laromestrocel as an Alzheimer’s treatment.