Rapamycin was undetectable in spinal fluid in the first trial that looked

Rapamycin has an unusually good case for a brain trial. It extends lifespan in animals and acts on the pathologies underlying Alzheimer's disease and related dementias in animal models, yet its clinical application remains underexplored [s1]. The obvious first question is whether the drug gets into the human brain at all. Until recently, nobody had measured it.

A phase 1 trial published in Communications Medicine in May 2025 did. It was a single-site, open-label study in people aged 55 to 85 with mild cognitive impairment or early-stage dementia, defined by a Global Clinical Dementia Rating of 0.5 to 1, all of whom took 1 mg of rapamycin daily for eight weeks [s1]. The primary aim was central nervous system penetrance, assessed by measuring drug levels in cerebrospinal fluid before and after treatment [s1].

The answer was no

In ten participants, with a mean age of 74 (SD 4) and 60% women, rapamycin was not detectable in cerebrospinal fluid either before or after treatment [s1].

That is the finding, and it is a hard one for the hypothesis. A drug that cannot be measured in the compartment where it is supposed to act has, at minimum, a delivery problem. The investigators report it plainly and conclude that the biological effects and clinical impact of repurposing rapamycin for Alzheimer's need to be better understood [s1].

The biomarkers moved the wrong way

The secondary results are more uncomfortable still. Over eight weeks, twenty mostly mild adverse events occurred, systolic blood pressure and HbA1c increased, and multiple red blood cell parameters decreased [s1]. There were no significant cognitive changes [s1].

Several disease and inflammatory biomarkers rose significantly after FDR correction. In cerebrospinal fluid, phosphorylated tau-181 increased by a mean of 2.64 pg/ml (95% CI 0.70 to 4.59), glial fibrillary acidic protein by 6262.21 (3787.44 to 9373.84) and neurofilament light by 367.19 (204.28 to 561.61) [s1]. In plasma, interferon gamma rose by 4.37 (3.01 to 5.74) and interleukin-5 by 0.33 (0.12 to 0.64) [s1].

Those markers are the ones neurologists watch as indicators of tau pathology, glial activation and neuronal injury. Rising values are not what a disease-modifying agent is supposed to produce.

The caveats are large and they cut in the trial's favour as much as against it. Ten people, eight weeks, no control group, and every participant knew what they were taking [s1]. Biomarkers fluctuate, and an open-label single-arm study cannot separate a treatment effect from the natural course of a progressive disease. What the trial establishes is not that rapamycin worsens Alzheimer's; it is that the assumption of benign, brain-penetrant action does not survive first contact with measurement.

A different dosing schedule, same measurement problem

A separate open-label pilot took the other approach the field favours — weekly rather than daily dosing. Thirteen patients with early-stage Alzheimer's received 7 mg weekly (eleven patients) or reduced doses of 2 mg and 4 mg (two patients) for 26 weeks, with blood sampled at four timepoints during week 13 [s2].

The pharmacokinetics were reassuring on their own terms. Between-person variability was moderate, with coefficients of variation of 0.28 to 0.40, and the 48-hour post-dose sample was the most stable (CoV 0.28) and correlated best with the pre-dose trough from the previous week (r = 0.72) [s2]. Concentrations stayed below immunosuppressive levels in every participant [s2]. The authors conclude that weekly dosing supports fixed-dose regimens in future trials and that the 48-hour sample is the best monitoring point [s2].

They also attempted to quantify rapamycin in cerebrospinal fluid and ran into methodological problems caused by the sensitivity limits of the assay [s2]. Two independent groups, two dosing schedules, and the same unanswered question about whether the drug reaches the brain.

What is being tested next

Neurocognitive trials of rapamycin have not been completed, and two are in progress: ERAP, a six-month single-arm open-label phase IIa study testing 7 mg weekly and measuring cerebral glucose metabolism, and REACH, which tests 1 mg daily against markers of Alzheimer's disease burden [s4].

A third is registered but not yet recruiting. A randomised, triple-masked phase 1 trial at the University of Missouri plans to enrol 225 women aged 45 to 65 who carry the ApoE4 variant, randomise them to sirolimus or placebo taken daily for about four weeks, and measure change in cerebral blood flow by MRI as its primary outcome [s3]. It was first posted in July 2026 with an estimated start in December [s3].

That design at least targets a mechanism the earlier studies could not test: whether the drug changes brain perfusion in people at genetic risk, before symptoms appear. Whether it can do so without reaching the cerebrospinal fluid is the question the first two trials have left on the table.

Rapamycin is not approved for Alzheimer's disease or cognitive decline. Nothing here is medical advice, and no dose in this article should be read as guidance.

Sources

1. [s1] Rapamycin treatment for Alzheimer's disease and related dementias: a pilot phase 1 clinical trial. Communications Medicine, published online May 20, 2025. https://doi.org/10.1038/s43856-025-00904-9
2. [s2] Pharmacokinetic analysis of intermittent rapamycin administration in early-stage Alzheimer's Disease. GeroScience, published online October 2025. https://doi.org/10.1007/s11357-025-01911-3
3. [s3] Effects of Sirolimus on Middle Aged Asymptomatic ApoE4 Carriers. ClinicalTrials.gov registry record NCT07728175, first posted July 27, 2026. https://clinicaltrials.gov/study/NCT07728175
4. [s4] What is the clinical evidence to support off-label rapamycin therapy in healthy adults? Aging (Albany NY), published online August 7, 2025. https://doi.org/10.18632/aging.206300