Background

This gene-silencing approach uses intrathecal injection of a small interfering RNA targeting the amyloid precursor protein mRNA as a potential treatment for Alzheimer’s disease and cerebral amyloid angiopathy. The goal is to reduce expression of APP in the central nervous system and thus depress production of substrates for amyloid deposition, e.g. Aβ42 in AD and Aβ40 in CAA.  

Mivelsiran fuses a synthetic siRNA to a proprietary 2'-O-hexadecyl (C16) lipophilic conjugate, which was designed to enhance cellular penetration and distribution in the CNS. In published preclinical work, C16-siRNAs delivered by intrathecal or brain injection in mice spread broadly through the brain and targeted neurons, astrocytes and microglia. The half-life of gene suppression was three to four months. In nonhuman primates, a single intrathecal injection of ALN-APP knocked down APP mRNA in the spinal cord and brain, and drove down CSF levels of the APP fragments sAPPα and sAPPβ by about 75 percent. The reductions persisted for two to three months, and gradually recovered to normal after about nine months. Intracerebroventricular injection in the CVN transgenic mouse model of brain and vascular amyloid deposition resulted in 50 percent reduction of APP mRNA and sAPPα, reduced Aβ40 deposition and inflammation, and normalization of brain glutamate levels and open-field behaviors (Brown et al., 2022).

This is the first siRNA therapy to be tested in people for a CNS-targeted disease. Alnylam’s marketed siRNA drugs target liver and other peripheral tissues.

Alnylam was co-developing this drug with Regeneron until 2024, when Regeneron opted out of further collaboration (press release).

Findings

In February 2022, Alnylam began a first-in-human Phase 1 study of the safety and target engagement of ALN-APP. The single- and multiple-dose trial will enroll 60 adults with MCI or mild dementia due to early onset AD, as defined by symptoms appearing before age 65. Participants must have a CDR global score of 0.5 or 1.0, and MMSE greater than 20, with no recent treatment with an amyloid-targeting antibody. Participants will receive a single, escalating dose of ALN-APP or placebo by intrathecal injection, or multiple doses over one year, against a primary outcome of adverse events. Secondary endpoints are CSF levels of sAPPα and sAPPβ, and pharmacokinetics. The trial will run through July 2025 at six sites in the U.S., Canada, Netherlands, and the U.K. 

On April 26, 2023, the company reported single-dose data on 20 patients in three dose cohorts (press release). AAdverse events were mild to moderate, the company said, without apparent elevation of CSF protein or white blood cells, and with neurofilament light chain readings comparable to placebo. ALN-APP caused dose-dependent reductions of CSF sAPPα and sAPPβ. The highest dose lowered sAPP proteins by up to 90 percent within two weeks, and maintained greater than 70 percent reduction for at least three months. At that time, multiple dose evaluation was under a partial clinical hold in the U.S., due to FDA concerns about findings in animal toxicology studies. Multiple ascending doses were approved to begin in Canada.

At the October 2023 CTAD conference in Boston, more data on the first three single-dose cohorts were presented (news). Each enrolled six or eight participants, who received 25, 50, or 75 mg siRNA, or placebo. The 50 and 75 mg doses reduced CSF Aβ42 and Aβ40 by 50 and 70 percent, respectively, after two months. Adverse events were mild to moderate; the majority were deemed related to lumbar puncture. Consistent with previous reports, the two highest doses reduced CSF sAPPα and sAPPβ concentrations by 69 and 82 percent, respectively, after two weeks. In the 75 mg group, sAPPα and β were 33 and 30 percent lower 10 months after dosing. 

In February 2024, Alnylam announced that the FDA had lifted the clinical hold to allow multiple dosing of up to 180 mg every six months, which is sufficient for the planned Phase 1 regimens (press release). A partial hold remains on higher or more frequent doses.

In May 2024, Alnylam began a Phase 2 trial in people with cerebral amyloid angiopathy. The study plans to enroll 200 people diagnosed with probable CAA based on MRI evaluation. Participants are to receive a single intrathecal infusion of mivelsiran or placebo, and to be followed for two years. People with MCI or mild AD in addition to CAA are eligible. Those with moderate or severe AD, and any who have had amyloid antibody treatment, are not. The primary endpoint is the rate of new cerebral lobe microbleeds. Secondary endpoints include signs of disease progression on MRI or fMRI, change in CSF sAPPα and sAPPβ, and adverse events. A separate cohort will enroll people with hereditary CAA caused by the E693Q APP mutation, who will not contribute to the primary endpoint. A 1.5-year open label extension is offered. The trial is running at thirteen locations in the United States and Canada, with completion of the placebo-controlled portion planned for July 2027.

At the October 2024 CTAD meeting, the company presented additional Phase 1 data for the 50 and 75 mg single-dose groups (conference news). On both doses, CSF sAPPβ remained 30 percent below baseline after one year. No ARIA was detected. The 75 mg dose reduced Aβ40 and Aβ42 by 27 and 40 percent relative to baseline, respectively, at six months.

Also in October 2024, Alnylam released initial results of multiple dosing in Phase 1 (press release). Patients who received 50 mg six months after the first dose, showed greater than 90 percent reduction in sAPPβ at seven months. No new safety issues arose, and the company reported no significant abnormalities in CSF total protein and white cell count, or on the exploratory biomarker neurofilament light chain. The company said it intends to initiate a Phase 2 study in patients with Alzheimer’s disease around the end of 2024.

For details on ALN-APP trials, see clinicaltrials.gov.