Background

This small molecule is an orally available protease inhibitor targeting the lysine proteases of the periodontal pathogen Porphyromonas gingivalis. Known as gingipains, these proteases penetrate gingival tissue and cause inflammation at the site of periodontitis (O'Brien-Simpson et al., 2009). Periodontitis has been linked epidemiologically to cognitive impairment, and P. gingivalis bacterial lipopolysaccharide has been detected in postmortem brain tissue of people with AD (Poole et al., 2013). Oral P. gingivalis has been called a risk factor for Alzheimer's disease (Kanagasingam et al., 2020). 

Cortexyme’s approach is based on the theory that P. gingivalis invades the brain, where gingipains contribute to Alzheimer’s pathology (see Sabbagh and Decourt, 2022). The company reported elevated gingipain in brain tissue from people with AD, and a correlation between levels of gingipain and tau proteins in postmortem middle temporal gyrus from AD and healthy control tissue. P. gingivalis DNA was detected in postmortem cortices from people with AD and healthy controls, and in CSF of AD patients (Jan 2019 news on Dominy et al., 2019). In the same study, they show that in mice, oral P. gingivalis infection led to the appearance of bacterial DNA in the brain, increased brain Aβ42 production, neuroinflammation, and hippocampal degeneration. The first three findings were reported to be reduced by atuzaginstat; results for hippocampal cell death were not reported.

In preclinical work from other labs, infection with P. gingivalis was reported to worsen AD pathology and cognitive impairment in AD transgenic mice, and to cause neuroinflammation, memory impairment, neurodegeneration, micro- and astrogliosis, increased brain Aβ and phospho-tau, and neurofibrillary tangles in wild-type C57Bl6 mice (Ishida et al., 2017; Ilievski et al., 2018; Ding et al., 2018). For a review of the preclinical literature, see Costa et al., 2021.

In human neurons grown in culture, P. gingivalis infection led to tau phosphorylation and degradation, synapse loss, and cell death (Haditsch et al., 2020).

P. gingivalis is associated with cardiovascular disease. In rabbits, oral infection was reported to increase arterial plaque and levels of the inflammatory marker CRP. Both were reversed by treatment with COR388 (2020 AAIC abstract). In aged dogs with periodontal disease, ninety days of COR388 reduced oral bacterial load and gum pathology (Arastu-Kapur et al., 2020). In addition, older dogs had bacterial antigens and ribosomal RNA in their brains, consistent with systemic infection seen in humans.

Findings

Two Phase 1 trials of atuzaginstat were completed by June 2019. In a single-dose study of 5 to 250 mg capsules in 34 healthy adults, the compound was safe and well-tolerated. A multiple-dose study assessed safety and tolerability in 24 healthy older adults (mean age of 60 years) and nine with AD (mean age 72). According to a company press release and a poster presentation at the 2018 CTAD conference, healthy adults received 25, 50, or 100 mg COR388 or placebo every 12 hours for 10 days; AD patients took 50 mg or placebo every 12 hours for 28 days. The pharmacokinetic profiles of COR388 in AD and controls were reported to be similar. All volunteers with AD had P. gingivalis DNA fragments in their CSF at baseline. COR388 caused no serious adverse reactions, and no one withdrew. Gingipains also were reported to degrade ApoE, and 28 days of treatment with COR388 was claimed to reduce CSF ApoE fragments (2020 AAIC abstract).

A Phase 2/3 trial (GAIN) evaluating a 48-week course of COR388 in 643 people with mild to moderate AD began in April 2019. Participants took either 40 mg, 80 mg, or placebo twice daily. The primary endpoint was to be ADAS-Cog11 score, and the ADCS-ADL was added later as a co-primary functional endpoint. Further outcomes included CDR-SB, MMSE, NPI, the Winterlight Speech Assessment, MRI brain scans, and change in periodontal disease status. Investigators assessed CSF Aβ and tau, plus P. gingivalis DNA and gingipains in CSF, blood, and saliva, before and after treatment. A dental substudy of 228 participants is assessing effects of COR388 on periodontal disease. This trial involves 93 sites in the U.S. and Europe. The U.S. sites are offering a 48-week open-label extension.

According to a presentation at the 2020 CTAD, GAIN was fully enrolled. At baseline, more than 80 percent of participants had CSF Aβ and tau levels consistent with amyloid positivity or an AD diagnosis. All had detectable antibodies to P. gingivalis in their blood. In the dental substudy, 90 percent had periodontal disease. In December 2020, an independent data-monitoring committee recommended continuing the trial after a planned futility analysis of 300 patients treated for six months (press release).

In February 2021, the FDA placed a partial clinical hold on GAIN because of liver abnormalities in some participants (press release). Dosing in the open-label extension was stopped, but the placebo-controlled portion of GAIN continued. Cortexyme characterized the liver effects as reversible and showing no risk of long-term effects.

In October 2021, Cortexyme announced top-line results indicating the trial had missed its co-primary endpoints of ADAS-Cog11 and ADCS-ADL (press release). The company reported a statistically significant 57 percent slowing of decline on the ADAS-Cog11 in a subgroup with detectable saliva P. gingivalis DNA at baseline who took the higher dose; a 42 percent slowing on the lower dose did not reach statistical significance. This prespecified subgroup analysis included 242 participants; it found no effect on the ADCS-ADL. Improvements in ADAS-Cog and other cognitive endpoints correlated with reductions in saliva P. gingivalis DNA, according to data presented at CTAD 2021 in November. The most common treatment-related adverse events were gastrointestinal, occurring in 12 to 15 percent of treated participants. The treatment groups had dose-related liver enzyme elevations greater than three times the upper limit of normal, in 7 and 15 percent of participants on low and high doses, respectively, with bilirubin elevation reported in two participants on the high dose. The elevations occurred mainly in the first six weeks of treatment, and all resolved without long-term effects. Discontinuations due to transaminase elevations numbered one on placebo, and five and 17 in the 40 mg and 80 mg groups, respectively. The overall dropout rate was 25 percent in the placebo group, and 40 percent in atuzaginstat groups. There were five deaths in the high dose arm, and one in the low dose. All were deemed unrelated to drug. There was no evidence of ARIA or other imaging abnormalities.

At CTAD, the company announced plans for a confirmatory trial, pending discussions with regulators. The plan was to test atuzaginstat in people with mild to moderate AD and evidence of P. gingivalis infection, at the lower dose of 40 mg twice daily, reached by titration to minimize liver effects. The company was also planning a trial in Parkinson’s disease to begin in 2022. These trials were never registered.

In September 2021, Cortexyme began a Phase 1 trial of a second-generation lysin-gingipain inhibitor, COR588 (press release). This compound is expected to require only once-daily dosing. Results were expected in May 2022.

In January 2022, the company announced that the FDA had placed a full clinical hold on atuzaginstat due to concerns about liver toxicity (press release). The company said it intended to develop its backup compound, COR588, for Alzheimer’s disease, pending Phase 1 results. In July 2022, Cortexyme announced that COR588 had met safety and tolerability endpoints in a single- and multiple-ascending dose study in healthy adults (press release).

In August 2022, Cortexyme discontinued the gingipain inhibitor program, and offered it for external licensing (press release). The company changed its name to Quince, and its focus to bone disease. In January 2023, Quince put out word that it had sold Cortexyme’s legacy small molecule protease inhibitor portfolio to Lighthouse Pharmaceuticals, a company co-founded by a former Cortexyme CEO (press release).

For all trials of atuzaginstat, see clinicaltrials.gov.