While Phase 2 results of crenezumab (see Part 2 of this series) grabbed headlines at this year's Alzheimer's Association International Conference, other unfolding stories warrant equal attention. Ana Graf from Novartis presented data from a Phase 2 dose-finding study for CAD106, an active immunotherapy against Aβ. This came the day after Novartis and Banner Health of Phoenix announced a partnership to test an active immunotherapy and a BACE inhibitor in cognitively normal adults who have two copies of the ApoE4 risk allele (see Part 1 of this series). Graf confirmed to Alzforum that this ApoE4 trial will test CAD106, but could not disclose the identity of the BACE inhibitor.
Novartis has been evaluating CAD106 in Alzheimer’s patients since 2005. The immunotherapy uses a carrier particle to present the N-terminal of Aβ (amino acids 1-6) to the immune system (see Aug 2006 conference news). The latest trial enrolled 121 people with mild AD to test the safety, tolerability, and immunogenicity of 150 or 450 mg of CAD106 injected with or without adjuvant. Secondary outcomes included amyloid PET, brain atrophy, serum Aβ, and CSF analysis of Aβ, total tau, and phospho-tau. Researchers gave 103 people up to seven injections of the immunotherapy over a period of 60 weeks, with final analysis at week 90. Fifteen people got placebo. This 7:1 randomization is unusual, said Graf, but it fit the study’s main safety and immunogenicity objectives.
Novartis monitored serological response, such as antibody titers. One person on the high dose and 15 on the low dose mounted no antibody response, and the company withdrew them from the trial. Overall, 54 percent on the low dose mounted a strong serological response, as did 76 percent on the high dose. About 8 percent of patients dropped out in the treatment group for safety reasons. Alum and MF59, the two adjuvants used, did not increase the antibody titers.
Graf said no major safety issues arose during the trial. The study detected amyloid-related imaging abnormalities, or ARIA, after dosing in five patients on CAD106, all of whom were strong serological responders. Four had ARIA-H, related to micro-hemorrhage, and one had ARIA-E, related to vasogenic edema. All five were asymptomatic; their ARIA was seen on MRI scans scheduled by the protocol. Even so, three of them were taken off the study because the protocol required a stop when at least two new micro-hemorrhages occurred, Graf wrote to Alzforum by email.
Serious adverse events related to CAD106 included dermatitis, atrial fibrillation, and acute psychosis. Adverse events occurred more often in patients on drug than placebo, but were not dose-dependent, said Graf. The use of alum seemed to minimize flu-like responses.
Plasma Aβ increased two- to fivefold in people who mounted a strong serological response compared to people on placebo, or to people on CAD106 who mounted no immune response. Amyloid PET imaging in 26 patients hinted that those with the highest antibody titers bound less florbetapir at 78 weeks than at baseline, and the correlation was statistically significant, said Graf. She said there were no changes in CSF Aβ, but CSF p-tau fell in those with the strongest antibody response.
Neumann told Alzforum that NB-360 differs from other BACE inhibitors in that it is based on a 3-amino-dihydro-oxazine core, rather than hydroxyethylene or ethanolamine. He said that while this core was difficult to synthesize, it may have advantages over the thiazine ring found in some other BACE inhibitors. The 3-amino-oxazine core allows for tweaks that render the compounds more brain permeable. NB-360 is highly selective over other aspartyl proteases, for example cathepsin D, and is potent in vivo in acute studies in mice, rats and dogs, according to the poster.
For a treatment study, Novartis tested the compound in APP51 mice, which express human wild-type amyloid precursor protein. The study used female mice beginning at 14½ months of age, when the animals have parenchymal Aβ deposits. Doses of 100 micromoles/kg of NB-360 for six weeks reduced Aβ staining in the brain, Neumann’s poster showed. Treated animals had fewer plaques, less Aβ40 and Aβ42 in the forebrain, and fewer activated microglia, as judged by Iba1 staining. The researchers did not record any behavioral changes. Neumann told Alzforum that proof of target engagement and a good toxicity profile are sufficient indications to develop the drug further.
These results are investigational, Neumann wrote to Alzforum; efficacy and safety have not been established for this compound, and next steps remain under review.—Tom Fagan
- Crenezumab Disappoints in Phase 2, Researchers Remain Hopeful
- Novartis to Partner with Banner Health on ApoE4 Prevention Trial
- Madrid: News from the Vaccine Front—Phase 1 Hopefuls
- Winblad B, Andreasen N, Minthon L, Floesser A, Imbert G, Dumortier T, Maguire RP, Blennow K, Lundmark J, Staufenbiel M, Orgogozo JM, Graf A. Safety, tolerability, and antibody response of active Aβ immunotherapy with CAD106 in patients with Alzheimer's disease: randomised, double-blind, placebo-controlled, first-in-human study. Lancet Neurol. 2012 Jul;11(7):597-604. Epub 2012 Jun 6 PubMed.