Eli Lilly has put its weight behind AstraZeneca’s BACE1 inhibitor. On September 16, the two companies announced they will collaborate on the Phase 2/3 AMARANTH trial of AZD3293 (see AstraZeneca press release). AZD3293 recently completed Phase 1 safety testing, with results rosy enough to tempt Lilly to ante up $500 million to buy in. The partners will share equally in all future costs and revenues. Lilly terminated testing of its own BACE inhibitor in 2013 due to liver toxicity (see Jun 2013 news story). In the once-crowded field of BACE inhibitors, only Merck’s MK-8931 maintains a lead on AZD3293, having passed initial safety benchmarks in its own Phase 2/3 EPOCH study last year (see Dec 2013 news story).
While adverse events have stopped several previous β-secretase and γ-secretase inhibitors, the light remains green for AZD3293. At the 13th International Geneva/Springfield Symposium on Advances in Alzheimer Therapy, held last March in Geneva, Samantha Budd of AstraZeneca's Cambridge, Massachusetts, office presented Phase 1 data from healthy volunteers. Participants tolerated single doses up to 750 mg, as well as 12 days of dosing with 50 mg per day, Budd reported. She will detail similar Phase 1 results from Alzheimer’s patients at the seventh Clinical Trials Conference on Alzheimer’s Disease (CTAD) in Philadelphia November 20 to 22. “We are very excited about that data,” she told Alzforum.
Like most BACE inhibitors, AZD3293 inhibits both BACE1 and BACE2. Academic studies have identified numerous substrates besides amyloid precursor protein for these two enzymes. BACE1 appears to play a role in axon guidance, myelination, and maintenance of muscle spindle fibers, to name just a few areas (see, e.g., Jun 2012 news story; Jul 2013 news story; and Dec 2013 news story). In knockout mice, a lack of BACE1 can lead to seizures and problems with learning (see Sep 2008 news story; Jul 2010 news story). So far, none of these issues have cropped up in short-term trials. “We looked very carefully for a number of problems indicated by the BACE1 knockout papers,” Budd told Alzforum, adding that they did not see any changes in myelin or muscle pathology, even after completing a toxicology assessment of AZD3293 at doses far exceeding expected human exposures. AstraZeneca’s interpretation of this is that the findings in knockout mice are likely due to developmental effects, she suggested.
Other researchers have raised the question of what happens to APP in the wake of BACE inhibition. Researchers expected α-secretase activity to rise and pick up the slack, but some animal studies have cast doubt on whether this happens (see Jun 2014 news story). However, Budd did see a rise in a cleavage fragments in Phase 1 participants, indicating that APP was indeed being shunted into this processing pathway. Recently, Christian Haass at the Ludwig-Maximilians-Universität in Munich noted that BACE inhibition in some experimental settings leads to the generation of a small synaptotoxic fragment (see Oct 2014 conference story). AstraZeneca is looking for the presence of such small fragments in cerebrospinal fluid using mass spectrometry, Budd noted.
Meanwhile, the AMARANTH trial will soon begin enrolling. AstraZeneca announced plans for the study earlier this year (see Mar 2014 news story). By going straight from Phase 2 to 3 with the same group of patients, the company hopes to get results more quickly, Budd said. The trial will enroll about 1,500 participants who have mild AD or mild cognitive impairment (MCI) due to AD, as judged by biomarker evidence of brain amyloid accumulation. Participants’ scores on the Mini-Mental State Examination must range from 21 to 28, indicative of mild impairment. “We are enrolling patients earlier in the spectrum of disease than many past studies did,” Budd noted.
A third of the patients will receive 50 mg AZD3293 once per day, a third 20 mg/day, and the rest placebo. Researchers will assess safety based on the first three months of Phase 2 data, while dosing continues. If AZD3293 clears this hurdle, the trial will run for two years. The primary endpoint will be change on the Clinical Dementia Rating-Sum of Boxes (CDR-SOB). Secondary endpoints will include numerous cognitive and functional measures, such as the ADAS-Cog 13, global CDR, Alzheimer’s Disease Cooperative Study Activities of Daily Living Inventory, and the Neuropsychiatric Inventory. AstraZeneca will track biomarker changes as well. All participants will receive MRI scans, with a subset also getting amyloid PET imaging and another group undergoing lumbar punctures to measure Aβ and tau in CSF. The trial does not include tau imaging at present, although Budd noted they are closely watching the data from tau tracers in development and could add that later.
Both Lilly and AstraZeneca have amyloid tracers, although it is not yet clear which one the AMARANTH trial will use. Lilly’s Amyvid received approval from the Food and Drug Administration in 2012 (see Apr 2012 news story). In 2011, AstraZeneca licensed their tracer, now called NAV4694, to Navidea Biopharmaceuticals in Dublin, Ohio, for development, but the agreement specifies that AstraZeneca can still use NAV4694 in clinical trials (see AstraZeneca press release). Since then, two other tracers, flutemetamol and florbetaben, have also received regulatory approval (see May 2014 news story).
Amyloid PET experts who have seen preliminary data from NAV4694 believe that it may be the best tracer yet because of its very low background (see May 2010 news story). Privately, they lament that it is not yet clinically approved or embedded in more therapeutic trials. NAV4694 is currently in a Phase 3 trial to evaluate the tracer’s accuracy against autopsy data, as well as a Phase 2 study in MCI patients to see how accurately a positive scan can predict progression to AD.
Cornelia Reininger, Navidea’s chief medical officer, noted that several research studies, such as the Australian Imaging, Biomarker and Lifestyle (AIBL) Flagship Study of Ageing (see press release), use the tracer. Siemens’ PETNET Solutions, a radiopharmacy network, manufactures NAV4694 at numerous locations worldwide.
“I’m very excited about having this tracer available to the medical community,” said Reininger, who previously developed florbetaben at Bayer Healthcare Pharmaceuticals. “With this tracer, scans are unambiguous. In a recent study, we read 107 scans in 75 minutes, and there was only one we had to look at twice. It’s yes/no.”—Madolyn Bowman Rogers
- Lilly Halts Phase 2 Trial of BACE Inhibitor Due to Liver Toxicity
- Merck BACE Inhibitor Clears a Safety Hurdle, Gets New Trial
- BACE Secrets: Newly Identified Substrates May Regulate Plasticity
- Paper Alert: BACE1 Required for Muscle Spindle, Motor Control
- Blocking BACE—Do Adult Mouse Phenotypes Predict Side Effects?
- BACE Knockouts Show More Synaptic Troubles
- Off BACE: Lack of Secretase Causes Seizures in Knockout Mice
- Scant α-Secretase Compensation for BACE Inhibition in Monkeys
- At Birthday Symposium, Massachusetts ADRC Looks to Future With New Data
- BACE Inhibitor Heads for Phase 2/3 Trials
- FDA Approves Amyvid for Clinical Use
- Three’s Company: Florbetaben Approved, Excludes AD Diagnosis
- Geneva: The AstraZeneca Ligand—The Fairest of Them All?
- BACE Inhibitors Barrel Forward—Next Hurdles: Safety, Efficacy
- Merck Launches Largest Trial of BACE Inhibitor in AD
- Wave of New BACE Inhibitors Heading to Phase 2
- Not the Usual Suspects: Tracking BACE Inhibition, Axon Role
- Barcelona: Out of Left Field—Hit to The Eye Kills BACE Inhibitor
- Paper Alert: First BACE Trial Published—New Inhibitor On Deck
- Rounding Second BACE: Another Route to Memory Loss?