Just as the failure of avagacestat in 2012 was fading into the realm of sour memories, researchers have revived it with a detailed data analysis. The drug’s last trial, a Phase 2, aimed to test this γ-secretase inhibitor in people with prodromal Alzheimer’s disease, but screeched to a halt amid reports of gastrointestinal problems, skin cancer, and lack of efficacy among participants. The new analysis, published on September 28 in JAMA Neurology and led by Vladimir Coric of Bristol-Meyers Squibb in Wallingford, Connecticut, offers a silver lining: CSF biomarkers worked as a selection tool for people likely to progress to dementia. As such, the trial supports diagnostic criteria for MCI due to Dementia (see Apr 2011 news). 

Researchers developed avagacestat for its ability to dampen Aβ42 production, but as with other γ-secretase inhibitors, the drug’s potential effects on processing of other substrates—notably Notch—were always a concern. Just prior to the termination of the prodromal AD trial, BMS had also halted another Phase 2 trial of the inhibitor in people with mild to moderate AD, due to similar side effects and worsening cognition in the treatment group (see Nov 2012 newsDec 2012 news). In addition, a Phase 3 trial of another γ-secretase inhibitor, Eli Lilly’s semagacestat, also crumbled due to side effects and lack of efficacy (see Aug 2010 news). Taking a hint, researchers are now focusing on developing γ-secretase modulators, which aim to only tweak the enzyme’s processing of APP while sparing its relationship with other substrates.

While researchers agree that broad γ-secretase inhibitors were a lost cause, Coric and colleagues wanted to glean as much information as they could from the avagacestat trial to aid in the design of future trials.

The prodromal AD study enrolled 263 participants who met clinical criteria for mild cognitive impairment (MCI) and had a CSF biomarker profile indicative of early AD. This was an Aβ42 level of less than 200pg/mL or a ratio of total tau to Aβ42 of greater than or equal to 0.39. Half of these patients were randomized to receive 125 mg daily doses of avagacestat for two years, and the others placebo. Another 102 people who met criteria for MCI but at screening proved to be biomarker-negative entered into a parallel observational cohort designed to measure the usefulness of biomarkers at predicting progression to dementia. Participants were monitored with cognitive and neuropsychological tests every 12 weeks to track executive function and memory, and progression to dementia. Three further CSF samples were taken throughout the trial. A subset of the participants also received F-18 florbetapir PET scans to measure amyloid burden.

Of the 132 participants on avagacestat, 49 (37.1 percent) experienced severe adverse events, compared with 31 of the 131 people (23.7 percent) on placebo. Non-melanoma skin cancers, i.e., basal and squamous cell carcinomas, accounted for much of the difference between treatment and placebo groups, although people on the drug also had higher rates of severe gastrointestinal problems. Mild to moderate adverse events that predominated in the treatment group included skin rashes, weight loss, diarrhea, and vomiting. More than half the treatment group also had renal problems that drove up glucose levels in their urine, but these resolved after treatment stopped. BMS reduced the daily dose to 50 mg and gave participants the option of dropping to 25 mg if symptoms persisted.  Still, 72 patients on drug dropped out--45 due to adverse events, versus 53 taking placebo, 14 of whom had adverse events as well.

While the reasons for the toxicity remain uncertain, the researchers suggested the most likely culprit was inhibition of Notch processing. In previous animal studies, Notch inhibition destroyed cells lining the gut and immune cells, and caused squamous cell carcinoma (see Milano et al., 2004; Wong et al., 2004Proweller et al., 2006). 

While the lower dose caused less severe side effects, it also failed to stem cognitive decline. Treatment and placebo groups stayed the same in activities of daily living, or on the ADAS-Cog, MMSE, or CDR Sum of Boxes.

Avagacestat treatment modestly affected CSF Aβ concentrations. At weeks 24 and 104 of the trial, people treated with the drug had a 10 to 15 percent reduction in CSF Aβ40 and a 5 to 9 percent reduction in Aβ42, the latter being not significantly different from the placebo group. Amyloid PET scans displayed nearly full concordance with pathology indicated by CSF Aβ42.

Volumetric MRI revealed that people in the treatment arm had faster whole brain and hippocampal atrophy than those on placebo and concomitant increase in ventricle size. The researchers pointed out that similar levels of atrophy occurred during trials of other therapies targeting amyloid (see Fox et al., 2005Salloway et al., 2014). 

Progression Predictor. Most people negative for amyloid based on CSF biomarkers Aβ42 and total tau (gray) did not progress to dementia, as opposed to half among CSF biomarker-positive groups (light blue and orange). [©2015 American Medical Association. All rights reserved.]

With regard to progression to dementia, the small difference between the avagacestat and placebo groups was not statistically different, although the treatment group showed a trend toward more progression after one year. However, the researchers found dramatic differences in progression between the randomized cohort and the observational cohort that was biomarker-negative at screening. After two years, rates of progression were around 30 percent in the biomarker-positive avagacestat and placebo groups, compared with just 6.5 percent in the observational, biomarker-negative group. This indicated that CSF biomarkers served as a useful screening tool to select for participants likely to develop AD.

While the CSF biomarkers in combination with clinical criteria did enrich for people who developed dementia, the researchers pointed out that not everyone in the prodromal AD groups progressed within the two-year time frame of this trial. Longer studies would be needed to fully determine the strength of these patient selection tools. The authors added that high variability in CSF biomarker measurements, and the use of an assay not yet approved as a diagnostic test, were limitations that could be improved in future trials. With fully automated diagnostic tests on the horizon (see Aug 2015 conference news), this could become a real possibility.—Jessica Shugart



No Available Comments

Make a Comment

To make a comment you must login or register.


News Citations

  1. Revised Diagnostic Criteria for Alzheimer’s Are Published
  2. Déjà Vu? AD Patients Again Look Worse on γ-Secretase Inhibitor
  3. Drug Company Halts Development of γ-Secretase Inhibitor Avagacestat
  4. Lilly Halts IDENTITY Trials as Patients Worsen on Secretase Inhibitor
  5. CSF Aβ Assays Remain Fickle: Robots to the Rescue?

Therapeutics Citations

  1. Avagacestat
  2. Semagacestat

Paper Citations

  1. . Modulation of notch processing by gamma-secretase inhibitors causes intestinal goblet cell metaplasia and induction of genes known to specify gut secretory lineage differentiation. Toxicol Sci. 2004 Nov;82(1):341-58. PubMed.
  2. . Chronic treatment with the gamma-secretase inhibitor LY-411,575 inhibits beta-amyloid peptide production and alters lymphopoiesis and intestinal cell differentiation. J Biol Chem. 2004 Mar 26;279(13):12876-82. PubMed.
  3. . Impaired notch signaling promotes de novo squamous cell carcinoma formation. Cancer Res. 2006 Aug 1;66(15):7438-44. PubMed.
  4. . Effects of Abeta immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease. Neurology. 2005 May 10;64(9):1563-72. PubMed.
  5. . Two phase 3 trials of bapineuzumab in mild-to-moderate Alzheimer's disease. N Engl J Med. 2014 Jan 23;370(4):322-33. PubMed.

Further Reading

Primary Papers

  1. . Targeting Prodromal Alzheimer Disease With Avagacestat: A Randomized Clinical Trial. JAMA Neurol. 2015 Nov;72(11):1324-33. PubMed.