PHILADELPHIA – Despite some tantalizing hints of efficacy, the antiamyloid antibody crenezumab failed overall to post significant imaging and biomarker results in a phase II study.
The drug did not effect any statistically significant changes in amyloid plaque burden on PET imaging in the prespecified analysis, Dr. Steven Salloway said at the Clinical Trials Conference on Alzheimer’s Disease. A secondary assessment using a different technique yielded slightly better, although still nonsignificant, results.
There was a slight increase in serum amyloid beta 42 (Abeta42), but no effect at all on serum total or phosphorylated tau, said Dr. Salloway, director of neurology and the Memory and Aging Program at Butler Hospital in Providence, R.I.
The lackluster findings, combined with crenezumab’s failure to meet cognitive and functional endpoints in this and a companion study announced last July, are prompting questions about the antibody’s future, Dr. Salloway said. “As a result of these clinical trials ... the company is in discussion with regulators about further development of crenezumab, although no decision has been made.”
These new data will not, however, endanger the landmark API (Alzheimer’s Prevention Initiative–Autosomal Dominant Alzheimer’s Disease Treatment Trial), now launching in Colombia. Crenezumab is the active treatment in this first-ever primary prevention trial.
In fact, these new imaging and biomarker data have led to a dosing protocol change for the API study. Originally, the study called for the 300-mg bimonthly subcutaneous dose; instead, it will use the IV dosing equivalent, a company spokesperson said in an interview.
“Based on the learnings from the crenezumab exploratory data, which may suggest greater benefit in [patients with mild Alzheimer’s disease] treated with the higher IV dose, Genentech is taking next steps to enable use of a higher subcutaneous dose in the API study,” said Carole Ho. She noted that the ongoing extension studies will also drop the subcutaneous dose in favor of IV dosing.
The 69-week study, dubbed BLAZE, looked at 91 patients with mild to moderate, imaging-proven Alzheimer’s disease. They were randomized to either crenezumab 300 mg subcutaneously every other week, 15 mg/kg intravenously once a month, or matching placebo in each arm.
The study group was a typical Alzheimer’s cohort, with a mean age of about 69 years. The Mini-Mental State Exam scores ranged from 18 to 26 (mean score, about 21).
At baseline, all patients had amyloid PET imaging, fluorodeoxyglucose (FDG)-PET imaging, cerebrospinal fluid (CSF) sampling, and brain MRI. Amyloid PET was repeated at week 47, and all of the tests were repeated again at study completion.
Dropout rates in each treatment group – especially the intravenous group – were rather high. In the subcutaneous group, 77% of the placebo arm and 88% of the active arm completed treatment. In the intravenous arms, 59% of the placebo group and 60% of the active group finished the study. There were two deaths, both in the active intravenous group.
The prespecified amyloid imaging used a cerebellar gray matter reference region. With this method, there were no significant differences in amyloid plaque burden by the end of the study.
Serial scans showed some odd variability that led researchers to question whether the reference region was accurately picking up any changes, Dr. Salloway said. In the subcutaneous group, the placebo group measurement stayed flat, as expected, but the active group measure took a significant dip at the interim scan, and then returned to baseline for the last scan, suggesting that amyloid dispersed and then reformed. In the IV arm, the placebo group took a significant dive from the interim scan to the final scan, suggesting that those patients were spontaneously clearing amyloid. The active IV group seemed to stay flat.
That prompted the investigators to reassess the results with a different reference region, a white matter region closer to the corpus callosum and centrum semiovale, Dr. Salloway said. This region is being used in the Alzheimer’s Disease Neuroimaging Initiative longitudinal cohort study. Data were sent to an independent investigator for a blinded reassessment.
The results obtained using this reference region were more in line with expectations, although not spectacular, Dr. Salloway said.
In the subcutaneous group, amyloid burden rose from baseline, with no difference at all between placebo and crenezumab. In the IV group, amyloid burden also rose in both arms. However, the curves showed some nonsignificant separation at the interim scan, which continued to the end of the study. The difference was still not significant at the final scan (P = .169).
When assessed by yet another white matter reference region, the results were similar, with no difference in the subcutaneous arm and a nonsignificant separation in the IV arm (P = .131).