The theory that GLP-1 analogues might play a role in treating cognitive impairment has existed for some time, and the Elad trial of Novo Nordisk’s Victoza was the first major test of the idea. The study failed, it was announced at the virtual CTAD meeting today: Dr Paul Edison of Imperial, which ran the trial, said there was no difference between treatment and placebo on cerebral glucose metabolic rate, the primary endpoint. Some of the secondaries were numerically positive: improvement on the ADAS-Exec z-score, a measure of cognitive function, was seen with Victoza, but there was no difference on other measures of dementia and competence at daily activities. Only 72 subjects were given the drug and 82 placebo, and the trial tracked them for just a year, which might have made finding signals hard. But academics seem keen on further investigation. Dr Clive Ballard of the University of Exeter said there was a "body of evidence growing for the need for a phase III programme” in this field, pointing to a recent Lancet paper on Trulicity. That job would presumably fall to Novo, which is expected to make a decision on exactly that undertaking in the coming months.