By Deena Beasley

SAN DIEGO Dec 2 (Reuters) - Novo Nordisk in 2020 launched pivotal trials of its GLP-1 drug semaglutide in Alzheimer's patients based on studies ​in humans, animals and real-world findings, a top company executive said on Tuesday, acknowledging ‌criticism that Novo's studies had design flaws.

Although the trials failed to show statistically significant slowing of cognitive decline in patients given the drug,‌ "we still think it was the right decision... a scientific question that needed an answer," Peter Johannsen, Novo's international medical vice president, said in an address at the Clinical Trials in Alzheimer's Disease meeting in San Diego.

Data, now consolidated on Novo's website, had shown evidence that the GLP-⁠1 hormone is involved in neurotransmission, ‌with multiple effects across the brain, he said.

While Alzheimer's is defined by the presence of toxic amyloid plaques in the brain, "there are still things we don'‍t know" about the pathology of the disease, Johannsen said. "This is a very complex disease with a lot of things going on with different genetic signatures."

Novo is slated to present on Wednesday initial results from the two 2-​year studies that tested Novo's GLP-1 diabetes pill Rybelsus against a placebo in nearly 4,000 Alzheimer'‌s patients.

Full results will be presented at a different medical meeting in March. The company issued a short press release last week saying the studies did not meet their goals.

COGNITIVE BENEFITS IN DIABETES PATIENTS

Johannsen said retrospective studies have shown cognitive benefits for diabetes patients using GLP-1s, which were first approved for blood-sugar control, with gains showing up after about a year of treatment, and building with longer-term use.

Some of those analyses did ⁠not specify which type of dementia a patient developed. Some ​of the real-world evidence was based on clinical diagnosis of ​Alzheimer's rather than more accurate testing and identification of amyloid plaques.

About 60% of people with dementia have Alzheimer's, according to the Alzheimer's Association, with the remaining cases ‍caused by vascular or other ⁠issues.

Johannsen noted potential "biases" in real-world analyses. The diabetes patients prescribed GLP-1s likely had access to endocrinologists, rather than only primary care, and may be in higher socioeconomic groups ⁠than the general population, he said.

Patients on GLP-1s for diabetes probably have better glycemic and metabolic control than those not ‌on the treatments, he said, possibly delaying them from seeking further help and being diagnosed ‌with dementia.

(Reporting By Deena BeasleyEditing by Bill Berkrot)