Delivery of the hormone in a tablet is still considered the “holy grail,” according to Mads Krogsgaard Thomsen, the chief science officer of Novo Nordisk. Yet Novo, the world’s largest maker of insulin, is reviewing its efforts to determine whether the product can be commercially viable, and how much investment would be needed.

Getting the right amount of any protein into a tablet is a formidable challenge; insulin is particularly difficult because too much or too little can drive blood sugar to dangerous extremes. So the Danish drugmaker has shifted its focus to a pill version of an experimental diabetes drug that stimulates insulin production in the pancreas.

“Coming from a company that’s sold injection-based drugs for 93 years, I have to say we don’t want to talk about the needle as a barrier,” Thomsen said in an interview in Munich. “But it is.”

That’s why an oral form of semaglutide, the next-generation drug in its blockbuster GLP-1 class of medicines, would be “a potentially transformational breakthrough,” Thomsen said.

Bagsvaerd, Denmark-based Novo, which traces its roots back to companies founded as early as 1923, said Friday that semaglutide has proven effective in reducing the risk of heart attacks, strokes and cardiovascular deaths in diabetics by 26 percent. An oral version of the drug could be available in 2020, according to Thomsen.

Illustrating a key advantage of developing a pill form of a GLP-1 medicine over insulin, Thomsen said the former treatment didn’t kill or even seriously injure a patient in a late-stage clinical trial who took 10 times the prescribed dosage. (The patient did throw up.)

“If you had that with an insulin, it would have taken 10 seconds and you would be dead,” he said. “Insulin is much more dangerous than GLP-1, and that means with oral insulin, getting the window right is not easy.”

That doesn’t necessarily mean the company is giving up on its ambitions for insulin in a pill.

“We are looking at all options,” Thomsen said, speaking on the sidelines of the European Association for the Study of Diabetes conference last week. The company in August said it would provide an update on its oral insulin program in the second half of the year.

To assess the future prospects of oral insulin, “you should probably follow what we do or don’t do,” he said.

In other words, smaller rivals such as Oramed Pharmaceuticals Inc. are “highly unlikely” to succeed if Novo doesn’t, according to Thomsen, because those companies are developing short-acting insulins or intermediate-acting insulins. Such treatments, usually taken at meal times, would have too much variation in absorption levels in its pill form to be safe.

Oramed responded to say that it’s moving forward with its oral insulin, which looks more promising after mid-stage trials. The Jerusalem-based company expects to enter late-stage trials next year, Chief Medical Director Roy Eldor had said earlier this month.

Novo, facing pressure on prices and increased competition in the U.S., its biggest market, will probably devote more of its research-and-development spending to partnerships with other companies or universities as it considers “adjacent areas” like cardiovascular medicine, Thomsen said.

The premium that drug purchasers are willing to pay for newer, more innovative treatments has fallen to about 20 percent above the price of older, more-widely used treatments such as Lantus, the aging bestseller from French drugmaker Sanofi, he said. It also will get harder to come up with diabetes medicines that are better than existing ones, he said.

“We will need some more shots on goal,” he said. “The R&D budget is set to increase in the company, but it will be composed maybe more of external collaborations, seeking out new boundaries and new targets.”