Loading ...

NEW ORLEANS — An investigational liver-targeted formulation of fast-acting insulin may reduce the risk for hypoglycemia in people with type 1 diabetes (T1D), new phase 2 data showed. 

The product, hepatic-directed vesicle (HDV) insulin (Diasome Pharmaceuticals, Inc.), is made of a specially designed phospholipid matrix that binds to a short-acting insulin such as lispro and preferentially directs it to the liver, thereby more closely mimicking insulin delivery in people without diabetes. 

The “very exciting thing” about liver-selective insulin, or liver-directed insulin, is that “by insulinizing the liver, you can take up glucose and store that glucose as glycogen, which can then be released during periods of hypoglycemia,” study lead investigator Klara Rachel Klein, MD, PhD, told Medscape Medical News. 

The reduction of hypoglycemia is a huge medical need in T1D, noted Klein, director of the Endocrinology, Diabetes, and Obesity Clinical Research unit at the University of North Carolina School of Medicine, Chapel Hill. 

“The only therapy that we have available for [T1D] is insulin, and there's this enormous risk of hypoglycemia that's associated with insulin use,” she explained. “Every day, people are walking a tightrope.” 

Klein presented the findings from the 27-site phase 2b OPTI-2 trial here at the American Diabetes Association (ADA) 2026 Scientific Sessions. 

Phase 2b OPTI-2 Trial

The randomized, double-blind trial compared HDV combined with lispro (HDV-LIS) with standard insulin lispro in 226 adults with T1D. Median baseline A1c was 7.4% in those randomly assigned to HDV-LIS (n = 112), and 7.3% in the lispro group (n = 114). Mean ages were 46.5 years and 43.9, respectively. All participants were on multiple daily insulin therapy using once-daily insulin degludec as basal insulin, and all wore a Dexcom G7 continuous glucose monitor (CGM) during the study. 

A 12-week dose-optimization period was followed by a 13-week maintenance period. The prespecified primary endpoint was a composite of A1c noninferiority at week 12 and superiority in hypoglycemia metrics during the last 6 weeks of dose optimization, including rate of nocturnal level 2 hypoglycemia (< 54 mg/dL), nocturnal percent time below 54 mg/dL, and proportion with less than 1% of a 24-hour day spent below 54 mg/dL. 

At week 12, A1c reductions were 0.34 percentage points for HDV-LIS vs 0.42 for standard lispro, a nonsignificant estimated difference of 0.08 percentage points (P = .26). At week 25, A1c percentage-point reductions were 0.31 and 0.38, respectively, estimated difference 0.07 (P = .40). Results at both timepoints met the FDA’s required < 0.1 percentage-point difference for A1c noninferiority. 

For the hypoglycemia endpoint, 14 of 15 prespecified CGM hypoglycemia metrics favored HDV-LIS and met noninferiority but narrowly missed statistically significant superiority during the 12-week titration period. However, during the 13-week maintenance period, HDV-LIS achieved statistical superiority for several prespecified secondary endpoints, including: 

• 28% reduction in 24-hour level 2 hypoglycemia (< 54 mg/dL) events (rate ratio 0.72, P = .014)
• 33% reduction in daytime level 2 hypoglycemia events (0.67, P = .009)
• 28% reduction in 24-hour percent time below 54 mg/dL (0.72, P = .017)
• 32% reduction in daytime percent time below 54 mg/dL (0.68, P = .010)
• 36% reduction in 24-hour extended hypoglycemia events (0.64, P = .029)

The incidence of level 2 hypoglycemia was 25% lower with HDV-LIS than with lispro throughout the study. No serious hypoglycemia events (level 3, requiring assistance) occurred in the HDV-LIS group, compared with five in the lispro group (P = .0598). 

Treatment-emergent adverse events were similar between groups (53.6% HDV-LIS vs 50.0% lispro). Serious treatment-related adverse events occurred in one person in the HDV-LIS group (0.9%) vs eight participants on standard lispro (7.0%). Adverse events leading to drug discontinuation occurred in none with HDV-LIS and two (1.8%) with lispro. There were no deaths, diabetic ketoacidosis events, or clinically meaningful hepatic safety signals in either group. 

A phase 3 study is planned, Klein said in her presentation. 

“At the end of dose optimization, all of the point estimates favored HDV. It just didn’t reach statistical significance, but it was a very ambitious study, and a small sample size, so I think it was all very reassuring,” she told Medscape Medical News. 

“OPTI-2 suggests that HDV-LIS can decouple glycemic control from hypoglycemia risk,” she said. “If confirmed in phase 3, HDV-LIS has the potential to help people living with T1D walk the tightrope between hyperglycemia and hypoglycemia.” 

In response to an audience member’s question about use in an insulin pump, she said that is certainly a possibility. 

Phase 3 Trial Justified

Asked to comment, session moderator Daniel J. Rubin, MD, professor of medicine at the Lewis Katz School of Medicine at Temple University, Philadelphia, told Medscape Medical News, “I think it's a really interesting mechanism and the data are promising. It’s a phase 2 trial and certainly justifies a larger phase 3 trial.”

He noted that there may be some concern about liver toxicity, although it wasn’t seen in this trial. “That’s the kind of thing that they would further explore in a phase 3, but there weren’t any safety signals here, which is reassuring…I’m looking forward to seeing the phase 3 results,” said Rubin, who is also director of clinical research and deputy chief of the section of endocrinology, diabetes, and metabolism at Temple. 

Klein told Medscape that she is working on another product, cadisegliatin (vTv Therapeutics), currently in phase 3 for T1D, that also works in the liver to reduce hypoglycemia, but in a different way. 

“One of the reasons people who live with [T1D] don't have as much glucose storage is because glucokinase requires insulin to be expressed. The idea is if you can just activate glucokinase in the liver, then you can increase glucose storage the same way that a liver-directed insulin would do — so it’s doing the same thing by different mechanisms.”

She noted, “CGM is a massive advance, but it still doesn't prevent hypoglycemia. Developing ways to try and treat diabetes and get people to be able to more confidently achieve glycemic targets without such significant fear of hypoglycemia would be an extremely significant advance.”

The study was sponsored by Diasome. Klein reported that she is a consultant for Roche Pharmaceuticals, Novo Nordisk, Antag Therapeutics, and Metsera, and is on advisory panels for vTv Therapeutics. Rubin reported no disclosures. 

Miriam E. Tucker is a freelance journalist based in the Washington, DC, area. She is a regular contributor to Medscape, with other work appearing in the Washington Post, NPR’s Shots blog, and Diatribe. She is on X (formerly Twitter) @MiriamETucker and BlueSky @miriametucker.bsky.social