Could Aiming Insulin at the Liver Be an Answer to Using Less?

Amid the debate about why insulin costs so much lies a second question: if the hormone can be delivered more efficiently, will people with diabetes need less insulin?

It’s not just a question that’s come up in the era of the artificial pancreas, as people with type 1 diabetes (T1D) work with technology that uses an algorithm to anticipate their blood glucose levels and make constant adjustments to their insulin needs.¹

A company called Diasome is pushing the idea further, by trying to take insulin to the place where it would go if created by islet cells and released by a human pancreas: straight to the liver.²

According to Robert Geho, MBA, Diasome’s co-founder and chief executive officer, the idea is to gain control over the element of insulin delivery that has thus far eluded patients and doctors—where insulin goes when it’s administered.

In an interview earlier this year with Evidence-Based Diabetes Management™(EBDM), Geho explained that patients can control how much insulin they take, and there have been many advances in how fast it acts. But until now, delivery is still under the skin, not targeted to the organ where insulin is normally stored in people without diabetes.

Years in development, Diasome has created a nanotechnology additive, consisting of thousands of phospholipid discs, which are added to insulin and bind to it, causing the insulin to be directed toward the liver with no safety issues. Through this method, called hepatocyte-directed vesicle (HDV) technology, the liver can now store and distribute insulin throughout the body. The process yields 2 important results: patients should need less insulin, and there is less risk of hypoglycemia. Recent results confirm this is especially true in patients with poor glycemic control.

Data presented in June at the American Diabetes Association (ADA) Scientific Sessions in San Francisco and published this week in Diabetes Care, an official ADA journal,³ are important steps toward bringing the company’s additive through the development process. These are data from ISLE-1, the phase 2b study that randomized 176 patients 2:1 to either use insulin lispro (Humalog) with the HDV additive (HDV-L) or insulin lispro (LIS) alone. The primary outcome of the study was to show glycated hemoglobin (A1C) noninferiority of HDV-L, with a prespecified margin of ≤0.4%. Results showed the following:

• After 26 weeks, the difference in change from baseline A1C for HDV-L was 0.09% (95% CI, ­­—0.18% to 0.35%), confirming noninferiority. Notably, those with higher A1C at baseline (≥8.5%) showed reductions of 0.5% in both groups, but those taking the additive did so with 25% less bolus insulin.
• Among the patients with higher baseline A1C, there was less hypoglycemia and lower insulin dosing with similar A1C outcomes between HDV-L and LIS. For those with A1C ≥8.5%, incidence rates of severe hypoglycemia were 69 events/100 person years for HDV-L and 97 events/100 person years for LIS (P = .03).
• Among those with better control at baseline, there was greater risk of hypoglycemia despite similar A1C outcomes. For those with A1C <8.5%, incidence rates of severe hypoglycemia were 191 events/100 person years with HDLV-L and 21 events/100 person years for LIS.
• At week 26, the mean ± SD basal insulin doses were 0.36 ± 0.20 U/kg/day for the HDV-L group and 0.43 ± 0.23 U/kg/day for the LIS group. The bolus doses were 0.33 ± 0.17 U/kg/day for the HDV-L group and 0.38 ± 0.21 U/kg/day for the LIS group.

Asked by EBDM™ why the additive appeared to produce a greater effect among patients with poorly controlled glucose, Geho replied via email that the results show that HDV had “physiological effect across the A1C spectrum.” Those with less glycemic control at the outset achieved reductions in A1C and a reduction in time with blood glucose below 54 mg/dL, “all with less insulin dosing,” he said.

“This is the combination of effects that a more physiological form of insulin should provide,” Geho said.

The higher levels of hypoglycemia in group with lower A1C levels at baseline could be due to patients using more basal insulin than needed, and thus Diasome is conducting OPTI-1, an open-label phase 2 study for patients with A1C between 6.5% and 8.5%. This trial is designed to provide additional dosing guidance based on the phase 2 data from ISLE-1, according to a statement from the company.⁴

Although ISLE-1 used HDV with insulin lispro, Geho said, “Diasome has designed HDV to be insulin agnostic, and we believe that HDV is most appropriately viewed by regulators as an adjunctive therapy to insulin.” It is also compatible with pumps, pens, vials, and syringes, he said.

Phase 3 testing will begin in 2020, he said.

References

1. US Food and Drug Administration. What is the pancreas? What is an artificial pancreas? www.fda.gov/medical-devices/artificial-pancreas-device-system/what-pancreas-what-artificial-pancreas-device-system. FDA website. Updated August 30, 2018. Accessed September 26, 2019.
2. Diasome website. www.diasome.com/. Accessed September 26, 2019.
3. Klonoff D, Bode B, Cohen N, Penn M, Geho WB, Muchmore DB. Divergent hypoglycemic effects of hepatic-directed prandial insulin: a six-month phase 2b study in type 1 diabetes. Diabetes Care. 2019; doi:10.2337/dc19-0152.
4. Diasome announces first patient enrolled in phase 2 study of hepatocyte directed vesicle nanotechnology for people with type 1 diabetes [press release]. Cleveland, OH: GlobeNewswire; March 28, 2019. globenewswire.com/news-release/2019/03/28/1781422/0/en/Diasome-Announces-First-Patient-Enrolled-in-Phase-2-Study-of-Hepatocyte-Directed-Vesicle-Nanotechnology-for-People-with-Type-1-Diabetes.html. Accessed September 26, 2019.