Liver-Directed Mealtime Insulin Benefits Some With Type 1 Diabetes

Miriam E. Tucker

October 23, 2019

An investigational liver-targeted formulation of fast-acting insulin may help minimize hypoglycemia in patients with poorly controlled type 1 diabetes, new research suggests.

Findings regarding Diasome Pharmaceutical Inc's hepatic-directed vesicle (HDV) insulin were published online September 24 in Diabetes Care. The study was conducted by David Klonoff, MD, clinical professor of medicine at the University of California, San Francisco, and colleagues.

The product is made of a specially designed matrix that passively attaches free insulin. It provides more normal insulin biodistribution by mimicking the usual physiologic portal vein delivery that occurs in people who do not have type 1 diabetes. It can be mixed with any commercially available short-acting insulin.

Among 141 patients with type 1 diabetes who completed the 26-week phase 2b study, there was no difference in change in A1c level from baseline between the 98 patients who received mealtime insulin lispro (Humalog, Lilly) mixed with HDV (HDV-L) and the 43 patients who received lispro alone.

However, for those with baseline A1c ≥8.5%, use of HDV-L resulted in less time in hypoglycemia and less use of insulin despite the same drop in A1c.

"There was noninferiority on A1c and superiority on hypoglycemia," and this "is important," because people "with type 1 diabetes often have to choose between better control or more hypoglycemia," study coauthor and Diasome chief medical officer Marc Penn, MD, PhD, told Medscape Medical News in an interview.

"The addition of HDV allows for that same half percent improvement in A1c with an actual reduction in hypoglycemia, so it's not a give-and-take with HDV," he observed.

Patients whose A1c level was lower than 8.5% and who received HDV-L actually experienced an increase in time spent in hypoglycemia, but that would be expected, owing to the increased liver bioavailability, Penn explained.

Another study, which is currently underway, is examining the effects of lowering the dose of basal insulin in patients with A1c of 6.5%–8.5% in order to avoid that, he said.

"I think there's a growing understanding of the potential that patients who are poorly controlled" are so "in part because they have a higher hypoglycemia risk, and the choice of better control vs hypoglycemia is a greater issue for them," he said.

From the current study, Penn said, the thinking is that "patients who are better controlled are going to need their basal adjusted first before going on a more potent insulin, whereas those with higher A1c will have more excess glucose around so they won't need that."

"In theory, hepatic specific insulins could be more effective in lowering postprandial insulin vs standard subcutaneous insulin injections," endocrinologist Harvey L. Katzeff, MD, senior medical director at IQVIA, commented to Medscape Medical News.

"However, this study does not confirm this hypothesis, and the differences observed were measurable but not clinically significant," said Katzeff, who spent 10 years at Merck providing medical oversight and monitoring of clinical trials involving diabetes and related conditions.

"Overall, I would say the authors have overestimated the differences between these two insulins and that the data presented only weakly support [their] conclusions."

Was Noninferiority Proven?

The patients in the study were mostly male. Of the HDV-L group, 62% were men, and of the lispro group, 72% were men. The mean baseline age was 46.7 and 44.1 years, respectively, and the mean baseline A1c level was 8.12% and 8.22%, respectively.

The HDV-L used in the study was 1% HDV-bound lispro and 99% unbound lispro. It was formulated by mixing 0.8 mL HDV into 10 mL commercial lispro. The comparator was lispro similarly diluted with water.

Mean change in A1C from baseline to week 26 was –0.09% with HDV-L and –0.16% with lispro, giving an estimated treatment difference of 0.09% for the modified intention-to-treat population.

There were no significant treatment effects from week 0 to week 26 in basal, bolus, or total insulin doses.

The authors say that these data confirm noninferiority on the basis of a prespecified 0.4% margin, but Katzeff disputes this.

"The noninferiority value should be 50% of the treatment effect difference in the study," he said.

"The statistician used 0.4% when the actual treatment effect was 0.16%. They should have used 0.08% as the noninferiority margin and need to reevaluate the data using the correct margin. It may be that they are not noninferior as reported," he said.

Hypoglycemia Effect Differed by Baseline A1c

Overall, there were no significant differences between the two treatment groups in time spent with glucose levels <54 mg/dL, as measured by continuous glucose monitoring during week 26. It was 1.6% for the HDV-L group, vs 1.5% for the lispro group.

However, differences were found in a prespecified subgroup analysis of those with baseline A1c of ≥8.5% vs <8.5%.

At week 26, for those with A1c ≥8.5%, reduction in A1c was similar for both treatment groups (~0.5%), but the HDV-L group used about 25% less bolus insulin compared to the group that used lispro alone.

The HDV-L group also spent less time in hypoglycemia, with blood glucose <54 mg/dL, at 0.7%, compared with 1.2% of time for the group that used lispro alone, at 26 weeks (P = .09).

In contrast, among those with A1c <8.5%, there were no differences in insulin dosage from baseline to week 26 for either treatment.

Among these patients, the lispro group actually spent less time <54 mg/dL at 26 weeks, at 0.6%, compared with 2.0% for those taking HDV-L (P = .16).

Penn, who also practices cardiology at Summa Health in Akron, Ohio, told Medscape Medical News, "I think the most important thing is the idea that insulinization of the liver isn't being done, but it's part of the natural way that insulin is delivered to the body, and restoring that is actually really important."

Diasome is currently in talks with the US Food and Drug Administration about designing phase 3 trials for HDV-L, which will initially focus on patients with type 1 diabetes and A1c >8.5%.

Katzeff believes the results from that trial will need to be more convincing than those of the current trial.

"If the phase 3 data results are similar to the phase 2 data, I do not believe this new insulin preparation would be a significant improvement over the presently available short-acting insulin preparations," he concluded.

Penn countered, "In the study's higher baseline A1c subgroup, the combination of simultaneous A1c decrease from baseline, reduction in bolus insulin dosing, and reduction in time spent in hypoglycemia is unparalleled in long-term type 1 diabetes insulin dosing studies. Given the unprecedented nature of such long-term clinical data, we believe that HDV has the potential to be the first 'no trade-off' mealtime insulin therapy. We intend to further confirm this in Phase 3 testing."

"In lower-A1c subjects who did not lower their insulin dosing, we conclude that a more physiological form of mealtime insulin demands an optimized ratio of basal-to-bolus insulin dosing, and perhaps, a different total amount. The hypoglycemia trends in the lower A1c subgroup motivated us to initiate a low-A1c Phase 2 dose optimization trial that we expect to complete in the first quarter of 2020, presumably to be followed by further testing in Phase 3," stated Penn.

Diabetes Care. Published online September 24, 2019. Abstract

For more diabetes and endocrinology news, follow us on Twitter and Facebook


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: