Daniel M. Keller, PhD

October 09, 2012

October 9, 2012 (Berlin, Germany) — In a randomized, double-blind, comparator-controlled, 2-way crossover trial, 2 rapid-acting insulin analogs (lispro or aspart) formulated with recombinant human hyaluronidase ("PH20") improved postprandial glycemic control when compared with insulin lispro alone in the treatment of type 2 diabetes mellitus (T2DM). In addition, glycated hemoglobin (HbA1c) levels were noninferior between groups when compared with the standard of care.

Speaking in a poster session here at the European Association for the Study of Diabetes (EASD) 48th Annual Meeting, senior author Douglas Muchmore, MD, vice president for endocrinology clinical development at Halozyme Therapeutics in San Diego, California, reported that patients receiving the new formulation experienced a similar rate of hypoglycemia, body weight change, and injection site pain as with insulin lispro alone.

This phase 2 study involved 121 T2DM patients who underwent titration with insulin glulisine for 4 to 8 weeks. They were then randomly assigned to insulin lispro (lispro), lispro + PH20, or insulin aspart (aspart) + PH20 for 12 weeks in treatment period 1. In treatment period 2, the lispro patients were crossed over to lispro + PH20 or to aspart + PH20. The patients initially on lispro + PH20 or aspart + PH20 were crossed over to lispro alone.

PH20 depolymerizes interstitial hyaluronan in the skin, which is a barrier to the dispersion of drugs injected or infused into the skin. Dr. Muchmore explained that rapid dispersion of insulin enhances its dissociation kinetics. Hyaluronidase also allows it to access a larger capillary bed area and accelerates its absorption into the systemic circulation.

Participants with T2DM had to be on a basal/bolus insulin regimen to be eligible for the trial, with an HbA1c of 7.0% to 8.5% and with a fasting blood glucose level of 80 to 120 mg/dL. The average age was 58.7 years, 60% of participants were male, and participants were overweight on average (body mass index of 34.7 kg/m2).

During the trial, all patients received insulin glargine twice daily, with a postprandial glucose target of less than 140 mg/dL. For the analysis of results, the lispro + PH20 and aspart + PH20 groups were combined as an "analog-PH20 group."

Analog-PH20 Improves Postprandial Control, Noninferior for HbA1c

HbA1c was well controlled with either lispro or analog-PH20, and there was no difference between the groups. Both groups at baseline had an HbA1c of 7.13 ± 0.49%. At the study endpoint, HbA1c for the lispro group was 6.66 ± 0.50%, and for the analog-PH20 group, 6.64 ± 0.63%. These results met a predefined criterion of noninferiority with a margin of 0.4% difference in change between the groups from baseline.

Hypoglycemia rates did not differ between the lispro and analog-PH20 groups, and no instances of severe hypoglycemia occurred. The lispro group experienced 7.66 glucose values of 70 mg/dL or less per 4-week period vs 7.92 events/4-week period in the analog-PH20 group (P = .41). For glucose less than 56 mg/dL, 1.78 events/4-week period and 1.99/4-week period (P =0.25) were experienced respectively.

Adverse events occurred in 42.9% in the lispro group vs 52.9% in the analog-PH20 group; gastrointestinal disorders affected 7.6% in the former group and 14.9% in the latter. For other adverse events, the groups were well balanced. The most common were infections, affecting about 24% in each group. About 2% of participants in each group experienced moderate or moderately severe injection site pain, and no reports of severe pain occurred.

Postprandial glucose excursions were more limited with analog-PH20 compared with lispro alone at breakfast (P < .001), dinner (P = .033), and overall (P < .001), but not at lunch (P = .51). Analog-PH20 reduced excursions by 21% to 23% after breakfast, dinner, and overall.

Basal, prandial, and overall daily insulin doses did not differ among the groups. Analog-PH20 allowed more patients to reach postprandial glucose goals at 1 and 2 hours, although the difference was significant only at 2 hours (28% with lispro and 45% with analog-PH20, P < .001).

Weight changes also did not differ between groups (gain of 3.35 - 3.44 lb).

The 2 treatments had similar immunogenicity profiles, with very little change from baseline in anti-insulin antibodies. At baseline 4/117 patients had anti-PH20 antibodies, with a geometric mean titer of 1:12. At the end of the study, 1 additional patient had anti-PH20 antibodies, with a geometric mean titer of 1:15.

Novel Approach

Graham Ellis, MBChB, specialist physician in metabolic endocrine medicine at the Helderberg Clinical Trials Centre in Cape Town, South Africa, commented to Medscape Medicine News that the PH20 approach is novel and may help to make the response to insulin more predictable.

"If you get a faster, shorter-acting insulin out of it, then it will have some clinical relevance. That's the main aim of it," he said.

"I think these things are important. One is trying to niche insulins to certain types of patients to get a more predictable response, and that's often the limiting thing when patients are on multiple daily dose injections," he explained. "The shorter acting [the insulin], you're going to get a more predictable response usually, and you can time that better with your meal. So I think that's kind of the paradigm you're trying to work in."

Dr. Muchmore told Medscape Medical News that the effects of the enzyme are very transient. "The half-life of the enzyme in rat skin is about 30 minutes, so the enzyme itself disappears very quickly," he said. "There is no systemic exposure to the enzyme because its half-life in the circulation is about 30 seconds." He said hyaluronan is a very active metabolic compartment, with complete restitution occurring in about 18 to 36 hours.

Halozyme has completed 2 phase 2 trials and has also investigated PH20 in type 1 diabetes; phase 3 trials will not proceed until Halozyme can find a partner who is interested in codevelopment of the drug.

The study was supported by Halozyme Therapeutics. Dr. Muchmore is an employee of the company. Dr. Ellis is an employee of a company that does clinical trials for various large pharmaceutical companies, including Amgen, Eli Lilly, Merck, Novo Nordisk, and Novartis, but not for Halozyme, and he had no participation in this trial.

European Association for the Study of Diabetes (EASD) 48th Annual Meeting: Abstract 907. Presented October 2, 2012.