The FDA Endocrinologic and Metabolic Drugs Advisory Committee voted 8-4 for the approval of insulin degludec and insulin degludec/aspart for the treatment of type 1 and type 2 diabetes. In addition, the committee voted unanimously that a cardiovascular outcomes trial be conducted to determine whether the cardiovascular signal observed with these drugs is real.
“I think this form of insulin is addressing a true need, and I think that if in clinical use the efficacy data that we’ve seen hold up in practical clinical use it’s likely to be useful and broadly used as a form of insulin in diabetes care. But, we do have alternatives now,” Robert J. Smith, MD, professor of medicine at Alpert Medical School of Brown University, who voted against approval, told the committee. “This would be an improvement, and while likely a broadly applicable improvement, it’s not of such magnitude that it merits taking on a significant level of risk. In regard to the CV risk signal, I acknowledge there’s a lack of certainty about it, but given the importance of CVD in type 1 and type 2 diabetes, I just personally feel a level of discomfort in supporting approval without further addressing the reality of that issue.”
Both drugs are developed by NovoNordisk.
Efficacy of study drugs
When the new drug applications were filed, 41 studies of insulin degludec and 21 studies of insulin degludec/aspart had been completed. Six studies of degludec and two of degludec/aspart were ongoing.
Eleven therapeutic confirmatory phase 3 trials of insulin degludec were conducted. Nine of these trials studied once-daily dosing, and in these trials, insulin degludec was noninferior to comparison insulins at reducing HbA1c, which was the primary endpoint. Additionally, insulin degludec was associated with improvements in fasting plasma glucose (FPG) compared with other insulins, and demonstrated statistical significance in five of the nine trials.
Similarly, in all five insulin degludec/aspart phase 3 trials, insulin degludec/aspart, whether given once or twice daily, was noninferior to other insulins in reducing HbA1c (primary endpoint).
In addition, insulin degludec/aspart twice-daily lowered FPG significantly more than biphasic insulin aspart 30 twice-daily. In one trial, however, FPG was significantly lower with insulin glargine vs. insulin degludec/aspart once-daily. Improvements in glycemic control were achieved with similar doses of the study drug and comparator insulins.
Kenneth D. Burman
“As a general comment, everyone would agree that hypoglycemia is an important issue to patients and clinicians, and it’s an important issue with regard to potential cardiovascular morbidity in the future, and something we want to avoid,” Kenneth D. Burman, MD, chief of the endocrine section at Washington Hospital Center and acting chairperson of the meeting, said in a summary.
The prespecified major adverse CV events composite endpoint for both insulin degludec and insulin degludec/aspart was CV death, stroke and acute coronary syndrome (myocardial infarction and unstable angina pectoris).
Eighty patients from all 16 trials experienced treatment-emergent major adverse CV events, with similar incidence rates for both insulin degludec and insulin degludec/aspart and comparator drugs. According to NovoNordisk, there was no consistent pattern in the estimated HRs for time to first major adverse CV event — some favored the study drugs and some favored the comparators.
An additional post-hoc analysis of major adverse CV events was conducted, based on a request from the FDA. After excluding unstable angina pectoris from the major adverse CV events composite endpoint definition, the estimated HR for insulin degludec, insulin degludec/aspart and the comparator drugs increased to 1.393 (95% CI, 0.757-2.565) based on 54 major adverse CV events (39 with the study drug and 15 with the comparator).
“Although I can’t reconcile some of the CV outcomes and this may be a signal, it may be noise; I think we owe it to ourselves and our patients to find out,” Thomas J. Weber, MD, training program director of endocrinology, diabetes and metabolism at Duke University Medical Center, said during the first vote on the CV outcomes trial.
“We are concerned, as in any CV trial, especially one not designed to look at CV endpoints, that there really is no understanding of lipids, C-reactive protein or any other analyses of cardiac echos or carotid sonograms that might be useful in future studies,” Burman said on behalf of the panel. “We’re skeptical to some degree about the results but don’t think they are definitive, and further studies ought to be performed.”
Burman voted in favor of approval of insulin degludec and insulin degludec/aspart.
Although the FDA is not required to follow the recommendation of the committee, it usually does.