乌鸦传媒

This article is a collaboration between 乌鸦传媒 and:

ORLANDO -- Human neutral protamine Hagedorn (NPH) insulin and long-acting basal insulin came out on par for serious hypoglycemia risks in an observational study.

Hypoglycemia-related emergency department (ED) visits or hospital admissions occurred in 11.9 per 1,000 person-years among people whose first insulin was basal -- glargine (Lantus) or detemir (Levemir) -- versus 8.8 per 1,000 person-years in those who initiated on NPH (P=0.07), Kasia Lipska, MD, MHS, of Yale University in New Haven, Connecticut, and colleagues reported at the American Diabetes Association (ADA) annual meeting.

A propensity score-matched analysis likewise showed no significant difference, whereas diabetes control at 1 year after initiation was modestly but significantly better with NPH (adjusted differences in differences −0.22%, 95% CI −0.09% to −0.37%).

"These findings suggest that the use of basal insulin analogs in usual practice settings may not be associated with clinical advantages for these outcomes," the researchers concluded in the simultaneously published paper in the.

That message goes further than an ADA working group looking into on soaring insulin prices, which had concluded that human insulin "may be an appropriate alternative" to the basal insulins at up to 10 times their list price, noted an accompanying editorial by Matthew Crowley, MD, MHS, and Matthew Maciejewski, PhD, both of the Durham VA Health Services Research and Development Center of Innovation in Durham, North Carolina.

"Along with pursuing measures to help control increasing insulin costs, reemphasizing NPH insulin as a front-line insulin option for most patients with type 2 diabetes could begin to bend the insulin cost curve for patients and insurers," they wrote.

As insulin analog prices have roughly tripled over the past decade or so, NPH has remained fairly cheap without requiring a prescription. "It's a miracle that such a drug exists in America," commented John Buse, MD, PhD, of the University of North Carolina at Chapel Hill.

"The insulin analogs have been shown in clinical trials to reduce the risk of hypoglycemia but modestly... and largely around the routine hypoglycemia, not as much data around severe hypoglycemia," he said, except with insulin degludec (Tresiba) for which the DEVOTE trial showed about a 40% reduction versus insulin glargine.

The study couldn't rule out a small difference in severe hypoglycemia or a difference in less severe events that didn't trigger medical care. But as Buse, who was not involved in the study, noted,"you'd need a lot of events, and particularly exposure to degludec, to really understand what the benefits are. If I had a patient where resources were limited, I would not feel at all bad about using NPH insulin."

The study included 39 hypoglycemia-related ED visits or hospital admissions among 1,928 patients who initiated long-acting insulin analogs and 354 among 23,561 patients in the type 2 diabetes population of Kaiser Permanente of Northern California. Degludec was not included in the study as it was approved after the study period, which ran from Jan. 1, 2006, through Sept. 30, 2015.

The average follow-up was 1.7 years, with patients censored at death, loss of health plan coverage, or change in insulin treatment.

Notably, "these results may not generalize to other health care systems because, counter to current prescribing trends in the United States favoring basal insulin analogs, the rate of NPH insulin initiation in the study population was particularly high (92%)," the editorialists wrote. "Clinicians' greater familiarity with NPH insulin in this health care system may have reduced the likelihood of hypoglycemia among NPH insulin users, which may partly explain the low event rates."

Also, the mean baseline hemoglobin A1c level was 9.4%. This may limit generalizability to patients starting with better-controlled type 2 diabetes or those using complex basal-prandial insulin regimens, the editorial added.

Other limitations included lack of data on hypoglycemic events that did not trigger health care use, which can still be important to patients, and some imbalance in baseline characteristics.