乌鸦传媒

ORLANDO -- For type 1 diabetes patients, switching from an insulin pump or multiple daily injections to (Afrezza) maintained glycemic control -- for those who used it as intended, the INHALE 3 trial showed.

Overall glycemic control was similar between treatments, with an HbA1c at 17 weeks of 7.6% on the inhaled insulin plus basal insulin degludec (Tresiba) and 7.5% with usual care (P=0.01 for noninferiority), reported Carol J. Levy, MD, of Icahn School of Medicine at Mount Sinai in New York City, at the American Diabetes Association (ADA) annual meeting.

However, secondary outcomes showed that the switch didn't work for everyone. "One size does not fit all," Levy said.

The panel of INHALE 3 investigators suggested that poor candidates might include those patients with a fear of hypoglycemia that keeps them from dosing at bedtime, or who have a lot of auto boluses given by their automated infusion pump suggesting that they're not proactive about management.

Overall, 26% of the inhaled insulin patients had their HbA1c worsen by at least 0.5% compared with 3% on usual care. For patients who had a baseline HbA1c over 7.0%, 21% randomized to inhaled insulin achieved below that goal by 17 weeks but the same number actually had worsening of at least 0.5% after switching to inhaled insulin.

Worsening appeared to be driven by high overnight values in those who skipped their correction dose before bed, explained Grazia Aleppo, MD, of Northwestern University in Chicago. Nighttime time-in-range values were higher for the half of usual care patients who used an automated insulin delivery pump than those who switched to inhaled insulin (61% vs 48%).

Outliers "didn't take their dose of TI [Technosphere inhaled insulin powder], let's be honest," Aleppo said. "There was supposed to be an optimization on the participant's side of taking the second dose of TI after dinner. And most importantly, it is the bedtime correction [that] makes a huge difference. For those who did, they got better results."

Another key takeaway was how to dose inhaled insulin. The product -- based on earlier trials showing superiority of the drug in type 2 diabetes over oral antidiabetic agents alone and in type 1 diseases -- suggests starting patients off at roughly the same inhaled insulin dose as their injected mealtime insulin.

INHALE 3 started patients off with a dose determined by rounding the usual rapid-acting insulin analog to the nearest whole number, multiplying by two, and then rounding down to the nearest 4-unit cartridge of inhaled insulin. Even that approximately doubled dose proved to be an underestimate, though.

"We knew going in that the earlier trials were not using enough of the TI," said Irl B. Hirsch, MD, of the University of Washington in Seattle, speaking at an ADA press conference. "We were ready to be more aggressive with going up on the dose."

In fact, the titrated dose at 17 weeks was 2.5 to three times higher than the baseline rapid-acting insulin analog, he reported at the session. With insulin degludec as the long-acting basal insulin, the basal-to-bolus ratio was about 30:70. "That's not what we teach, but that's what we saw," Hirsch said.

A meal challenge as part of the trial showed that the inhaled insulin reduced postmeal hyperglycemic spikes compared with usual care. Patients need to be willing to give a correction dose 1 to 3 hours after a meal, particularly for a high protein or high fat meal, Hirsch noted.

The findings will help the clinical team know how to guide patients to get the most from inhaled insulin, he concluded. "TI is not for everyone, but for some it is really an important option."

The phase IV randomized controlled trial included 123 adults with type 1 diabetes enrolled at 19 U.S. sites. All participants were using continuous glucose monitoring along with multiple daily injections, an automated insulin delivery system, or a pump without automation.

They were randomly assigned to an insulin regimen of insulin degludec plus inhaled insulin or continuation of usual care. The primary outcome of the trial was at 17 weeks, followed by a 13-week extension phase in which participants in both groups used the degludec-inhaled insulin regimen.

Mean participant age was 45, with a mean diabetes duration of 23 years. Women accounted for 54% of patients, and 89% were white. At baseline, automated insulin delivery pumps were used by 48% of the patients, multiple daily injections of insulin by 44%, a sensor-augmented pump by 6%, and predictive low-glucose suspend systems by 2%.

Comment