HAMBURG, Germany — A hybrid closed-loop insulin therapy system significantly improved glycemic control among pregnant women with type 1 diabetes, the randomized AiDAPT trial showed.

Meeting the primary endpoint, the mean percentage of time that the maternal Glucose level was in the target range was 68.2% with the hybrid closed-loop system and 55.6% with standard insulin therapy (mean adjusted difference 10.5 percentage points, 95% CI 7.0-14.0, P

This equated to an additional 2.5 hours per day spent in the pregnancy-specific target glucose range, the authors noted in the New England Journal of Medicine, where the study was simultaneously published.

“We would recommend that hybrid closed-loop from here on in should be offered to all pregnant women with type 1 diabetes, definitely during pregnancy and … ideally starting pre-pregnancy,” Murphy said. “Whether we use faster-acting insulins, whether we use pumps, and despite the massive improvements we’ve seen in clinical care that can be attributed to real-time [continuous glucose monitoring] — none of those things by themselves are enough to get enough women to target early enough in pregnancy.”

In an accompanying editorial, Satish Garg, MD, and Sarit Polsky, MD, MPH, both of the University of Colorado in Denver, agreed that most pregnant women with type 1 diabetes struggle to reach target glucose goals, noting that “the importance of these findings cannot be understated.”

Beyond more time spent in target glucose range, closed-loop users also reaped other benefits compared with standard care, which involved multiple daily injections or use of an insulin pump:

• Less time in hyperglycemia (>140 mg/dL): mean difference of -10.2% (95% CI -13.8 to -6.6)
• More overnight time in range (11 p.m. to 7 a.m.): mean difference of 12.3% (95% CI 8.3-16.2)
• Lower HbA1c levels: mean difference of -0.31% (95% CI -0.50 to -0.12)

A total of six severe hypoglycemia events were reported in closed-loop users versus five in the standard-care group. Each group also had one diabetic ketoacidosis event. There were 12 device-related adverse events in the closed-loop group — seven related to closed-loop therapy — that occurred, which included things like incorrect insulin boluses and overnight loss of Bluetooth connectivity.

For reasons “that were not entirely clear,” Murphy also said there was a slightly earlier gestation of delivery in the intervention arm, with babies delivered on average 4.5 days earlier.

The trial used the CamAPS FX hybrid closed-loop system, currently approved for use during pregnancy in the U.K. and the European Union. The system automatically adjusts insulin delivered via an insulin pump using real-time glucose-sensor measurements. It is not currently available in the U.S.

The CamAPS FX system allowed pregnancy-specific glucose targets of 81 to 90 mg/dL, “which were the targets used from 16 to 20 weeks of gestation onward,” Garg and Polsky pointed out.

Other hybrid closed-loop systems have FDA approval, like Insulet’s Omnipod 5 and Medtronic’s MiniMed 780G, but aren’t approved for use during pregnancy and can’t be adjusted low enough for pregnancy-specific targets.

What distinguishes the CamAPS FX system’s algorithm from all others is this lower personal glucose target, Murphy noted. “We recommended that women [in the trial] use a personal glucose target of 5.5 at starting, reducing to 5.0 by the end of the first trimester, and between 4.5 and 5.0 as soon as they were safe from risk of hypoglycemia and happy to do that. There’s no other commercially available system that can do that.”

Garg and Polsky pointed out that while many clinicians use closed-loop systems off-label during pregnancy, “manufacturers of closed-loop systems available in the United States have made no commitments to doing proper prospective studies in this high-risk group.”

“Regulators (such as the FDA) may need to consider other ways (e.g., studies of real-world data) to approve closed-loop systems for high-risk groups (e.g., toddlers, older adults, and pregnant persons) and to encourage device manufacturers to perform necessary clinical trials in these groups,” they added.

The 124-participant open-label AiDAPT trial measured glucose levels from 16 weeks’ gestation through delivery. At the time of randomization, baseline HbA1c was 7.7% (an A1c over 6.5% was needed for inclusion). Mean participant age was 31, and over 90% were white. More than half had diabetic retinopathy, and less than 10% had nephropathy and/or neuropathy.

At baseline, 73-76% of patients were using the Abbott FreeStyle Libre continuous glucose monitor, 20-23% used a Dexcom device, and 2-7% were using a Medtronic device. More closed-loop participants were using an insulin pump at baseline (52% vs 40%), whereas the opposite was true for multiple daily injections (44% vs 59%). Only a few used an automated insulin delivery system. “Approximately half of the participants were using multiple daily injections, and that is really important — so half of these trial participants were pump naive,” said Murphy.

Editorialists Garg and Polsky pointed out that participants underwent randomization around 11 weeks’ gestation, missing “the crucial period of organogenesis.”

“Evidence suggests that a higher percentage of time in the target glucose range in the first 10 weeks of gestation reduces the risk of infants being born large for their gestational age, which raises the question of whether earlier closed-loop initiation could further affect neonatal health,” they wrote.

Murphy also noted that the glycemic benefits were quickly apparent. “We recruited women at 10 weeks [gestation], randomized at 11, trained at 12. By 12 weeks, we already had a 5% difference between the intervention and the control group,” she said.