Glycemic Variability in Pregnant Individuals Using Assisted Hybrid Closed-Loop Therapy Versus Sensor-Augmented Pump Therapy

Introduction

Pregnancies complicated by type 1 diabetes are subject to variations in insulin sensitivity from fluctuating maternal and placental hormone levels throughout gestation. ¹ Hybrid closed-loop (HCL) therapy has been shown to increase time in target glucose range and decrease episodes of hypoglycemia in non-pregnant individuals when compared to sensor-augmented pump therapy (SAPT), ² but its use has been rarely studied in pregnancy. ³ Some preliminary HCL studies in pregnancy have shown that the therapy appears to help women attain glycemic goals of pregnancy, but its effects on glucose variability (GV) are not well established.^3,4 In this analysis, we aim to compare the glycemic risk and continuous glucose monitor (CGM) variability metrics between HCL therapy and SAPT use during pregnancy.

Research Design and Methods

The research design for the Pregnancy Intervention with a Closed-Loop System (PICLS) study has been previously described. ⁵ Briefly, the PICLS study was a pilot, two-center, open-label, randomized controlled trial comparing HCL therapy and SAPT in pregnancy and early post-partum using the Medtronic 670G insulin pump and Medtronic Guardian 3 sensor, which, at the time of study start, was the only Food and Drug Administration (FDA)-approved HCL system on the market in the United States. Participants were randomized between 14 and 18 weeks of gestation and used their assigned therapy through four to six weeks post-partum.

The PICLS study’s primary objectives were to compare safety parameters and the main glycemic indices (eg, time below, in, and above range) between groups and have been described in detail elsewhere. ⁵

The secondary objectives of the PICLS study included comparisons of alternative glycemic variables, which we herein present. CGM data were compared between groups for each phase of the study (run-in phase, first trimester, second trimester, third trimester, and post-partum) for the secondary glycemic metrics (low blood glucose index [LBGI], high blood glucose index [HBGI], mean amplitude of glycemic excursions [MAGE], continuous overall net glycemic action [CONGA _n ], and J-index).

Results

There were 23 participants who completed the study (n = 12 SAPT, n = 11 HCL). There were no statistically significant differences in measures of glycemic risk (LBGI and HBGI) or in measures of glycemic variability (MAGE, CONGA, or J-index) observed between the HCL and SAPT groups during any point in the study in pregnancy or post-partum (Table 1).

Table 1.

Measures of Glycemic Risk and Glucose Variability of SAPT and HCL Therapy in Pregnancy.

	SAPT (n = 12)	HCL (n = 12) a	P-value
Measures of glycemic risk
LBGI, mean % (SE)
Run-in phase	5.48 (0.52)	5.67 (0.53)	.80
First trimester	5.40 (0.52)	5.92 (0.53)	.49
Second trimester	4.83 (0.52)	4.09 (0.53)	.32
Third trimester	4.06 (0.52)	3.44 (0.53)	.41
Post-partum	4.86 (0.53)	3.69 (0.53)	.13
HBGI, mean % (SE)
Run-in phase	4.54 (0.6)	5.32 (0.62)	.37
First trimester	5.23 (0.6)	6.12 (0.62)	.31
Second trimester	4.78 (0.6)	5.61 (0.62)	.34
Third trimester	3.57 (0.6)	4.58 (0.62)	.25
Post-partum	5.70 (0.62)	6.42 (0.62)	.42
Measures of glucose variability
MAGE, mean mg/dL (SE)
Run-in phase	84.4 (5.5)	98.8 (5.6)	.08
First trimester	91.9 (5.5)	100.9 (5.6)	.26
Second trimester	85.4 (5.5)	94.5 (5.6)	.25
Third trimester	69.2 (5.5)	81.8 (5.6)	.12
Post-partum	92.9 (5.6)	99.5 (5.6)	.41
CONGA n , mean (SE)
Run-in phase	39.9 (2.0)	42.0 (2.0)	.47
First trimester	41.3 (2.0)	44.3 (2.0)	.30
Second trimester	36.4 (2.0)	40.1 (2.0)	.20
Third trimester	29.6 (2.0)	35.2 (2.0)	.06
Post-partum	40.0 (2.0)	42.6 (2.0)	.38
J-Index, mean mg/dL2 (SE)
Run-in phase	26.2 (2.3)	29.5 (2.3)	.31
First trimester	29.3 (2.3)	32.0 (2.3)	.42
Second trimester	29.5 (2.3)	34.6 (2.3)	.12
Third trimester	25.1 (2.3)	30.7 (2.3)	.10
Post-partum	32.1 (2.3)	38.0 (2.3)	.08

Abbreviations: CONGA_n, continuous overall net glycemic action; HBGI, high blood glucose index; HCL, hybrid closed-loop therapy; LBGI, low blood glucose index; MAGE, mean amplitude of glycemic excursions; SAPT, sensor-augmented pump therapy; SE, standard error.

a

In the HCL group, one participant experienced a fetal loss in the second trimester, leaving data from 11 participants for second and third trimesters and post-partum.

Conclusions

In this pilot study, the fact that glycemic variability of HCL wear in gestation with a system that has a higher-than-recommended glucose target for pregnancy (120 mg/dL study device vs 95 mg/dL fasting glucose target in pregnancy) ⁶ was not worse than that seen for the same pump and CGM being used without automation (SAPT) is reassuring. However, results should be interpreted with caution given the small sample size. In addition, it is difficult to know if the small clinical differences (that were not statistically significantly different) between groups may be important in terms of adverse pregnancy outcomes. In summary, in the PICLS study, the use of the first HCL system with proper oversight and assistive techniques was not associated with significant differences in secondary metrics of glycemia compared to SAPT.