Towards Precision Medicine for Diabetes in Pregnancy

Gestational diabetes mellitus (GDM) affects up to one in six pregnancies in
Europe,(1) and close to one in two pregnancies in women with obesity.(2) GDM
increases the risk for pregnancy complications including preeclampsia,
C-section and birth injuries. Moreover, impairments in glucose metabolism
during pregnancy pertain after pregnancy(3-5) and increase the risk for obesity
and type 2 diabetes mellitus in both the mother(6, 7) and the baby.(8-11)
In 2016, Powe and Hivert(12) demonstrated that GDM is heterogeneous in its
pathophysiology and outcomes. In their cohort study, among the GDM-patients,
30% had impaired glucose-stimulated insulin secretion (pancreatic β-cell
function), but were insulin sensitive (herein referred to as *GDM-Secretion*),
whereas 50% were insulin resistant, ie low insulin sensitivity, but had normal
insulin secretion (*GDM-Sensitivity*). In 2019, we have confirmed heterogeneity
of GDM in a Dutch cohort (at MUMC), with comparable distributions (20%
GDM-Secretion, 50% GDM-Sensitivity).
To date, the treatment of GDM is universal for all GDM patients. First
hyperglycemia is targeted by lifestyle modification including dietary
modification. For patients with persistent hyperglycemia, pharmacological
intervention is initiated using metformin, an insulin-sensitizer. In case of
insufficient repression of hyperglycemia by metformin, insulin is supplemented.
Insulin is the last step of treatment because it can cause hyperinsulinemia and
hypoglycemia in the mother and the infant. Metformin on the other hand passes
the placenta and can have long-lasting effect on the infants metabolism.
Importantly, the choice and efficacy of treatments is questioned because
maternal, infant and long-term offspring outcomes for GDM are significantly
worse as compared to women without GDM-diagnosis(13). Specifically, for
patients requiring additional insulin-therapy, which is 15-46% based on current
literature, delayed regression of hyperglycemia may adversely affect neonatal
outcome. We have recently demonstrated that the total hyperglycemic exposure
during pregnancy (as AUC) is most relevant towards infant outcomes that
spot-assessments of hyperglycemia,(14) supporting the notion that a delay in
appropriate treatment will likely contributes to deleterious metabolic
imprinting of the fetus.
Unequivocally, authors of clinical studies,(15-19) meta-analysis(20-22) and
practical guidelines(23, 24) call for interventions more personalized to the
patient, but surprisingly little guidance is offered as to how personalization
should be attempted. In most reports, personalization of lifestyle modification
refers to modulation of the intervention to facilitate adherence and thereby to
increase the effect sizes on gestational weight gain or diet quality. The
effect sizes of such intervention are small and poorly effective in reducing
GDM. To our knowledge, only two studies (in the US) are currently performed to
consider the heterogeneity of GDM in its treatment; both are currently ongoing,
one using dietary intervention (PI: Powe, NCT04187521) and one pharmacological
intervention (PI: Feghali NCT03029702), yet, as far as published records show,
with different study designs, in different health acre contexts (more frequent
insulin resistance and insulin-treatment, poorer health care), and in different
study populations, ie US vs NL, likely with very different lifestyles.

In the present study, we propose to investigate the efficacy of pharmacological
treatment of GDM (using metformin) by the pathophysiological cause of
hyperglycemia, namely reduced insulin sensitivity or reduced insulin secretion.

Primary Objective: to assess differences in efficacy of metformin-treatment in
GDM patients with insulin resistance vs low insulin secretion on glucose control
Secondary Objective(s): to assess differences in efficacy of
metformin-treatment in GDM patients with insulin resistance vs low insulin
secretion on maternal and infant pregnancy outcomes, including but not limited
to gestational weight gain, complications during delivery, and lfiestyle
factors.

We will use an intuitive prospective, observational study design to compare
efficacy of metformin-treatment in GDM-patients with low insulin sensitivity to
GDM-patients with low insulin secretion. Participants will be recruited at
24-28 weeks gestation, when GDM is assessed in clinical practice,(2, 25)
followed-up at a second visit late in pregnancy (35-37 weeks gestation) until
their 6-week postpartum control visit. Recruitment will be performed at the
Gynecology clinic of Zuyderland Medical Center. Outcomes will be assessed at
both pregnancy visits, and selected outcomes again at the postpartum visit.

The associated risks are minimal, because treatment follows clinical practice.
The addition catheters being placed carry the risk of inducing slight bruises.
Given the observational nature of this study on patients treated per routine
practice, study-participation invokes no group-related risk.