We thank Yin et al. (1) for their hypothesis on mechanisms linking gestational diabetes mellitus with childhood cancer. Yin et al. (1) proposed that hyperinsulinemia, rather than hyperglycemia, explains the association of gestational diabetes with childhood cancer in our study. They also identified genetic pathways by which gestational diabetes could activate fetal monocytes and granulocytes, leading to hematopoietic cancer.
Using a sample of 25 pregnant women, Yin et al. (1) found elevated C-peptide levels in the cord blood of newborns exposed to gestational diabetes compared with unexposed newborns. Glucose levels were similar in both groups. C-peptide is an indicator of insulin secretion. As C-peptides were measured in cord blood originating from the fetus and maternal insulin does not cross the placenta, the authors concluded that fetal hyperinsulinemia rather than hyperglycemia is the factor leading to cancer. However, it is important to recognize that fetal insulin is secreted in response to maternal glucose crossing the placenta (2) and that insulin secretion is not fully effective in the first and second trimesters (3). As the fetal pancreas is not mature until the third trimester (3), fetal hyperglycemia can occur earlier in pregnancy before stabilizing later. Higher insulin secretory activity with normal serum glucose is merely evidence that mature fetal pancreatic β-cells can maintain glucose homeostasis at birth (2,3). Moreover, infants exposed to gestational diabetes are expected to shift to normoglycemia or hypoglycemia rapidly after birth, once maternal glucose influx stops (2). The findings of Yin et al. prove that fetuses were exposed to hyperglycemia but are not sufficient to support a specific pathway leading to cancer.
The gene expression analyses performed by Yin et al. demonstrate how umbilical cord blood may bring new insight on early determinants of childhood cancer. The possibility that gestational diabetes mellitus increases the expression of CXCL8 in fetal monocytes and granulocytes, eventually leading to cancer, appears plausible. CXLC8 is involved in inflammatory and carcinogenic processes, with data indicating that glucose and insulin can elevate CXCL8 in monocytes (4). The complexity of these pathways, one element of which is the role of other genes, will require closer investigation in future studies.
Methodological concerns should, however, be considered. The study by Yin et al. (1) was restricted to women who had caesarean deliveries. Caesarean birth is associated with pediatric cancer even with control for gestational diabetes mellitus (5). It is suspected that infants who bypass the vaginal canal and are not exposed to maternal flora have an altered microbiome, which may affect oncogenic regulation (5). Thus, future studies should use caution to ensure that women are representative of a general population rather than women with caesarean delivery.
In sum, we agree that hyperinsulinemia may play a role in the pathway between gestational diabetes and childhood cancer but cannot exclude the possible involvement of hyperglycemia. Further studies will be needed to follow up on potential genetic pathways involved.
See accompanying article, p. e153.
Article Information
Funding. This work was funded by the Canadian Institutes of Health Research (PCC-170244). N.A. was supported by the Fonds de Recherche du Québec-Santé (296785).
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
