Comment on Marcoux et al. Varying Impact of Gestational Diabetes Mellitus on Incidence of Childhood Cancers: An Age-Stratified Retrospective Cohort Study. Diabetes Care 2022;45:1177–1183

We were interested by the recent article in Diabetes Care from Marcoux et al. (1), who reported that children exposed to gestational diabetes mellitus (GDM) had 1.47 times higher risk of any cancer (95% CI 1.21–1.79), 1.61 times higher risk of blood cancer (95% CI 1.09–2.36), and 1.48 times higher risk of acute myeloid leukemia (95% CI 0.51–4.25) before 2 years of age. Interestingly, the associations between high cancer risks and GDM faded in older children. In their discussion, the authors proposed that this could be related to fetal exposure to maternal hyperglycemia and hyperinsulinemia. However, the precise mechanisms underlying GDM and the increased risks of childhood cancer remain uncertain.

To investigate possible mechanisms between maternal GDM and childhood blood cancer, 12 GDM patients and 13 healthy pregnant women undergoing cesarean section were enrolled at the Second Xiangya Hospital of Central South University. These two groups were well matched in terms of maternal age, fetal sex, and fetal weight. Umbilical cord blood mononuclear cells and sera were isolated from fresh umbilical cord blood. Albeit with similar cord blood glucose levels, fetal sera from GDM mothers exhibited higher C-peptide levels (286.96 ± 132.42 pmol/L vs. 178.39 ± 60.12 pmol/L, P = 0.045), indicating the existence of hyperinsulinemia in the cord blood of the GDM group.

Furthermore, to explore the effects of maternal GDM on fetal immune cells, umbilical cord blood mononuclear cells from four GDM patients and three matching healthy pregnant women were collected for single-cell transcriptome sequencing. Differential gene expression analysis between the GDM and control groups showed that G0S2, CXCL8, NAMPT, IFITM2, NFKBIA, FCGR3B, PTGS2, SOD2, DUSP1, and SAT1 were the top 10 differentially expressed genes (DEGs) (|log₂ fold change| ≥0.585 and P < 0.05). According to KEGG enrichment analysis, DEGs were enriched primarily in the insulin signaling pathway. Surprisingly, cancer-related pathways, such as pathways in cancer and acute myeloid leukemia, were also significantly enriched. In addition, signaling pathways of mitogen-activated protein kinase, WNT, Janus kinase, STAT, and ERBB, all of which were associated with leukemia (2–4), were also demonstrated to be upregulated by KEGG pathway enrichment analysis. After subcluster analysis, only DEGs of monocytes and granulocytes were enriched in pathways in cancer, with CXCL8, NFKBIA, and PTGS2 being the common DEGs between the GDM and control groups. Notably, both CXCL8⁺IL1B⁺ monocytes and CXCL8⁺CD16⁺ granulocytes were increased in the GDM group (26.8% vs. 19.4% and 29.5% vs. 7.7%, respectively). Subsequent flow-cytometric results further confirmed the expansion of CXCL8⁺IL1B⁺ monocytes in the GDM group and showed an obvious increase of CXCL8⁺IL1B⁺ monocytes in the high-glucose stimulation experiment. Collectively, our findings suggest that maternal GDM reprograms the fetal monocytes and granulocytes with a profile of highly expressed CXCL8 (5), which may be associated with an elevated risk of childhood blood cancer.

We agree with the hypothesis of Marcoux et al. (1) that maternal hyperinsulinemia (rather than hyperglycemia) plays a role in the association between GDM and pediatric cancer risk. Furthermore, our results imply that GDM promotes the maternal–fetal proinflammatory activation of monocytes and granulocytes, which may be involved in the pathogenesis of childhood blood cancer, particularly acute myeloid leukemia.