Oral 63rd Endocrine Society of Australia Annual Scientific Meeting 2020

Proteomic signatures in early pregnancy to detect women at increased risk of gestational diabetes (#26)

Natassia Rodrigo 1 2 3 , Mark Larance 4 , Sarah Glastras 1 2 3
  1. Endocrinology, Royal North Shore Hospital, St Leonards, NSW, Australia
  2. University of Sydney, Sydney, NSW, Australia
  3. Kolling Institute of Medical Researh, Sydney, NSW, Australia
  4. Charles Perkins Centre, University of Sydney, Sydney, NSW, Australia

Background and Aims:

Gestational diabetes mellitus (GDM) increases pregnancy-related complications, obstetric interventions and adult-onset diabetes, in both mother and child. Established clinical risk factors for GDM lack specificity for its development. The addition of biomarkers may improve the prediction of GDM from early pregnancy. We aimed to determine if serum biomarkers would improve GDM prediction when added to established clinical risk factors.

Materials and Methods:

Pregnant women with at least one clinical risk factor for GDM (e.g. ethnicity, BMI) were recruited, a fasting blood sample collected, and a 75g oral glucose tolerance test performed at 12-18 weeks (blinded) and 24-28 weeks gestation.  Metabolic and lipid profiles were analysed by standard laboratory measures and discovery proteomic profiles were obtained using mass spectrometry. ROC curve analysis (using probability scores) was carried out to determine the utility of adding metabolic markers to predict GDM from early pregnancy.

Results: 23/93 (24.7%) women developed GDM. Using logistic regression (LR), independent clinical risk factors conferring risk of GDM were BMI>30, ethnicity, past history of GDM (P<0.05). Independent metabolic markers at 12-18 weeks associated with GDM included fasting insulin, fasting glucose, 60min glucose on OGTT, and HDL (P<0.01). ROC curve analysis using clinical risk factors alone showed moderate predictability for GDM (AUC 0.735). Addition of these serum markers to the model improved risk prediction markedly (AUC 0.958). Discovery proteomic analyses demonstrate predictive potential for 7 protein markers, including adiponectin and matrix metalloproteinase-3 (MMP3).

Conclusions: Metabolic and proteomic serum markers at 12-18 weeks gestation were distinctly different and predictive of GDM development. Both adiponectin and MMP3 have known roles in diabetes pathogenesis. Further studies are needed to determine if these novel proteomic signatures have clinical utility in stratifying women most likely to develop GDM and benefit from early intervention.