Development and Validation of a Non-Invasive Microbiome-Based Diagnostic Tool for Early Detection of Gestational Diabetes Mellitus

PROJECT SUMMARY Brooklyn InnoSeq (doing business as MaMome) is developing a microbial screening test for the early detection of Gestational Diabetes Mellitus (GDM) to reduce the risk of adverse pregnancy outcomes. GDM is defined as diabetes diagnosed within the mother in the second or third trimester of pregnancy provided that overt diabetes early in pregnancy has been excluded [1]. GDM is the most common medical complication and metabolic disorder of pregnancy with a reported incidence of 18% among pregnant women worldwide [1-4]. The risk of gestational diabetes increases with each pregnancy. The current diagnosis standard for GDM is a 2-step Oral Glucose Tolerance Test (OGTT) administered between weeks 24-28 of gestation [5]. However, the OGTT methodology is a diagnostic tool for GDM and not a tool for the prevention of this disease. Currently, there are no effective clinical tools to accurately predict GDM or to detect latent Type II diabetes in early pregnancy as many women do not receive screenings before they are pregnant [6]. One of our primary innovations has been in the identification of microbial biomarkers associated with GDM in earlier pregnancies through microbiome profiling and machine learning technologies. These biomarkers will provide the basis for developing a clinically viable non-invasive microbiome screening test to predict GDM development significantly earlier than the current OGTT method. The sampling procedure of our screening test does not require any significantly invasive measures, extending its utilization not just in hospitals with healthcare providers, but also to the home via a user-friendly sampling kit. With this approach, we can initiate the appropriate interventions and therapeutics to prevent GDM and subsequent associated complications. Our previous study, consisting of 66 pregnant women, has identified 5 bacteria in the first trimester that are predictive of GDM. Our Phase I project aims to validate these biomarkers with an independent cohort and to refine machine learning approaches based on a newly enrolled cohort. In aim 1, we will prospectively enroll 250 pregnant women, collecting stool samples at the first trimester and at the OGTT diagnosis window. This will be followed by metagenomic whole genome shotgun sequencing (mWGS) to identify microbiome species and (or) microbiome genes that are associated with GDM. In aim 2, we will develop a predictive model for GDM using machine learning approaches. Our ultimate goal is to create a predictive microbiome screening test to detect GDM within Trimester 1. With approximately 4.6 million annual births in the United States [7], patients and physicians would benefit greatly from more accurate, non-invasive tests at home before GDM is established. We will first target OB/GYNs to use our screening test, focusing on developing relationships in the East Coast US markets as well as partnering with global distributors to enter a larger market.