461: Novel protein biomarker combinations that predict preterm birth

Abstract

Preterm birth (PTB) remains a substantial cause of perinatal mortality and morbidity. Existing methods of preterm birth prediction and current biomarkers fail to identify the majority of asymptomatic women who will deliver prematurely. We propose that select protein biomarkers detected in the cervicovaginal fluid (CVF) are reflective of spontaneous preterm birth. Our aim was to develop and validate an immunoassay-based test composed of a combination of novel protein biomarkers with potential to predict preterm birth (<37 weeks) from mid-trimester. : Samples of CVF from asymptomatic pregnant women were collected between 19 and 26 weeks of gestation and subsequent pregnancy outcomes were recorded including gestation at delivery. The majority of these women were considered to be at risk of preterm birth, including women with a past history of spontaneous preterm birth, previous cervical surgery and twin pregnancy. Exclusion criteria were cervical cerclage and progesterone therapy. CVF samples were obtained from women attending their routine antenatal care. The study included a total of 90 samples from 76 women, of whom 28 spontaneously delivered <37 weeks and 48 delivered at term. The samples were split into a discovery cohort of 35 term and 22 preterm cases and a validation cohort of 13 term and 6 preterm cases. A total of 10 biomarkers were screened by immunoassay and assessed for their ability to predict PTB based on a unique algorithm of biomarker sub-panels. The optimal multivariate model based on specific biological pathways, such as inflammation, hormonal and epithelial-mesenchymal transition, produced mini biomarker panel tests. When combined these achieved a sensitivity of 81% and specificity of 91% for the initial discovery cohort. These results were confirmed in the validation cohort. These results once re-validated in a larger cohort, have the potential of generating a novel test that would predict preterm birth weeks or even months prior to any clinical presentation for risk of preterm birth. Furthermore, the specific pathways represented by these biomarkers may suggest options for future preventative treatments for preterm birth.