OCT 15, 2020 10:30 AM CDT

Detection and Treatment Prediction of Pancreatic Cancer Using a Glycan Biomarker​

Speaker

Abstract

CA19-9 is widely known as being the best marker for pancreatic cancer having, on average, a 75% specificity and sensitivity. Three versions of a candidate biomarker related to CA 19-9 (sTRA), detected through screening via on-chip glycon profiling and their value in pancreatic cancer diagnosis are described. The distinct structural components of the sTRA versions, contrary to CA19-9, gives them a stem characteristic, indicating a potential role as subtype markers. Plasma levels of the biomarkers correlated with 45 matched samples, indicating the blood sample can be a good indicator of the primary tumor level. Combining the biomarkers together can detect more types of cancers than each alone, leading to improved sensitivity from 47% to 63%. Additionally, a statistically significant sensitivity increases over CA19-9 alone was noted in a blinded set. This is most rigorous validation so far of improvement upon CA 19-9 using upfront predefined thresholds and classification rules with statistically significant improvement. CA 19-9 provides prognosis value but lacks treatment prediction value. In contract, sTRA, demonstrates value for differentiating subtypes of cancers that is more resistant to chemo.

Learning Objectives:

  • Understand the gaps in pancreatic cancer screening and how to overcome those gaps
  • sTRA demonstrates differentiating value when compared to CA19-9