Biomarker Discovery for Early Detection of Pancreatic Ductal Adenocarcinoma (PDAC) Using Multiplex Proteomics Technology

Early detection of pancreatic ductal adenocarcinoma (PDAC)
can
improve survival but is hampered by the absence of early disease symptoms.
Imaging remains key for surveillance but is cumbersome and may lack
sensitivity to detect small tumors. CA19-9, the only FDA-approved
blood biomarker for PDAC, is insufficiently sensitive and specific
to be recommended for surveillance. We aimed to discover a blood-based
protein signature to improve PDAC detection in our main target population
consisting of stage I or II PDAC patients (n = 75)
and various controls including healthy controls (n = 50), individuals at high risk (genetic and familial) for PDAC
(n = 47), or those under surveillance for an intraductal
papillary mucinous neoplasm (n = 36). Roughly 3000
proteins were measured using Olink multiplex technology and conventional
immunoassays. Machine learning combined biomarker candidates into
4- to 6-plex signatures. These signatures significantly (p < 0.001) outperformed CA19-9 with 84% sensitivity at 95% specificity,
compared to CA19-9’s sensitivity of 53% in the target population.
Exploratory analysis was performed in new-onset diabetes (n = 81) and chronic pancreatitis (n = 50)
patients. In conclusion, 41 promising biomarker candidates across
multiple signatures were identified using proteomics technology and
will be further tested in an independent cohort.