Incorporating CA19-9 testing in a prospective study of pancreatic cancer surveillance (PCS) among high-risk individuals (HRIs).

Abstract

10527 Background: For individuals at an increased risk of pancreatic cancer (PC), imaging-based PCS is recommended. However, despite PCS among such individuals, when PC is diagnosed, it is found at an advanced stage in over 25% of individuals. Blood-based biomarkers of PC, such as carbohydrate antigen 19-9 (CA19-9) and HbA1C, can be abnormal prior to radiographic findings among individuals with PC. However, it is unknown whether prospective assessment of these biomarkers would allow for earlier detection or interval PC of PC among HRIs undergoing PCS. Methods: We conducted a single-institution (Dana-Farber Cancer Institute) prospective study of HRIs undergoing imaging-based PCS with annual MRCP and/or EUS (NCT06122896). All participants signed informed consent; the institutional IRB approved the study. CA19-9 was measured at baseline and every 6 months. Individuals with normal imaging but abnormal CA19-9 value underwent additional imaging, and short-interval repeat CA19-9 measurement. Personalized, genotype-adjusted, CA19-9 reference ranges were calculated based on fucosyltransferase (FUT) enzymes FUT3 and FUT2 variant analysis. Herein, we report outcomes from baseline PCS assessment. We measured the specificity of unadjusted and adjusted CA19-9. Results: 231 HRIs (mean age 62, 71% female, 47% with a previous history of cancer) underwent baseline PCS. 105 HRI (46%) had a pathogenic germline variant (PGV) predisposing to PC and met 2023 NCCN eligibility criteria for PCS surveillance, 77 (33%) HRI had a family history of PC without a predisposing PGV; 49 HRI (21%) had other indications for PCS. 153 (66%) HRIs underwent baseline PCS with MRCP [Table]; the remainder underwent an EUS. No HRIs were diagnosed with PC on baseline imaging. 98 (42%) HRI had pancreatic cyst(s) detected on baseline imaging. All abnormal CA19-9 tests (unadjusted = 5, genotype-adjusted = 1) were considered false positives as additional imaging did not reveal additional pancreatic pathology. The specificity of unadjusted and genotype-adjusted CA19-9 was 97.8% (95% CI: 95.0-99.3) and 99.5% (95% CI 97.3-100). Two individuals with concern for a focal lesion on EUS underwent immediate workup with imaging (n=1) and biopsy (n=1), both with reassuring findings. One individual was found to have a paraganglioma. Conclusions: Baseline unadjusted CA19-9 testing led to unnecessary follow-up studies in 2% of HRI undergoing imaging-based PCS. Genotype-adjusted CA19-9 may have superior specificity compared to unadjusted CA19-9. Longitudinal biomarker assessment, including serial CA19-9 and HbA1c, is ongoing. Clinical trial information: NCT06122896 . Characteristic n (%) Baseline imaging MRCPEUS 231 (100)153 (66)78 (34) Individuals with cystic lesion(s) detected 98 (42) Individuals with cyst(s) > 1cm 21 (9) MPD > 4mm 2 (0.9) Elevated unadjusted CA19-9 (> 35 U/mL) 5/229 (2.2) Elevated genotype-adjusted CA19-9 1/205 (0.5) MPD: main pancreatic duct.