Hepato-pancreato-biliary
Noah STEPHEN Brown, MD
General Surgery Resident
University of Utah Hospital
South Salt Lake, Utah, United States
Noah STEPHEN Brown, MD
General Surgery Resident
University of Utah Hospital
South Salt Lake, Utah, United States
Noah STEPHEN Brown, MD
General Surgery Resident
University of Utah Hospital
South Salt Lake, Utah, United States
Matthew Firpo, PhD
Research Associate Professor
University of Utah, Department of Surgery, United States
Courtney L. Scaife, MD
Professor of Surgery
Huntsman Cancer Institute, United States
In 2018 the AJCC 8th edition introduced the N2 disease designation for staging of pancreatic adenocarcinomas. N2 disease has been shown in several studies to indicate markedly worse overall survival. Our institution has previously reported on a multianalyte serum biomarker panel for early detection of pancreatic adenocarcinoma. This study intended to identify if these pre-operative serum biomarkers may predict post-operative nodal status.
Methods:
We searched our existing institutional pancreas cancer database for patients with histologically confirmed pancreatic adenocarcinoma surgically resected between January 2004 and October 2019. These patients and their biological samples have already been screened using a 31 analyte panel for previous supervised learning to detect early stage disease. Included in the previously collected data about these patients is the number of nodes examined and positive for PDAC in their resected specimens. Given this information we recategorized these patients using the updated AJCC 8th edition introducing N2 disease. The individual analytes were then screened for their ability to distinguish N2 disease.
Results:
N2 disease indicated a significantly poorer post operative overall survival compared to N0 and N1 disease combined with median survival of 29.8 months (IQR 22.6-41.1) in N2 cases versus 53.6 months (IQR 19.2-137.3) in N0/N1 cases (Hazard Ratio 2.295 95% Confidence Interval, 1.234-4.269, p=0.009). Basigin (BSG) was significantly elevated in N2 disease (mean 17.45, SD 13.53) compared to N0 disease (mean 12.09, SD 11.47), p=0.014 by Dunn's test) while Leucine-rich alpha-2-glycoprotein 1 (LRG1) was significantly decreased in N2 disease (mean 3446.21, SD 2719.12) compared to N0 disease (mean 5727.25, SD 3236.40, p=0.025).
Conclusions:
These results suggest that by differentiating N2 from N0 disease, BSG and LRG1 could be useful in preoperatively identifying candidates that would benefit most from resection. This offers a foundation for future studies to combine biomarkers and clinical factors into a machine learning algorithm to reliably distinguish N2 disease in the preoperative setting, which could affect the pre-surgical discussion and provide vital prognostic information to patients.