|
Back to 2018 Program and Abstracts
A SOLUBLE PROTEIN SIGNATURE CAN PREDICT POST-SURGICAL SURVIVAL IN PANCREATIC ADENOCARCINOMA
Michael H. Gerber*1, William Gooding2, Daniel Delitto1, Jessica Cioffi1, Bayli DiVita1, Shannon M. Wallet1, Ryan M. Thomas1, Jose Trevino1, Steven J. Hughes1
1Department of Surgery, University of Florida, Gainesville, FL; 2University of Pittsburgh, Pittsburgh, PA
Objective: Overall survival in resectable pancreatic cancer is poor, and prediction of long term survival benefits from surgery remains rudimentary. We hypothesize that a soluble immunological protein signature from the tumor microenvironment can predict survival in surgical patients with pancreatic adenocarcinoma.
Methods: Tissue lysates from resected surgical specimens of pancreatic ductal adenocarcinoma (n=36) were analyzed for concentrations of 36 cytokines, chemokines, and growth factors. Uniquely informative analytes were discovered using a logistic regression analysis. A covariate penalized logistic model using the identified analytes was created. The prediction values were split using the median into high risk and low risk groups. The overall survival was plotted using Kaplan-Meier survival curves which were then analyzed using the log-rank test.
Results: Analysis of the 36 analytes identified 7 informative analytes that were used in the survival algorithm. The two groups consisted of 18 patients each, and the median overall survival in the high risk group was 7 months compared to a median overall survival in the low risk group of 29 months (P-value <0.001) (see figure).
Conclusion: Long term survival post-surgery can be predicted in pancreatic adenocarcinoma using soluble immune related proteins from the tumor microenvironment. This finding raises the potential for the use of protein signatures for a precision approach to patients with pancreatic cancer.
Back to 2018 Program and Abstracts
|
