Abstract
BACKGROUND: No prognostic biomarker is currently used in clinical management of patients with surgically resected pancreatic cancer other than CA-19-9. In this study, we tested the prognostic value of GATA6 measured with immunohistochemistry and digital assistance. PATIENTS AND METHODS: One hundred and ninety-three patients with resected pancreatic ductal adenocarcinoma from the ESPAC-4 trial of adjuvant gemcitabine and capecitabine were included. Two pathologists independently assessed GATA6 protein expression by immunohistochemistry in tissue microarray cores, manually and with digital assistance. Overall survival was compared across GATA6 levels using multivariate Cox proportional hazard regressions, with exploratory analyses evaluating recurrence-free survival and differential treatment effects. RESULTS: Interobserver concordance improved with digitally assisted scoring (kappa 0.72 versus 0.25, P < 0.001). Median overall survival was 24.3 months [95% confidence interval (CI) 19.2-32.1 months] with low GATA6 expression versus 35.2 months (95% CI 29.9-53.0 months) with high GATA6 expression (adjusted hazard ratio 1.60, 95% CI 1.08-2.38, P = 0.02). Similar results were observed for recurrence-free survival (adjusted hazard ratio 1.45, 95% CI 0.99-2.14, P = 0.06). GATA6 expression was not associated with differential treatment effects. CONCLUSIONS: GATA6 expression measured by immunohistochemistry with digital assistance was a prognostic biomarker among people with pancreatic adenocarcinoma treated with surgical resection and adjuvant chemotherapy.