The prognostic effect of clinical staging in pancreatic adenocarcinoma

BACKGROUND: The importance of pancreatic cancer staging is uncertain. The aim of this report was to evaluate the accuracy of combined standard imaging techniques in predicting the pathologic stage and to evaluate the prognostic effect of clinical staging to identify patient groups in which laparoscopy and laparotomy could be beneficial. METHODS: Fifty-four patients were included in this analysis. The techniques used for clinical staging were endoscopic retrograde cholangiopancreatography, abdominal computed tomographic scan, and ultrasonography. All patients underwent both clinical and surgical/pathologic staging. A comparison was performed between presurgical stage and surgical/pathologic stage. The prognostic effect of different factors on survival was evaluated with both univariate (log-rank) and multivariate (Cox) analysis. RESULTS: Sensitivity and specificity for vascular involvement were 73.9% and 96.3%, respectively. Sensitivity and specificity for nodal involvement were 63.6% and 95.4%, respectively. A total of 33.3% of patients showed a higher than expected pathologic stage, and 3.7% showed a lower than expected pathologic stage, by comparing clinical and pathologic evaluation. A highly significant correlation was observed between clinical T stage (P = .0067) and tumor diameter (P = .0037) and patient survival. Maximal prognostic differentiation was observed by dividing patients into two groups based on imaging results: group A (favorable prognosis) and group B (unfavorable prognosis). The median survival was 25.1 and 8.0 months for group A and B, respectively. Five-year survival was 20.1% and 0%, respectively (multivariate analysis: P = .0007). CONCLUSIONS: Integrated standard imaging studies achieved reasonable diagnostic accuracy in our analysis. A single classification based on clinical stage and tumor diameter evaluated by imaging predicts prognosis in patients with pancreatic carcinoma.