Clinical staging in pancreatic adenocarcinoma underestimates extent of disease

Background/objectives: We sought to identify the reliability of AJCC clinical staging was in comparison to pathologic staging in surgically resected patients with pancreatic cancer. Methods: We used the National Cancer Database Pancreas from 2004 to 2016 and evaluated patients who underwent resection for PDAC with all documented components of clinical and pathologic stage. We first evaluated the distribution of overall clinical stage and pathologic stage and then evaluated for stage migration by assessing the number of patients who shifted from a clinical stage group to a respective pathologic stage group. To further characterize the migratory pattern, we assessed the distribution of clinical and pathologic T-stage and N-stage. Results: In our cohort of 28,338 patients who underwent resection for PDAC, AJCC clinical staging did not reliably predict pathologic stage. Stage migration after resection was responsible for discrepancies between the distribution of overall clinical stage and pathologic stage. The predominant migration was from patients with clinical stage I disease to pathologic stage II disease. Most patients with clinical T1 and T2 disease were upstaged to pathologic T3 disease and over half of patients with clinical N0 disease were upstaged to pathologic N1 disease after resection. Discussion: Clinical staging appears to overrepresent early T1, T2, and N0 disease, and underrepresent T3 and N1 disease.