Tumor Budding Is a Predictor of Nodal Metastasis and Tumor Recurrence in T1 Esophageal Adenocarcinoma
Jonathan A Nowak, Agoston Agoston, Yifan Zheng, Raphael Bueno, Robert D Odze, Amitabh Srivastava. Brigham and Women's Hospital, Boston, MA
Background: Endoscopic mucosal resection (EMR) is increasingly used for management of patients with T1 esophageal adenocarcinoma (EAC). Although the overall outcome is favorable in these patients, a subset shows recurrence and an adverse outcome. The aim of this study was to identify pathologic features that can help predict an adverse outcome in early stage EAC and which can, therefore, be used to optimize patient selection for EMR versus surgery and chemoradiation therapy.
Design: 42 patients with T1 EAC treated by primary surgical resection at a single institution between 1995-2005 were identified. A pre-determined set of pathologic variables were examined for each case including: tumor size, tumor grade, tumor type (intestinal versus non-intestinal), depth of invasion (above first, between first and second, and below second muscularis mucosae), presence of lymphovascular invasion, presence of lymph node metastases, and presence and extent of tumor budding (criteria similar to Ueno et al (2003) used previously in colorectal cancer). Follow up data including overall survival and time to recurrence was obtained by medical chart review, and was analyzed by both a Cox Proportional Hazards Model and Fisher's Exact Test. Correlation between potential predictive variables was evaluated by the Spearman Rank-order Correlation Coefficient.
Results: Mean patient age was 64.4 years (range: 46-87 yrs) and M:F ratio was 9.5:1. 22/42 patients (52%) died during follow up and 4 patients (9.5%) had a documented recurrence (3 distant and 1 combined local and distant). The presence of tumor budding was a strong predictor of tumor recurrence (Hazard Ratio [HR]=14.21 for time to recurrence, p-value 0.022) and the presence of ≥10 buds per 20x field was associated with a nearly 20 fold increase in risk of tumor recurrence (HR=19.99, p-value 0.007). The amount of tumor budding also correlated positively with nodal metastases (Spearman rank-order correlation coefficient 0.6, p-value <0.001). The presence of lymph node metastases was also associated with a decreased time to recurrence (HR=7.75; p-value 0.041). High tumor grade correlated both with reduced overall survival (HR=3.06, p-value 0.015) and with tumor recurrence (Fisher Exact Test p-value 0.012).
Conclusions: Our findings suggest that tumor budding is a strong predictor of both tumor recurrence and nodal metastasis and can be used to identify a subset of T1 EAC patients at high risk for an adverse outcome.
Monday, March 4, 2013 11:00 AM
Proffered Papers: Section D, Monday Morning