Background: Research in early esophageal adenocarcinoma focused on prediction of lymph node metastases in order to stratify patients for endoscopic treatment instead of esophagectomy. Although distant metastases were described in rates of up to 13% of patients within a follow-up of 3 years, their prediction has been neglected so far.
Methods: In a secondary analysis, a cohort of 217 patients (53 T1a and 164 T1b) treated by esophagectomy was analyzed for histopathological risk factors. Their ability to predict the combination of lymph node metastases at surgery as well as metachronous locoregional and distant metastases (overall metastatic rate) was assessed by uni- and multivariate logistic regression analysis.
Results: Tumor invasion depth was correlated with both lymph node metastases at surgery (τ = 0.141; P = .012), tumor recurrences (τ = 0.152; P = .014), and distant metastases (τ = 0.122; P = 0.04). Multivariate analysis showed an odds ratio of 1.31 (95% CI 1.02-1.67; P = .033) per increasing tumor invasion depth and of 3.5 (95% CI 1.70-6.56; P < .001) for lymphovascular invasion. The pre-planned subgroup analysis in T1b tumors demonstrated an even lower predictive ability of lymphovascular invasion with an odds ratio of 2.5 (95% CI 1.11-5.65; P = 0.028), whereas the predictive effect of sm2 (odds ratio 3.44; 95% CI 1.00-11.9; P = 0.049) and sm3 (odds ratio 3.44; 95% CI 1.00-11.9; P = 0.049) tumor invasion depth was similar.
Conclusions: The present report demonstrates the insufficient risk prediction of histopathologic risk factors for the overall metastatic rate.