Abstract
Background The aim of our study was to develop a prognostic index score for patients undergoing
surgical resection for esophageal cancer that accurately determines survival with
specific clinicopathological characteristics.
Methods Clinical, histological, and demographical variables of 475 patients were entered
in an univariate and multivariate regression model, followed by individual calculation
of the Prognostic Indicator Score and model validation via simulation.
Results Significant variables included in the scoring system were number of positive lymph
nodes, pT, pL, R, obesity, and American Society of Anesthesiologist classification.
Survival probability and its associated hazard function was significantly different
between the scores, with an increase of hazard ratio ranging from 2.56 (score 2) to
20 (score 6 or higher). Comparing histological cancer entities revealed statistical
significance only between stage IIA versus IIB in squamous cell and stage IIIA versus
IIIB in adenocarcinoma.
Conclusions According to our methodology, an individualized follow-up by each possible score
might allow interdisciplinary selection of patients for treatments based on expected
survival. This may represent a breakthrough in patient selection for currently available
treatments and clinical studies.
Keywords
prognostic index score - esophageal cancer - surgical approach - operative - computing
methodologies - model validation
