Predictors for Tumor Recurrence After Primary Definitive Surgery for Oral Cancer

Abstract

The purpose of this study was to identify significant predictors for oral squamous cell carcinoma recurrence. This Ambispective cohort study was performed in consecutive metastasis-free patients treated for oral squamous cell carcinoma with curative intent from 1998 through 2003. Variables included gender, age, tumor site, macroscopic pattern of the lesion, coexisting disorders (diabetes, hepatic and heart disorders, other tumors/diseases), degree of differentiation, and pathologic TNM stage. Tumor recurrence was considered the dependent variable (outcome). The distribution of recurrences was assessed with χ(2) test. Survival times were estimated by Kaplan-Meier curves and differences were examined with log-rank test. Multiple Cox regression study was also performed. The significance level chosen for all tests was P < .05. One hundred eighteen patients entered the study. Tumor recurrence was 44.9% during the follow-up period (10% local, 29.7% regional, and 5% distant). The mean period that had elapsed before recurrence was 15 months (1.5 to 81.8), with most recurrences (66%) during the first year after treatment (84.9% before 2 years). Multivariate Cox regression analysis indicated the presence of a coexisting disorder (P = .022) as the most relevant prognostic factor for relapse, because patients with associated diseases had a 2.43-fold risk of recurrence. Tumor stage (P = .037), degree of differentiation (P = .042), and macroscopic pattern of the lesion (P = .022) were also identified as prognostic factors for relapse. The risk profile for oral cancer recurrence includes patients younger than 60 years with coexisting diseases whose primary tumor occurred as an ulcerated lesion, and diagnosed at an advanced stage with a poorly differentiated tumor.