World Congress of the International Academy of Oral Oncology edition:1 location:Amsterdam date:17-20 May 2OO7
In the last two decades, an evolution in the definition of the levels of the TNM classification and the stage grouping guidelines has been observed. These changes have been driven by empirical observations. We wanted to evaluate the prognostic effect of these changes in a patient group from a nationwide database in the Netherlands.
Patients and Methods
In a group of 231 patients with major salivary gland carcinoma treated by the members the Dutch Head and Neck Oncology Group with a median follow-up of 68 months and a 5 year disease specific survival of 74% (SE 3%), the stage grouping criteria of the subsequent 1987 through 2002 editions of the UICC TNM Classifications were applied. Also the suggestions to relocate T4N0M0 and T3N1M0 from Stage IV to Stage III disease (Numata et al, 2000) were analysed.
Using Kaplan-Meier survival analysis for overall survival, observations were made on distribution, discrimination, correct ordering and concordance.
Results and Conclusion
Suboptimal distribution and discrimination is observed when classifying patients according to the 1987-1992 UICC guidelines. Using the 1997 UICC guidelines an improved discrimination is observed, but Stage III is found to be nearly empty (3% of patients). Following the proposal of Numata in the patients of the Dutch Head and Neck Oncology group, a clearly better distribution and discrimination results, dealing adequately with the stage III problem in the 1997 classification. The 2002 UICC guidelines result in an upstaging of patients belonging to Stage I in earlier editions, but still leave much to be desired as to discrimination and distribution. A combination of the changes proposed in the 2002 UICC edition and the changes proposed by Numata produces a superior discrimination and distribution in our patient group and seems the way to go in the future.