Aims: To evaluate the current UICC/AJCC Staging System for nasopharyngeal carcinoma and to search for ways of improving the system.
Materials and methods: This is a retrospective analysis of 2687 consecutive patients treated in five public centres in Hong Kong during the period 1996-2000. All patients were staged by computed tomography, magnetic resonance imaging, or both. The prognostic significance of the current stage assignment on various aspects of tumour control was evaluated.
Results: T-category, N-category and stage-group were all significant prognostic factors for major end points (P < 0.01). However, the distinction of prognosis between Stage I and II was insignificant (5-year cancer-specific survival being 92% vs 95%; P = 0.13). Multivariate analyses (corrected for age and sex) revealed lack of significance between T2a and T1 in hazards of local and distant failures, N3a and N2 in distant failure and subgroups of T1-2N0 in cancer-specific deaths. Corresponding down-staging of T2a to T1, N3a to N2, and subgroup T2N0 to stage I, resulted in more even and orderly increase in the hazard ratio of cancer-specific deaths (from 1 for stage I to 1.98 for II, 3.5 for III, 6.08 for IVA and 8.62 for IVB), better hazard consistency among subgroups of the same stage and more balanced stage distribution.
Conclusions: The current UICC/AJCC Staging System could be further improved by the modifications suggested; validation of the current proposal by external data is urgently awaited.