PURPOSE: To use magnetic resonance imaging to re-evaluate and improve the 6th edition of the International Union Against Cancer/American Joint Committee on Cancer staging system for nasopharyngeal carcinoma. METHODS AND MATERIALS: We performed a retrospective review of the data from 924 biopsy-proven nonmetastatic nasopharyngeal carcinoma cases. All patients had undergone magnetic resonance imaging examinations and received radiotherapy as their primary treatment. RESULTS: The T classification, N classification, and stage group were independent predictors. No significant differences in the local failure hazards between adjacent T categories were observed between Stage T2b and T1, Stage T2b and T2a, and Stage T2b and T3. Although the disease failure hazards for Stage T1 were similar to those for Stage T2a, those for Stage T2b were similar to those for Stage T3. Survival curves of the different T/N subsets showed a better segregation when Stage T2a was downstaged to T1, T2b and T3 were incorporated into T2, and the nodal greatest dimension was rejected. The disease failure hazard for T3N0-N1 subsets were similar to those of the T1-T2N1 subsets belonging to Stage II; the same result was found for the T4N0-N2 subsets in the sixth American Joint Committee on Cancer staging system. However, the staging system we propose shows more consistent hazards within the same stage group and better survival discrimination among T categories, N categories, and overall stages. CONCLUSION: Using the 6th American Joint Committee on Cancer staging system produces an acceptable distribution of patient numbers and segregation of survival curves among the different stage groups. The prognostic accuracy of the staging system could be improved by recategorizing the T, N, and group stage criteria.
PURPOSE: To use magnetic resonance imaging to re-evaluate and improve the 6th edition of the International Union Against Cancer/American Joint Committee on Cancer staging system for nasopharyngeal carcinoma. METHODS AND MATERIALS: We performed a retrospective review of the data from 924 biopsy-proven nonmetastatic nasopharyngeal carcinoma cases. All patients had undergone magnetic resonance imaging examinations and received radiotherapy as their primary treatment. RESULTS: The T classification, N classification, and stage group were independent predictors. No significant differences in the local failure hazards between adjacent T categories were observed between Stage T2b and T1, Stage T2b and T2a, and Stage T2b and T3. Although the disease failure hazards for Stage T1 were similar to those for Stage T2a, those for Stage T2b were similar to those for Stage T3. Survival curves of the different T/N subsets showed a better segregation when Stage T2a was downstaged to T1, T2b and T3 were incorporated into T2, and the nodal greatest dimension was rejected. The disease failure hazard for T3N0-N1 subsets were similar to those of the T1-T2N1 subsets belonging to Stage II; the same result was found for the T4N0-N2 subsets in the sixth American Joint Committee on Cancer staging system. However, the staging system we propose shows more consistent hazards within the same stage group and better survival discrimination among T categories, N categories, and overall stages. CONCLUSION: Using the 6th American Joint Committee on Cancer staging system produces an acceptable distribution of patient numbers and segregation of survival curves among the different stage groups. The prognostic accuracy of the staging system could be improved by recategorizing the T, N, and group stage criteria.