Staging of nasopharyngeal carcinoma investigated by magnetic resonance imaging

Radiother Oncol. 2006 Apr;79(1):21-6. doi: 10.1016/j.radonc.2006.03.015. Epub 2006 Apr 19.

Abstract

Background and purpose: To investigate the American Joint Commission on Cancer (AJCC) sixth edition staging system of nasopharyngeal carcinoma (NPC) by Magnetic Resonance Imaging (MRI).

Patients and methods: One hundred and fifty-nine non-disseminated biopsy-proven NPC patients were studied with MRI before treatment. Retrieval of MRI information enabled us to restage all patients accurately according to the sixth edition of the AJCC staging system. Splitting the respective T and N stages by the significant defining factors identified, the cancer death hazard ratios were modeled by the Cox model in SPSS 10.0 for windows (SPSS Inc, Chicago, IL).

Results: Single site of skull base abnormality (HR = 3.91, 95% CI: 0.74-20.56) has a superior result to others involved in T3 (HR = 5.83, 95% CI: 1.24-27.29). Involvement of either anterior or posterior cranial nerves solely (HR = 6.02, 95% CI: 1.55-35.60) was not found to be as a poor prognostic indicator as others involved in T4 (HR = 7.81, 95% CI: 1.81-33.63). Less than or equal to 3 cm of N1 (HR = 4.01, 95% CI: 0.48-33.83) and N2 (HR = 4.72, 95% CI: 0.62-35.78) have a better result than >3 cm of N1 (HR = 8.09, 95% CI: 0.95-68.97) and N2 (HR = 10.58, 95% CI: 1.32-84.62), respectively.

Conclusions: Perhaps, it is better to down-stage single site of skull base abnormality from T3 to T2, and involvement of either anterior or posterior cranial nerves solely from T4 to T3, meanwhile, < or =3 cm of N2 down-stage to N1, >3 cm of N1 up-stage to N2.

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Carcinoma / pathology*
  • Carcinoma / secondary
  • Female
  • Follow-Up Studies
  • Humans
  • Magnetic Resonance Imaging*
  • Male
  • Middle Aged
  • Nasopharyngeal Neoplasms / pathology*
  • Neoplasm Invasiveness
  • Neoplasm Staging / methods*
  • Neoplasm Staging / standards
  • Proportional Hazards Models
  • Reproducibility of Results
  • Retrospective Studies
  • Sensitivity and Specificity