[Radiologic criteria of MRI in diagnosing retropharyngeal lymph nodal metastasis of nasopharyngeal carcinoma]

Zhonghua Yi Xue Za Zhi. 2006 May 16;86(18):1265-8.
[Article in Chinese]

Abstract

Objective: To explore the reasonable radiologic criteria of retropharyngeal lymph nodal (RLN) metastasis of nasopharyngeal carcinoma (NPC).

Methods: 275 patients with newly diagnosed NPC were enrolled. All patients underwent Magnetic resonance imaging (MRI) before treatment to measure the sizes of maximal and minimal axial diameters of each node. Determination of malignancy was based on the results of MRI follow-up. Receiver operating characteristic (ROC) curve was used to determine the optimal axial diameter and cutoff size criterion of RLN metastasis.

Results: Among the 275 patients, a total number of 468 RLNs were assessed. 240 positive and 228 negative nodes were later confirmed by the MRI follow-up. The mean sizes of minimal and maximal axial diameters for positive nodes were significantly higher than those for negative nodes (9.9 mm +/- 4.8 mm vs 3.3 mm +/- 1.2 mm, 12.9 mm +/- 5.8 mm vs 5.7 mm +/- 2.0 mm, all P < 0.01). The reasonable criterion for diagnosing RLN metastasis was the minimal axial diameter of > or = 6 mm. The sensitivity, specificity and accuracy of this criterion were 84.1%, 94.3% and 89.1%, respectively. There were 3 nodes whose minimal axial diameter was < 6 mm in 50 RLNs with central necrosis. Central necrosis combined with the minimal axial diameter criteria slightly increased the sensitivity at a stable specificity.

Conclusion: The reasonable radiologic criteria for assessing RLN metastasis of NPC are: (1) node with a minimal axial diameter > or = 6 mm; and (2) any node with central necrosis.

Publication types

  • English Abstract

MeSH terms

  • Adult
  • Aged
  • Female
  • Follow-Up Studies
  • Humans
  • Lymph Nodes / pathology
  • Lymphatic Metastasis / diagnosis*
  • Magnetic Resonance Imaging / methods*
  • Male
  • Middle Aged
  • Nasopharyngeal Neoplasms / pathology*
  • Reproducibility of Results
  • Sensitivity and Specificity