Staging primary head and neck cancers with (18)F-FDG PET/CT: is intravenous contrast administration really necessary?

Eur J Nucl Med Mol Imaging. 2009 Sep;36(9):1417-24. doi: 10.1007/s00259-009-1127-8. Epub 2009 Apr 8.


Purpose: The aim of our study was to prospectively evaluate whether intravenous contrast media in integrated positron emission tomography and computed tomography (PET/CT) with (18)F-fluorodeoxyglucose ((18)F-FDG) significantly contributes to evaluation of primary head and neck cancers compared with unenhanced PET/CT, regional contrast-enhanced CT of head and neck (neck CE-CT) and regional magnetic resonance imaging of head and neck (neck MRI).

Methods: Subjects were 42 consecutive patients (35 men, 7 women; age range: 36-91 years) with biopsy-proven primary head and neck cancers. Lesion detection of primary and nodal sites and TNM classification were assessed on a per-patient basis. McNemar test and kappa statistics were employed for statistical analyses.

Results: Forty patients (95%) were successfully followed up: 24 patients had nodal disease and 3 had distant metastasis. Contrast-enhanced and unenhanced PET/CT detected 98 and 95% of the primary tumours, respectively, and both detected 92% of patients with nodal disease, which revealed no statistically significant difference. Accuracy for T status was 75 and 73%, respectively, which proved significantly more accurate than neck CE-CT, which had an accuracy of 53% (p = 0.0133 and 0.0233, respectively). Neck MRI correctly classified the T status in 58% of patients; however, no statistically significant difference was found between PET/CT and neck MRI. Contrast-enhanced PET/CT, unenhanced PET/CT, neck CT and neck MRI correctly staged the N status in 90, 90, 79 and 90% of patients, respectively, with no statistically significant difference. Overall TNM classification was correctly classified in 68 and 65% of patients, respectively. Weighted kappa values between enhanced and unenhanced PET/CT for primary tumour detection, nodal detection, T status and N status were 0.655, 1.000, 0.935 and 1.000, respectively.

Conclusion: We found almost perfect correlation between enhanced and unenhanced PET/CT for lesion detection and initial staging of primary head and neck cancers. Routine contrast administration for PET/CT imaging may not be justified.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Contrast Media* / administration & dosage
  • Female
  • Fluorodeoxyglucose F18*
  • Head and Neck Neoplasms / diagnostic imaging*
  • Head and Neck Neoplasms / pathology
  • Humans
  • Injections, Intravenous
  • Male
  • Middle Aged
  • Neoplasm Metastasis
  • Positron-Emission Tomography
  • Prospective Studies
  • Radiopharmaceuticals*
  • Tomography, X-Ray Computed


  • Contrast Media
  • Radiopharmaceuticals
  • Fluorodeoxyglucose F18