RSNA 2003 Scientific Papers > Is PET/CT More Accurate Than PET or CT Alone? A Prospective ...
  Scientific Papers
  SESSION: Neuroradiology/Head and Neck (Head and Neck Cancer: Staging and Recurrence)

Is PET/CT More Accurate Than PET or CT Alone? A Prospective Study

  DATE: Wednesday, December 03 2003
  START TIME: 11:30 AM
  END TIME: 11:37 AM
  LOCATION: Room N228
  CODE: K14-984

Barton Branstetter MD
Pittsburgh , PA
Todd Blodgett MD
Lee Zimmer MD
Subha Raman MD
Carl Snyderman MD
Carolyn Meltzer MD

Head and neck neoplasms, CT
Positron emission tomography (PET)

Purpose: Combined PET/CT imaging is available at many institutions, but there have been no definitive studies demonstrating that the combined modality is more accurate than PET or CT alone in the evaluation of head and neck tumors. We designed a prospective study to determine whether patient outcome would better correlate with combined PET/CT than with PET or CT alone.

Methods and Materials: Consecutive patients (n=65) with known or suspected malignancies of the head and neck were evaluated on a commercial PET/CT scanner. A neuroradiologist with head and neck experience prospectively reviewed the CT images without the PET data. A nuclear medicine physician with PET experience prospectively reviewed the PET images without the CT data. Lesions were categorized on a 5-point scale: benign, probably benign, equivocal, probably malignant, or malignant. The two reviewers then evaluated the fused PET/CT images and rendered a consensus interpretation. The patients were followed for an average of 9 months (range: 6 to 12 months), and the presence or absence of tumor was determined with biopsy, radiologic follow-up, or clinical assessment. This follow-up was considered the gold standard for receiver-operator characteristic (ROC) curves analyzing individual lesions (per lesion basis) and overall tumor presence (per patient basis). Statistical significance was determined with an area test applied to the correlated, ordinal category data.

Results: Follow-up was available for 64 patients; the remaining patient was excluded from the analysis. 125 lesions were identified in 57 / 64 patients. 20 lesions were evaluated by biopsy, 15 lesions were evaluated with radiologic follow-up, and 90 lesions were assessed clinically. For analysis of individual lesions, ROC curves demonstrated that CT was statistically inferior to PET, and PET was statistically inferior to PET/CT. For overall assessment of tumor presence, CT was statistically inferior to PET, and there was a trend suggesting that PET was inferior to PET/CT.

Conclusion: This prospective trial demonstrates improved accuracy of PET/CT over PET or CT alone in the assessment of head and neck malignancies.




Questions about this event email: