Three hours delayed 18F-FDG PET in evaluating solitary pulmonary nodules

  • Dr Wen-Sheng Huang, Department of Nuclear Medicine, Tri-service General Hospital, Taipei, Taiwan, Taiwan
  • Dr Chi-Yung Chang, Department of Nuclear Medicine, Tri-service General Hospital, Taipei, Taiwan, Taiwan
  • Dr Wen-Shan Liu, Department of Radiation Oncology, Chung Shan Medical University Hospital, Taichung, Taiwan, Taiwan
  • Dr Rong-Tong Liu, Department of Surgery, Chung Shan Medical University Hospital, Taichung, Taiwan, Taiwan
  • Dr Jung-Kung Lee, Department of Nuclear Medicine, Chung Shan Medical University Hospital, Taichung, Taiwan, Taiwan
  • Dr Ching-yee Wong, Department of Nuclear Medicine, Oakland University William Beaumont Hopsitals, Royal Oak Michigan, United States

    • Objective: Solitary pulmonary nodules (SPN) sometimes cause diagnostic dilemma in FDG PET. Different metabolic behavior of FDG at different time point set between benign and malignant tissues enables us to discriminate each other. We observed the triple-time point SUV in separating malignant from benign SPN.

    Methods: 41 patients with SPN seen on CT were enrolled and underwent FDG-PET in 2 wks using a dedicated PET system. Final diagnoses were made based on histology (n=38) or at least 24 m of CT follow-up (n=3). Three sets of FDG PET were acquired at 1 h (scan 1), 2 h (scan 2), and 3 h (scan 3) after 370 MBq FDG injection and were reconstructed using OSEM. The maximal standardized uptake values (SUVmax) at 3 time points, i.e. SUV1h, SUV2h, and SUV3h of each SPN were measured. Receiver-operating- characteristic (ROC) curves of the 3 parameters were evaluated by comparing the areas under the curves (AUC) using paired t-tests. Sensitivity, specificity, and positive and negative predictive values of these analyses were determined at the optimal cutoff points in reference to the ROC curves.

    Results: The averaged SUVmax in malignancy (mean±SD) was 4.60±3.25 on scan 1, 5.28±3.50 on scan 2, and 5.97±3.66 on scan 3. However, none of the ROC curves derived from the 3 parameters revealed significantly greater AUC than the other (SUV1h, SUV2h, and SUV3h =0.85, 0.89, and 0.92, respectively).

    Conclusion: The SUV3h might not achieve significantly better discrimination power on SPN for differentiating malignant from benign lesions than that of SUV2h.