Objectives: This study was performed to assess the role of retrospective PET-MRI fusion with Ga-68-DOTA(0)-Phe(1)-Tyr(3)-octreotide (Ga-68-DOTATOC) PET and Gd-EOB-DTPA MRI in the detection of hepatic metastases from neuroendocrine tumours (NET).
Methods: Twenty-two consecutive patients with suspected liver metastases from histopathologically proven NET were examined with Gd-EOB-DTPA MRI and multiphase contrast-enhanced Ga-68-DOTATOC PET/CT. PET and MRI images were retrospectively fused using commercially available software. Two physicians experienced in nuclear medicine and radiology analysed the images to assess diagnostic confidence and characterise liver lesions.
Results: A total of 181 lesions were detected. PET-MRI showed a sensitivity of 91.2% (significantly superior to PET/CT; P < 0.05) and a specificity of 95.6% (significantly superior to MRI; P < 0.05). PET/CT had a sensitivity of 73.5% and a specificity of 88.2%. MRI had a sensitivity of 87.6% and a specificity of 86.8%. The area under the curve was 0.98 for PET-MRI, 0.96 for MRI, and 0.89 for PET/CT (P < 0.05).
Conclusions: Retrospectively fused PET-MRI was superior to multiphase contrast-enhanced Ga-68-DOTATOC PET/CT and Gd-EOB-DTPA MRI in the detection of NET liver metastases. It was more sensitive than PET/CT and more specific than MRI. Fused PET-MRI therefore seems well suited for surgical and interventional treatment planning of NET liver metastases.
Key points: • Ga-68-DOTATOC PET-Gd-EOB-DTPA MRI fusion can improve imaging of liver metastases of neuroendocrine tumours. • This technique appears more sensitive than PET/CT for staging NET hepatic metastases. • Ga-68-DOTATOC PET-Gd-EOB-DTPA MRI fusion is more specific than MRI alone.