The study included 50 consecutive patients who underwent routine FDG-PET/CT for potentially treatable lung cancer following staging CT. The subjects also received PET/MRI scans the same day.
Two experienced readers were blinded to the other modality's results and interpreted PET/MR images independently. In addition, two different readers with no knowledge of the PET/MRI findings reviewed PET/CT images separately for tumor-node-metastasis (TNM) staging.
Receiver operator characteristics (ROC) curve analysis showed that PET/MRI had sensitivity of 97% and specificity of 92% for determining resectability.
Interobserver agreement for PET/MRI was between "substantial" and "perfect" for TNM staging by the two readers. Intermodality agreement between PET/CT and PET/MRI was between "substantial" and "almost perfect" for TNM staging.
"In clinical staging, it is not uncommon for MRI to characterize uncertain findings seen at PET/CT or enhanced CT," commented lead author Dr. Francesco Fraioli, from the department of nuclear medicine and radiology, in an email to AuntMinnie.com. "The interval time between two separate scans might potentially result in delays in patient evaluation and staging by the local multidisciplinary team. The one-stop-shop of PET/MRI can fill this gap, providing at the same time multiparametric metabolic, functional, and morphological information combining radioactive tracers with different MR sequences."
University College London Hospitals installed the first PET/MRI scanner in the U.K. almost three years ago. "The machine has been generally well-accepted by our clinicians for staging, providing that the patient already has had a diagnostic CT," Fraioli added.
He and his colleagues have since looked the role of a "standard" PET/MRI protocol, performed similarly to PET/CT to assess tumor resectability.
"We plan to progress our study by investigating the optimal diagnostic protocol for lung cancer staging, and particularly the development of new, specific dedicated lung MR protocols," he said.
