The ability of 18Ffluorodeoxyglucose positron emission tomography (18F-FDG PET) / magnetic resonance imaging (MRI) to have high diagnostic competence in the evaluation of diverse malignant tumors has been impressive. But this hybrid imaging technique cannot improve upon the gold standard of MRI to reveal brain metastases in newly diagnosed patients with non-small cell lung cancer (NSCLC), a multi-institutional German team of radiologists has found.
At initial diagnosis, a large percentage of patients with NSCLC also have distinct metastases. Because the presence of brain metastases impacts treatment strategy, cerebral staging is recommended for all stage III or IV patients. Radiologists from the University Hospital of Essen, the University of Duisburg-Essen, and the University Hospital of Dusseldorf conducted a study of 83 consecutive patients who had an initial cerebral staging with 18F-FDG PET/MRI that was part of a whole-body staging with initial 18F-FDG PET/computed tomography (CT) scan and subsequent whole-body 18FFDG PET/MRI scan.
Two neuroradiologists analyzed the images in consensus in random order, initially interpreting the MRI sequences followed four weeks later with the integrated 18FFDG PET/MRI datasets. They were asked to determine the absence or presence of brain metastases, the number of brain metastases, and the size of each.
MRI alone correctly identified brain metastases in 15 patients, with a total of 39 (100%) lesions. By comparison, brain metastases could only be identified in six of these patients with PET images and only 15 lesions. Thirteen of the 39 lesions were smaller than 5 mm; only one of these was identified on PET.
Lead author Cornelius Deuschl, MD, of the Institute of Diagnostic and Interventional Radiology and Neuroradiology at the University Hospital of Essen, and co-authors stated that in addition to 18F-FDG PET/MRI not improving diagnostic accuracy in the staging of these patients, the PET component of the integrated protocols could lead to uncertainty in diagnosis due to the potential for a high number of false-negative findings.
The sensitivity of PET correlated to the size of metastases. For metastases less than 5 mm in size, PET sensitivity was 8% compared to 27% for metastases between 5-10 mm, and 82% for metastases greater than 10 mm.
Diagnosing brain metastases: MRI still better than PET/MRI in NSCLC patients. Appl Radiol.