Whole-body MRI is more sensitive but less specific than FDG-PET/CT for cancer detection, according to researchers from China and Europe. Findings suggest a complementary rather than exclusive role in oncologic imaging for both modalities and validate recent studies suggesting close follow-up since either test can miss metastases.
The clinical literature points at whole-body MRI as the most accurate test for the detection of bone metastases compared with whole-body FDG-PET/CT in patients with different types of cancer. New data suggest both tests may actually offer valuable diagnostic and staging information, although neither one is infallible, said principal investigator Dr. Till Heusner, a radiologist and nuclear medicine physician at the University Hospital Essen in Germany.
Heusner and colleagues enrolled 55 patients with malignant melanoma and 54 with non-small cell lung cancer who underwent whole-body imaging with MRI and FDG-PET/CT for detection of bone metastases. Patients also underwent biopsy and slightly more than a year of clinical follow-up, which served as the standard of reference for the study.
The investigators found whole-body MRI more sensitive for metastatic detection. They deemed both tests similarly accurate, however, considering MRI's higher sensitivity also meant an increased rate of false- positive findings compared with PET/CT. They also found evidence of metastatic lesions missed by both modalities at initial diagnosis that were confirmed a year later.
Eleven patients with either melanoma or NSCLC had bone metastases. Whole-body PET/CT and MRI correctly detected metastases in five and seven patients, respectively. MRI was false positive in six patients, while PET/CT was false positive in one. Sensitivities, specificities, positive and negative predictive values, and accuracies for detection of bone metastases were 45%, 99%, 83%, 94%, and 94%, respectively, with PET/CT and 64%, 94%, 54%, 96%, and 91% with MRI.
The differences between the two imaging modalities were not statistically significant (p>0.05).
Patient interaction could in some cases be worth 1000 whole-body images, according to Heusner. Typical false-positive cases with all modalities involved seemingly bony lesions in patients who had recently sustained traumatic injuries.
