Value of PET/MRI in CNS malignancies

Background: Clinical nuclear medicine services now include PET/MRI hybrid imaging: a technology that combines the delicate structural and functional characterization of tissue provided by MRI with the sensitivity of PET imaging of metabolism. However, there remain questions regarding which clinical indications are more accurately diagnosed by PET/MRI hybrid imaging versus PET plus MR performed separately. The major clinical question addressed in this study is whether there is a significant diagnostic benefit for hybrid PET/MRI imaging in patients with new or recurrent low or high-grade brain tumor and other brain malignancies.

Methods: From 2016-2017, 32 patients with CNS pathologies underwent clinical PET/MR scans at University of Pittsburgh Medical Center. Patient population included epilepsy (5), primary brain tumors (20), and metastatic brain or skull base tumor with questionable brain involvement (7). The 27 patients with primary brain tumors or metastatic brain or skull base tumors with questionable brain involvement were included in the study. All patients underwent PET/MR from a Siemens 3T mMR after injection of approximately 10 mCi F-18 FDG. Eight had fully diagnostic MR after injection with approximately 20 cc MultiHance IV. The remaining 19 had MR for attenuation correction plus T2_FLAIR, DWI, T1_MPRAGE and DWI_ADC sequences. Patients' tumor types were as follows: A-Meningioma (1), B-Maxillary sinus/nasopharyngeal (5), C-Low-grade glioma (4), D-High grade glioma (15), E- Metastatic tumor (2). An experienced nuclear medicine and radiologist physician (JMM) separately interpreted the MRI+ PET and PET/MRI images of all patients in sagittal, axial and coronal sections in a blinded manner. Patient diagnosis was first interpreted blindly from randomized PET and MR scans, separated from the original PET/MR sans. Diagnostic agreement between the separate interpretations of the PET and MRI scans versus the hybrid PET/MRI scans were scored. The added benefit from PET/MR compared to PET+MR was rated as High (benefit that changed or greatly increased the certainty of diagnosis), Moderate (benefit that increased the certainty of diagnosis), or Low (no change in certainty of diagnosis). High grade tumors were separated into positive (D+) or negative for recurrence (D-). Results: Table below shows added value of PET/MR compared with PET+MR. Figure illustrates images from a patient with suspected malignant degeneration of an oligodendroglioma (category D-) that was shown to not have high grade disease by PET/MR, but was considered as possible high grade degeneration by PET+MR.

Conclusions: In summary, diagnostic accuracy in patience with questionable recurrent viable high-grade glioma improved most by PET/MR, as there was greater certainty in not overcalling normal grey matter tissue close to the tumor bed as a region of tumor recurrence. Those that benefited the least were patients with clearly definable high-grade recurrence or clearly definable high grade metastatic brain tumor, since the presence of viable tumor was unequivocally identified equally on both PET+MR and PET/MR. Figure. 58-year-old female with left frontal oligodendroglioma diagnosed 10 years ago, underwent recent RT for suspected malignant degeneration. Recent MRI (not shown) showed multiple enhancing foci in the tumor periphery. PET imaging alone (top row) shows increased uptake in the lateral left frontal lobe. PET/MR (bottom row) shows the uptake corresponds to a normal appearing gyrus on MRI (middle row). Follow-up clinical assessment and MRI scans showed tumor stability.

• Download figure
• Open in new tab
• Download powerpoint