PET/MR for Molecular Morphometric Characterization of Lymphnode Metastases in Bladder Cancer

Objectives Diffusion-Weighted Whole Body Imaging with Background Signal Suppression (DWIBS) is an advanced application of DWI to image a large volume to characterize elevated micro-cellularity in metastatic lymph nodes. The feasibility of DWIBS in assessing cancer involvement of pelvic lymph nodes with high cellularity has been previously shown, however it has been found to be lacking in specificity due to false positive rates related to inflammatory changes and other sources of increased cellularity in lymphnodes. Therefore we aimed to assess the combination of morphologic imaging with the functional characteristics and metabolic information provided by high resolution digital 18F-FDG PET/CT for staging and detection of lymphnode metastases in bladder cancer.

Methods Prior to cystectomy, 10 patients had 18F-FDG whole body PET performed on a pre-commercial release digital PET/CT, Vereos (dPET) acquired using an initial dynamic acquisition for 30 min covering the pelvis followed by two whole body sweeps with 90 seconds per bed position performed at 50 and 75 minutes post-injection. The target injection dose was 13 mCi FDG. High resolution reconstruction of all PET data was performed, using a 2 mm voxel length volume with point spread function correction and Gaussian filter enabled. Patient positioning was carefully monitored in order to facilitate the co-registration with the DWIBS MRI data sets. MRI was performed on an Ingenia CX 3T using 32-channel phased-array coils. The DWIBS sequence was performed in addition to T2W-MRI, DWI, and APT-MRI sequences prior to the administration of a Gadolinium-based contrast agent for DCE-MRI and post-contrast morphologic imaging. PET and MR images were manually co-registered and evaluated for detection and characterization of lymphnodes suspicious for metastatic disease.

Results The matched assessment of visual lymph nodes after co-registration of the DWIBS MRI with the FDG-PET/CT was readily completed for all patient data sets. The metabolic characteristics of each lymph node was assessed by SUVmax and metabolic volume determination and correlated to the visual impression of the lymph node based on the DWIBS MRI reads. We found that reader training for appropriate evaluation of the DWIBS images using cases obtained from the volunteer subject / patient population improved the interpretability of the co-registered data, reducing inter- and intra-reader variability. Initial results point to the ability to better delineate lymphnodes with increased DWIBS signal based on correlation to established PET/CT interpretation approaches.

Conclusions DWIBS MRI enables morphologic imaging of lymphnodes with signal intensity related to tissue cellularity and as such provides high resolution anatomic detail and sensitivity in detecting metastatic lymphnodes. The limited specificity of this MR imaging approach can potentially be compensated for by pairing it with high definition digital PET. We found that a combined imaging approach is readily feasible and of excellent clinical utility. Appropriate reader training and validated quantitative analysis may allow this multi-modality technique to provide a breakthrough for non-invasive N-staging in bladder cancer.