Introduction: Mantle cell lymphoma (MCL) is a rare subtype of non-Hodgkin lymphoma and requires both bone marrow biopsy and fluorine-18 fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) to correctly stage the disease at diagnosis. However, accurate identification of bone marrow involvement by lymphoma on FDG PET/CT scans has not been previously demonstrated. We hypothesized that a voxel-based analysis of the iliac bones on the FDG PET/CT scan might provide insight into bone marrow involvement.
Patients and methods: A retrospective cohort study of patients with a diagnosis of MCL who had undergone both single iliac bone marrow biopsy and FDG-PET/CT scan from 1 study site were included in the development phase (n = 10). An additional 12 patients from a second institution were included in the validation phase. Using a semiautomated workflow, a voxel-based data set of FDG uptake within the bilateral iliac bones was captured for each patient. In the development phase, empirical receiver operating characteristic curves for each data set were fit. We then identified the standardized uptake value (SUV) threshold cutpoints at which the sensitivity and specificity were optimized to 100%. In the validation phase, we evaluated the performance of these candidate SUV threshold cutpoints in 15 additional patients from a second institution.
Results: We found that 1 cutpoint, > 38% of voxels with activity < 0.95, outperformed all the other candidate cutpoints, correctly classifying all patients except for 1 (overall sensitivity, 100%; specificity, 87.5%).
Conclusion: The ability to correctly identify bone marrow involvement using FDG PET/CT-based voxel analysis provides promise as a novel noninvasive method of accurate staging.
Keywords: Non-Hodgkin lymphoma; PET/CT; Standardized uptake value; Threshold; Voxel analysis.
Copyright © 2018 Elsevier Inc. All rights reserved.