The present study was designed to assess the additional value of 2-deoxy-2[18F]fluoro-D-glucose ([18F]FDG) positron emission tomography/computed tomography (PET/CT) to magnetic resonance imaging (MRI) in the treatment response assessment of multiple myeloma (MM). We performed a meta-analysis of all available studies to compare the detectability of treatment response of [18F]FDG PET/CT and MRI in treated MM. We defined detecting a good therapeutic effect as positive, and residual disease as negative. We determined the sensitivities and specificities across studies, calculated the positive and negative likelihood ratios (LR), and made summary receiver operating characteristic curves (SROC) using hierarchical regression models. The pooled analysis included six studies that comprised 278 patients. The respective performance characteristics (95% confidence interval (CI)) of [18F]FDG PET/CT and MRI were as follows: sensitivity of 80% (56% to 94%) and 25% (19% to 31%); specificity of 58% (44% to 71%) and 83% (71% to 91%); diagnostic odds ratio (DOR) of 6.0 (3.0–12.0) and 1.7 (0.7–2.7); positive LR of 1.8 (1.3–2.4) and 1.4 (0.7–2.7); and negative LR of 0.33 (0.21–0.53) and 0.81 (0.62–1.1). In the respective SROC curves, the area under the curve was 0.77 (SE, 0.038) and 0.59 (SE, 0.079) and the Q* index was 0.71 and 0.57. Compared with MRI, [18F]FDG PET/CT had higher sensitivity and better DOR and SROC curves. Compared with MRI, [18F]FDG PET/CT had greater ability to detect the treatment assessment of MM.
Challenges and caveats of a multi-center retrospective radiomics study: an example of early treatment response assessment for NSCLC patients using FDG-PET/CT radiomics