Comparisons of detectability between whole-body MRI and I-123/131 mIBG images in patients with metastatic pheochromocytoma and paraganglioma

1111 Objectives: Whole-body scan and single-photon emission computed tomography/X-ray computed tomography (SPECT/CT) using meta-iodo-benzylguanidine (mIBG) have an important role in evaluation of pheochromocytoma and paraganglioma (PPGL), while the utility of whole-body magnetic resonance image (WB-MRI) for various tumors have been reported. Since a comparative study between WB-MRI and mIBG images in patients with metastatic PPGL has not been done, this study aimed to evaluate the detectability of lesions with WB-MRI compared with mIBG imaging. Methods: Thirteen patients with a resected primary lesion and histologically proven PPGL (age, 54.1 ± 15.1 years old; male/female = 8/5, pheochromocytoma/paraganglioma = 9/4) and those with suspected of having metastatic spread were included. All patients underwent both WB-MRI and I-123 mIBG images within 3 months as a pre-treatment evaluation of I-131 mIBG therapy. Two expert reviewers (9 and 11 years of experience) evaluated the presence of metastatic lesions in lungs, liver, bones, lymph nodes, and other locations (Table 1) on WB-MRI (WB-diffusion weighted image, apparent diffusion coefficient map, T1-weighted image, T2-weighted image, short inversion time inversion recovery weighted image), I-123 mIBG images (whole-body scan and SPECT/CT images 24 h after I-123 mIBG injection, 222 MBq), and post-therapy I-131 mIBG images (whole-body scan and SPECT/CT images 72-96 h after I-131 mIBG injection, 5550-7400 MBq/person). The gold standard was defined independently by another expert reviewer (15 years of experience) integrating all modalities (I-123 mIBG image, post-therapy I-131 mIBG image, WB-MRI, plain/contrast-enhanced CT images, F-18 FDG-PET/CT images). The diagnostic measures of sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV) and receiver operating characteristics (ROC) - area under the curve (AUC) for detecting metastatic lesions were calculated. The diagnostic tests were compared with ROC-AUC, and DeLong’s tests were performed for statistical analysis. P values < 0.05 were considered statistically significant. The consistency of the two reviewers was evaluated by Cohen’s kappa statistics. Results: The sensitivity, specificity, accuracy, PPV, NPV and AUC for detecting metastatic lesions were 83%, 98%, 91%, 98%, 86%, 0.91 in WB-MRI images, 84%, 94%, 89%, 93%, 86%, 0.89 in I-123 mIBG images, 88%, 92%, 90%, 92%, 88%, 0.90 in post-therapy I-131 mIBG images, respectively (Table 2,3). The AUC did not differ significantly among modalities. The kappa coefficients of reviewers of WB-MRI, I-123 mIBG, and post-therapy I-131 mIBG images were also comparable; 0.75, 0.75, and 0.78, respectively. The P values of the AUC with WB-MRI versus I-123/131 mIBG were 0.49 and 0.76, respectively (Table 3). WB-MRI images identified lesions with a detectability comparable to I-123 mIBG and post-therapy 131 mIBG images in patients with PPGL. Some lesions could only be detected by either I-123/131 mIBG or WB-MRI images (Figure 1,2). WB-MRI identified some additional lesions adjacent to organs with physiological high mIBG uptake on I-123/131 mIBG scintigraphy. Conclusions: Diagnostic accuracy of WB-MRI was comparable to mIBG images and might have advantages for detecting lesions around organs with physiological mIBG accumulation, although we need to pay attention to various findings other than metastasis. WB-MRI could play a complementary role in the diagnosis of metastatic PPGL.