Background: The aim of this study was to assess the accuracy of a novel imaging modality, three-dimensional (3D) metabolic and radiologic gathered evaluation (MeRGE), for localizing parathyroid adenomas (PAs).
Methods: Consecutive patients presenting with primary hyperparathyroidism who underwent both thin-slice cervical computed tomography (CT) and (99m)Tc-sestamibi (MIBI) scanning were included. 3D-CT reconstruction was obtained using VR-RENDER, which was used to perform 3D virtual neck exploration (3D-VNE). The MIBI scan was then fused with the 3D reconstruction to obtain 3D-MeRGE. Sensitivity, specificity, and accuracy were assessed. Parathyroid gland volume and preoperative parathormone (PTH) levels were analyzed as predictive factors of correct localization (i.e., correct quadrant).
Results: A total of 108 cervical quadrants (27 patients) were analyzed. Sensitivities were 79.31, 75.86, 65.51, and 58.61 % with 3D-MeRGE, 3D-VNE, MIBI, and CT, respectively. Specificity was highest with CT (94.93 %) followed by 3D-VNE (92.4 %). MIBI and 3D-MeRGE had the same specificity (88.6 %). 3D-MeRGE and 3D-VNE achieved higher accuracy than MIBI or CT alone. Mean PTH values were significantly higher in patients with lesions that were correctly identified (true positive, TP) than in those whose lesions were missed (false negative, FN) with 3D-VNE (219.60 ± 212.77 vs. 98.75 ± 12.76 pg/ml; p = 0.01) and 3D-MeRGE (217.69 ± 213.76 vs. 09.75 ± 20.48 pg/ml; p = 0.02). The mean parathyroid gland volume difference between TP and FN was statistically significant with all modalities except CT.
Conclusions: 3D-MeRGE and 3D-VNE showed high accuracy for localization of PAs. 3D-MeRGE performed better than MIBI or CT alone for detecting small adenomas and those with a low PTH level.