Background: To determine the utility of available radionuclide imaging modalities for preoperative parathyroid localization, we compared the accuracy of 4 types of technetium-99m ((99)Tc) sestamibi-based scans.
Methods: Over 5 years, 833 patients with sporadic primary hyperparathyroidism underwent either (99)Tc-sestamibi with planar views (Planar; n = 138); sestamibi single photon emission computed tomography (SPECT; n = 165); SPECT with thyroid (123)I-subtraction phase (SPECT/(123)I; n = 350); or SPECT combined with conventional CT (SPECT/CT; n = 180). The accuracy of each modality was determined on the basis of intraoperative parathyroid pathology, defined as single adenoma or multigland disease.
Results: Planar scans had significantly more false negatives (no tumor visualized) than SPECT-based scans (P < 0.01), but positive predictive values were similar. A false-negative scan was present in 38% of Planar scans, 27% SPECT, 4% SPECT/(123)I, and 17% SPECT/CT, with single adenoma found at operation in 77%, 64%, 53%, and 74%, respectively. When a scan had a single focus of uptake, SA was found at that location in 77%, 85%, 68%, and 87%, respectively. SPECT, SPECT/(123)I, and SPECT/CT did not significantly reduce the false-positive rate.
Conclusions: We recommend obtaining multiplanar SPECT-based imaging, which offers 3-dimensional localization and improved detection of parathyroid tumors when compared with Planar scans. A negative scan did not predict multigland disease.