Journal of nuclear medicine : official publication, Society of Nuclear Medicine最新文献

Currently, PET scanner validation phantoms, methods, and acceptance criteria for clinical trials are not standardized. This situation generates substantial inefficiencies with many scanners being tested multiple times for different trials. Herein we propose a standardized PET scanner validation paradigm for clinical trials. Methods: At present, active PET scanner validation programs administered by the European Association of Nuclear Medicine Research Ltd. (EARL), the Society of Nuclear Medicine and Molecular Imaging Clinical Trials Network (CTN), and Australia New Zealand Society of Nuclear Medicine Australasian Radiopharmaceutical Trials Network are reviewed in detail to identify similarities, differences, strengths, and weaknesses. PET criteria that help define the quantitative performance characteristics most critical for clinical trials are identified. Historical quantitative scanner performance capabilities are reviewed, including increasing availability of primary and secondary standard activity measurements for calibration purposes. Methodologies for these phantom-based measurements are reviewed, and standardized approaches are recommended. Results: Phantom requirements, acquisitions, reconstruction, analysis, and acceptance criteria have all been developed to be reasonably aligned with current standard scanner validation approaches, while at the same time recommending improvements and clarifications where programmatic differences were identified. A scanner validation program based on the measurement of radionuclide specific scanner calibration and harmonized recovery coefficient performance is proposed. Quarterly calibration verification of ¹⁸F and annual calibration of other radionuclides are recommended. Accuracy of ±5% for ¹⁸F calibration and ±10% for other radionuclides are proposed acceptance criteria. Annual verification of EARL 2-concordant recovery coefficient performance using a National Electrical Manufacturers Association NU2 image quality phantom or CTN5 phantom imaged at an 8:1 target-to-background contrast is recommended, although contrast recovery coefficients, rather than recovery coefficients, are advised. Conclusion: An internationally standardized PET scanner validation paradigm is proposed. International adoption of such a system combined with a data-sharing system would create a more efficient, robust, uniform, and trustworthy scanner validation environment for clinical trials while improving clinical trial qualification efficiency, decreasing costs and mitigating duplication of testing.

The aim of this retrospective study was to evaluate the correlation between findings from somatostatin receptor (SSTR) PET/CT and histopathology in patients with suspected intracranial meningiomas. Methods: We conducted a retrospective analysis of 8,077 SSTR imaging studies recorded in our institutional database between 2006 and 2021. In total, 223 SSTR PET/CT scans were performed for suspected meningioma, and 240 lesions were matched with histopathology results within 4 mo. Reports from SSTR PET/CT scans and histopathology were retrospectively reviewed to assess the presence of intracranial meningiomas. The positive and negative predictive values, sensitivity, specificity, and overall diagnostic accuracy of SSTR PET/CT were calculated. The SUV_max, SUV_mean, and SUV_peak were determined for each lesion. Results: In 222 (92.5%) of 240 lesions, meningioma was accurately identified by SSTR PET/CT and confirmed by histopathology. In 7 cases (2.9%), SSTR PET/CT suspected meningioma was not confirmed by histopathology (false-positive). Furthermore, in 11 cases (5%), meningioma was neither suspected by SSTR PET/CT nor confirmed by histopathology (true-negative result). There were no false-negative findings in our cohort. SSTR PET/CT demonstrated a sensitivity of 100% (95% CI, 98.4%-100%) and a specificity of 61.1% (95% CI, 35.8%-82.7%) in detecting meningiomas. Positive predictive value was 96.9% (95% CI, 93.8%-98.8%), and negative predictive value was 100% (95% CI, 71.5%-100%). The overall diagnostic accuracy was 97.1%. The receiver-operating-characteristic analysis for SUV_max in predicting histopathology results showed an area under the curve of 94%, indicating an excellent ability of SUV_max to distinguish between positive and negative histopathologic findings. Conclusion: SSTR PET/CT is a precise imaging modality for detecting intracranial meningiomas, as demonstrated by its high sensitivity. However, in 2.9% of cases, despite a positive PET/CT result, histopathology did not confirm the presence of a meningioma. Integration of MRI, histopathology, and SSTR PET/CT supports informed treatment decisions.