Evaluating diagnostic accuracy and determining optimal diagnostic thresholds of different approaches to [68Ga]-DOTATATE PET/MRI analysis in patients with meningioma

Abstract

Multiple approaches with [68Ga]-DOTATATE, a somatostatin analog PET radiotracer, have demonstrated clinical utility in evaluation of meningioma but have not been compared directly. Our purpose was to compare diagnostic performance of different approaches to quantitative brain [68Ga]-DOTATATE PET/MRI analysis in patients with suspected meningioma recurrence and to establish the optimal diagnostic threshold for each method. Patients with suspected meningioma were imaged prospectively with [68Ga]-DOTATATE brain PET/MRI. Lesions were classified as meningiomas and post-treatment change (PTC), using follow-up pathology and MRI as reference standard. Lesions were reclassified using the following methods: absolute maximum SUV threshold (SUV), SUV ratio (SUVR) to superior sagittal sinus (SSS) (SUVRsss), SUVR to the pituitary gland (SUVRpit), and SUVR to the normal brain parenchyma (SUVRnorm). Diagnostic performance of the four methods was compared using contingency tables and McNemar's test. Previously published pre-determined thresholds were assessed where applicable. The optimal thresholds for each method were identified using Youden's J statistics. 166 meningiomas and 41 PTC lesions were identified across 62 patients. SUV, SUVRsss, SUVRpit, and SUVRnorm of meningioma were significantly higher than those of PTC (P < 0.0001). The optimal thresholds for SUV, SUVRsss, SUVRpit, and SUVRnorm were 4.7, 3.2, 0.3, and 62.6, respectively. At the optimal thresholds, SUV had the highest specificity (97.6%) and SUVRsss had the highest sensitivity (86.1%). An ROC analysis of SUV, SUVRsss, SUVRpit, and SUVRnorm revealed AUC of 0.932, 0.910, 0.915, and 0.800, respectively (P < 0.0001). Developing a diagnostic threshold is key to wider clinical translation of [68Ga]-DOTATATE PET/MRI in meningioma evaluation. We found that the SUVRsss method may have the most robust combination of sensitivity and specificity in the diagnosis of meningioma in the post-treatment setting, with the optimal threshold of 3.2. Future studies validating our findings in different patient populations are needed to continue optimizing the diagnostic performance of [68Ga]-DOTATATE PET/MRI in meningioma patients.Trial registration: ClinicalTrials.gov Identifier: NCT04081701. Registered 9 September 2019. https://ichgcp.net/clinical-trials-registry/NCT04081701 .

Conflict of interest statement

The authors declare no competing interests.

© 2022. The Author(s).

Figures

Figure 1

Flow chart of the included subjects and lesions as confirmed by each diagnostic criterion.

Figure 2

Axial images of [68Ga]-DOTATATE PET AC (A), 3D T1 post-gadolinium MR (B), fused PET/MR T1 (C–E) windowed SUV 0–15. This representative patient from the cohort is a 70-year-old female with a history of WHO grade II meningioma s/p resection and proton therapy 6 years prior to imaging and an additional proton therapy a year prior. The fused PET/MR images demonstrate a lesion suspicious for a meningioma in the left anterior temporal pole with SUV of 13.6 (C, red arrow). The more posteriorly located enhancing lesion (C, white arrow) demonstrates SUV of 4.5 and was suspicious for post-treatment change, given that the SUV of the superior sagittal sinus was 2.3. The subsequent resection and biopsy of the two lesions a year later confirmed the suspected diagnosis of recurrent meningioma and radiation necrosis, respectively. The superior sagittal sinus (D, arrow) demonstrates SUV of 2.3 and the pituitary gland (E) demonstrates SUV of 12.4.

Figure 3

Meningioma (M) versus Post-treatment change (PTC) SUV, SUVRsss, SUVRpit, and SUVRnorm with mean and standard deviations. Mean SUV, SUVRsss, SUVRpit, and SUVRnorm of meningioma was 15.8, 11.5, 0.92, and 324.7 respectively. **** indicates statistical significance with p 

Figure 4

Meningioma SUV, SUVRsss, SUVRpit, and SUVRnorm stratified by WHO Grades with mean and standard deviations. Kruskal–Wallis test showed no statistically significant relationship between WHO Grade and the four SUV metrics.

Figure 5

ROC Curves of the four methods: SUV, SUVRsss, SUVRpit, and SUVRnorm. ROC analysis revealed area under the curve of 0.932, 0.910, 0.915, and 0.800 for SUV, SUVRsss, SUVRpit, and SUVRnorm, respectively (all P