Abstract
2998
Introduction: To establish the diagnostic criteria for differentiating lymph node metastasis (LNM) in patients with papillary thyroid carcinoma (PTC) based on 18F-fluorodeoxy-D-glucose positron emission tomography-magnetic resonance imaging (18F-FDG PET/MRI), and to investigate the clinical efficacy for determining the surgical scope of lymph node (LN) dissection using the established criteria.
Methods: The imaging and clinical data of 12 patients with PTC from May to November in 2021 were analyzed retrospectively. These patients were examined by PET/MRI and ultrasound (US) within 2 weeks to determine the N stage of the disease. The size of all visible LNs in both sides of the neck was measured on MRI axial fat-suppressed T2-weighted images and the maximum standardized uptake value (SUVmax) in hybrid PET images. Based on postoperative histopathology, the optimal quantitative diagnostic criteria of MRI and PET/MRI for differentiating LNM were established by logistic regression analysis and receiver operating characteristic (ROC) curve analysis. For determining the surgical scope, the neck LNs were divided into central, left lateral and right lateral regions, according to surgical guidelines. The regional LNM diagnosis of US was made by experienced physicians based on clinical experience and negotiation. The sensitivity, specificity and accuracy of the established MRI and PET/MRI quantitative criteria were compared with those of US diagnosis.
Results: 12 patients with PTC were enrolled, with 19 LN regions being dissected by surgery. Postoperative pathology showed 15 malignant regions and 4 benign regions, including 58 malignant LNs and 113 benign LNs. Univariate and multivariate logistic regression analysis showed that PET SUVmax and short diameter on MRI (SD-MRI) were independent predictors of malignancy in patients with PTC. According to ROC curve analysis, the cutoff value of SD-MRI was 5.7 mm (area under curve, AUC=0.806; Youden Index, YI=0.460). For the LNs ≥5.7 mm, the cutoff value of SUVmax was 1.6. For the LNs <5.7 mm, the cutoff value of SUVmax was 1.8. By hybriding the above “Dual threshold” quantitative diagnostic criteria (SD-MRI ≥5.7 mm & SUVmax>1.6 or SD-MRI <5.7 mm & SUVmax>1.8), the AUC and YI of hybrid PET/MRI could be improved to be 0.901 and 0.803 (Figure A). Based on the regional level analysis, the sensitivity/ specificity/ accuracy of LNM diagnosis made by US as well as quantitative MRI and PET/MRI “Dual threshold” criteria were 73.3%/ 100%/ 78.9%, 80.0%/ 25.0%/ 68.4% and 86.6%/ 75.0%/ 84.2%, respectively (Figure B). Compared with US, PET/MRI “Dual threshold” criteria can avoid the omission of 10.5% (2/19) LNM region, but increase the excessive dissection of 5.3% (1/19) benign LN region.
Conclusions: By integrating PET modality, PET/MRI "Dual threshold" demonstrated better accuracy than size-based MRI quantitative criteria for differentiating LNM in patients with PTC. For determining the LNM dissection scope, PET/MRI "Dual threshold" had reduced omission rate and better accuracy compared with US. This result indicates the clinical application potential and rationality of hybrid PET/MRI devices for PTC.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.