This single-center post hoc retrospective analysis included 69 therapy-naïve women with suspected breast lesions (81 breast lesions, 797 regional lymph nodes, 94 distant metastases). The study compared the diagnostic accuracy of [18F]AlF-NOTA-FAPI-04 PET/CT with conventional imaging (ultrasonography, mammography, breast MRI, whole-body CT, and bone scan).
For primary breast lesions, conventional imaging showed a sensitivity of 93.75% and specificity of 100.00%, while visual PET/CT had a sensitivity of 98.44% and specificity of 70.59%. The AUC for conventional imaging was 0.970 versus 0.845 for visual PET/CT (P = 0.034), indicating significantly lower performance for PET/CT in this setting.
For regional lymph node metastases (LNM), visual PET/CT demonstrated significantly higher sensitivity (95.95% vs 73.99%) and specificity (97.60% vs 95.41%) compared to conventional imaging. Staging accuracy was also improved with PET/CT. Median follow-up was 4.5 months.
Safety and tolerability were not reported. Key limitations include the single-center design and retrospective analysis, which may introduce bias. The findings suggest that FAPI PET/CT may enhance detection of regional LNM, but prospective studies are needed before clinical adoption.
View Original Abstract ↓
Conventional imaging has limitations in differentiating benign from malignant breast lesions and in staging breast cancer (BC). This study compared the diagnostic and staging performance of [18F]AlF-NOTA-FAPI-04 PET/CT with conventional imaging in suspected breast lesions.
This single-center post hoc retrospective analysis included therapy-naïve participants with suspected breast lesions who underwent [18F]AlF-NOTA-FAPI-04 PET/CT between July 2023 and February 2025. All patients underwent ultrasonography, mammography, breast MRI, whole-body CT, and bone scan. Conventional imaging results were based on consensus visual interpretation integrating these modalities. PET/CT images were assessed visually and semi-quantitatively to assess the diagnostic accuracy. Lesions were confirmed by histopathology and follow-up (median, 4.5 months). Diagnostic performance for primary breast lesions, regional lymph nodes metastases (LNM), and distant metastases was evaluated on lesion-based, whereas visual scoring and staging (the AJCC eighth edition) comparisons were patient-based. Receiver operating characteristic/Delong and McNemar χ² analyses were performed using SPSS Statistics 29.0. A two-sided P
Sixty-nine women (median age, 50 years) with 81 breast lesions, 797 regional lymph nodes, and 94 distant metastases were included, and all 69 patients underwent ultrasonography, mammography, breast MRI, whole-body CT, and bone scan. For primary breast lesions, conventional imaging showed sensitivity and specificity of 93.75% and 100.00%, whereas visual PET/CT showed 98.44% and 70.59%. Although LBRbreast achieved the highest area under the curve (AUC), its AUC was not significantly different from that of conventional imaging (1.000 vs. 0.970, P = 0.078). In contrast, visual PET/CT showed a significantly lower AUC than conventional imaging (0.845 vs. 0.970, P = 0.034). For regional LNM detection, visual PET/CT achieved sensitivity and specificity of 95.95% and 97.60%, which were significantly higher than the corresponding values of conventional imaging (73.99% and 95.41%, all P
[18F]AlF-NOTA-FAPI-04 PET/CT outperforms conventional imaging in detecting regional LNM and improving pathological staging accuracy in BC.