Abstract
OBJECTIVE: To systematically evaluate the diagnostic efficacy of Al¹⁸F-NOTA-FAPI-04 PET/CT in breast cancer and explore the correlation between tracer uptake parameters and pathological features of breast cancer based on single-centre, retrospective data. METHODS: This single-center retrospective study enrolled 58 female patients with suspected primary breast cancer who underwent Al¹⁸F-NOTA-FAPI-04 PET/CT before surgery or core needle biopsy (February 2023-November 2025). Lesion maximum/mean standardized uptake values (SUVmax/SUVmean) were measured via region-of-interest analysis. Semi-quantitative positivity thresholds (SUVmax > 2.5 or tumor-to-background ratio ≥ 1.5) were used for lesion classification. Diagnostic efficacy (sensitivity, specificity, accuracy) was calculated with pathological results as the gold standard, and 95% confidence intervals (CIs) were determined via the Clopper-Pearson exact method. Differences in uptake parameters across pathological types, molecular subtypes, and immunohistochemical (ER/PR/HER2/Ki-67) statuses were analyzed using Mann-Whitney U or Kruskal-Wallis H tests. RESULTS: Of 58 patients, 49 had breast cancer and 9 had benign lesions. Al¹⁸F-NOTA-FAPI-04 PET/CT achieved a sensitivity of 95.9% (95% CI: 86.3%-99.5%), specificity of 88.9% (95% CI: 51.9%-99.7%), and accuracy of 94.8% (95% CI: 85.8%-98.9%). All 53 primary lesions were detected (detection rate: 100%), and the detection rate for 87 lymph node metastases was 96.5%. SUVmax/SUVmean of malignant lesions (median [IQR]: 13.20 [9.55-17.85]/8.15 [5.68-10.92]) were significantly higher than those of benign lesions (2.13 [1.56-2.89]/1.35 [0.98-1.86], both P < 0.001). No statistically significant differences in uptake parameters were observed across pathological types, molecular subtypes, or ER/PR/HER2/Ki-67 statuses (all P > 0.05), though numerical trends existed; this may be attributed to Type II error due to small subgroup sizes and insufficient statistical power (29.7%-62.3% for detecting medium-to-large effect sizes). No correlation was found between lesion size and uptake parameters (r = 0.186 for SUVmax, r = 0.165 for SUVmean, both P > 0.05). CONCLUSION: Al¹⁸F-NOTA-FAPI-04 PET/CT exhibits high sensitivity and accuracy for breast cancer diagnosis and lymph node metastasis detection, with uptake independent of blood glucose levels. Its core clinical roles include complementary initial diagnosis in high-risk subgroups (dense breasts, diabetes), primary nodal staging, and problem-solving in equivocal cases. However, the lack of significant correlation between uptake parameters and pathological features should be interpreted cautiously due to potential Type II error. These characteristics make it a versatile imaging tool for breast cancer diagnosis and staging in regional tertiary care settings, with further validation needed in larger, balanced cohorts.