[18F]AlF-NOTA-FAPI-04 PET/CT Uptake of the Primary Tumor on PET/CT Imaging Might Distinguish Triple Negative Breast Cancer

Introduction: The objective of this study was to investigate the ability of [¹⁸F]AlF-NOTA-FAPI 04 PET/CT for molecular typing and pathological grading of primary breast cancer.

Methods: In this study, 33 newly diagnosed women with breast cancer who underwent [¹⁸F]AlF-NOTA-FAPI-04 examination were prospectively included, and a total of 35 breast cancer lesions were pathologically verified. Eight metrics extracted from PET images were studied: standardized uptake values (include SUVmax, SUVmean and SUVpeak), metabolic tumor volume (MTV), total leision FAPI (TLF), tumor background ratios (TBR_{blood pool}, TBR_muscle and TBR_bone). Two-sample t test and Mann-Whitney U test were used to determine PET parameters associated with pathological factors closely related to poor prognosis: hormone receptor-negative, HER-2 presence, Ki67≥30%, and triple-negative phenotype. Receiver operating characteristic (ROC) curve, AUC analysis and reclassification measures were used to evaluate the classification effect of different parameters on breast tumor features.

Results: [¹⁸F]AlF-NOTA-FAPI 04 demonstrated significant uptake in malignant breast cancer, as indicated by SUVmax (15.67±5.43), SUVmean (8.12±3.00), and SUVpeak (11.17±4.35). These parameters were all found to be positively correlated with the pathological grade of the primary lesion (P=0.0073, P=0.0051, and P=0.0044, respectively). Interestingly, SUVmax, SUVmean and SUVpeak were higher in estrogen receptor-negative tumors (P=0.0018, P=0.0024 and P=0.0025) and progesterone receptor-negative tumors (p = 0.0044, P=0.0089 and P= 0.0013). However, none of the PET indicators demonstrated significant discriminatory ability for HER-2 positive tumors. Furthermore, SUVmax, SUVmean, and SUVpeak showed positive correlations with Ki-67. In addition, SUVmax, SUVmean, SUVpeak, TBR_{blood pool}, TBR_muscle and TBR_bone were all higher in triple negative breast cancer (TNBC) than in non-TNBC, and these differences were statistically significant. The results of the ROC analysis revealed that TBR_bone exhibited the highest Area Under the Curve (AUC) of 0.8089, surpassing both SUVmax and SUVpeak, which showed AUCs of 0.7956 and 0.7867, respectively. TBR_bone was employed for discriminating triple-negative from non-triple-negative breast cancer, with a determined cutoff value of 19.19 and a corresponding Youden index of 0.6089.

Conclusions: [¹⁸F]AlF-NOTA-FAPI 04 shows promise in detecting aggressive breast cancer, especially in triple-negative breast cancer. Positive correlations with hormone receptor, grade and Ki-67 suggest its relevance in assessing tumor characteristics. Variations in uptake between triple-negative and non-triple-negative breast cancer, particularly with TBR_bone, hint at its potential clinical use in subtype identification. These findings underscore [¹⁸F]AlF-NOTA-FAPI 04 as a non-invasive tool with implications for breast cancer diagnosis. Further validation in larger studies is needed to confirm its clinical utility.