Low-activity total-body LAFOV-PET/CT in breast cancer staging and restaging: a preliminary clinical evaluation and potential adoption of ultra-fast protocol

Abstract Purpose This study aimed to evaluate (i) the diagnostic performance, image quality, and quantitative robustness of total-body long axial field-of-view PET/CT (LAFOV-PET/CT) using relatively low administered activity (1.8 MBq/kg), and (ii) the feasibility of ultra-fast acquisition protocols using retrospective count reduction in patients with invasive ductal carcinoma (IDC). Methods Twenty-four consecutive women with biopsy-proven IDC referred for staging or restaging underwent [¹⁸F]FDG total-body PET/CT using a relatively low administered activity (1.8 MBq/kg) on a uEXPLORER LAFOV system. A reference 8-minute acquisition was retrospectively reconstructed to simulate 4- and 2-minute scans, enabling evaluation of reduced-count conditions. Image quality, SUV-based quantitative metrics, and diagnostic performance were assessed to determine whether reduced-count reconstructions compromise lesion detectability. Image quality was independently assessed by two experienced readers using a 5-point Likert scale. Quantitative analysis included SUVmax, SUVmean, and tumor-to-background ratio. Diagnostic performance was evaluated on a per-lesion basis using histopathology and/or definitive lesion characterization by contrast-enhanced CT and clinical follow-up within 3 months as reference standard. Receiver operating characteristic (ROC) analysis was performed using SUVmax, with AUC confidence intervals estimated by the DeLong method and robustness assessed through leave-one-out sensitivity analysis. Results A total of 23 lesions were analyzed. PET/CT demonstrated a sensitivity of 93.3%, specificity of 100%, and overall accuracy of 95.7%. ROC analysis showed excellent discriminative performance (AUC 0.97; 95% CI: 0.90–1.00), with an exploratory SUVmax cut-off of 3.37. Sensitivity analyses confirmed the robustness of both AUC and cut-off values. Qualitative image quality did not significantly differ among 8-, 4-, and 2-minute acquisitions ( p > 0.05). Quantitative PET parameters remained stable across all acquisition durations, indicating preserved quantitative reliability even with ultra-short scans. Conclusions Low-activity total-body LAFOV PET/CT using uEXPLORER provides in this exploratory study potential excellent diagnostic performance, high image quality, and robust quantification for staging and restaging of invasive ductal breast cancer, even with acquisition times as short as 2 min. These findings support the feasibility of ultra-fast acquisition protocols and simulated low-activity imaging conditions, while preserving image quality, quantitative reliability, and lesion detectability. Further validation in larger prospective studies is warranted.