RESUMEN
Previous studies have suggested that early changes in blood flow (BF) in response to neoadjuvant chemotherapy and evaluated with 15O-water are a surrogate biomarker of outcome in women with breast cancer. This study investigates, in the triple-negative breast cancer subtype, the prognostic relevance of tumor BF changes (ΔBF) in response to chemotherapy, assessed using a short dynamic 18F-FDG PET acquisition. METHODS: Forty-six consecutive women with triple-negative breast cancer and an indication for neoadjuvant chemotherapy were prospectively included. Women benefited from a baseline 18F-FDG PET examination with a 2-min chest-centered dynamic acquisition, started at the time of 18F-FDG injection. Breast tumor perfusion was calculated from this short dynamic image using a first-pass model. This dynamic PET acquisition was repeated after the first cycle of chemotherapy to measure early ΔBF. Delayed static PET acquisitions were also performed (90 min after 18F-FDG injection) to measure changes in tumor glucose metabolism (ΔSUVmax). The association between tumor BF, clinicopathologic characteristics, and patients' overall survival (OS) was evaluated. RESULTS: Median baseline tumor BF was 21 mL/min/100 g (range, 6-46 mL/min/100 g) and did not significantly differ according to tumor size, Scarf-Bloom-Richardson grade, or Ki-67 expression. Median tumor ∆BF was -30%, with highly scattered values (range, -93% to +118%). A weak correlation was observed between ΔBF and ∆SUVmax (r = +0.40, P = 0.01). The median follow-up was 30 mo (range, 6-73 mo). Eight women developed recurrent disease, 7 of whom died. Low OS was associated with menopausal history (P = 0.03), persistent or increased tumor vascularization on the interim PET (ΔBF cutoff = -30%; P = 0.03), non-breast-conserving surgery (P = 0.04), and the absence of a pathologic complete response (pCR) (P = 0.01). ΔBF and pCR provided incremental prognostic stratification: 3-y OS was 100% in pCR women, 87% in no-pCR women but achieving an early tumor BF response, and only 48% in no-pCR/no-BF-response women (ΔBF cutoff = -30%, P < 0.001). CONCLUSION: This study suggests the clinical usefulness of an early user- and patient-friendly 2-min dynamic acquisition to monitor breast tumor ΔBF to neoadjuvant chemotherapy using 18F-FDG PET/CT. Monitoring tumor perfusion and angiogenesis response to treatment seems to be a promising target for PET tracers.