Early assessment of axillary response with (18)F-FDG PET/CT during neoadjuvant chemotherapy in stage II-III breast cancer: implications for surgical management of the axilla

Background: If all initially node-positive patients undergo axillary lymph node dissection (ALND) after neoadjuvant chemotherapy (NAC), overtreatment may occur in patients with complete response. Positron emission tomography–computed tomography (PET/CT) during NAC may predict axillary response and select patients appropriate for less invasive treatment after NAC. We evaluated the value of sequential 18F fluorodeoxyglucose (FDG) PET/CTs during NAC for axillary response monitoring in stage II–III breast cancer. Methods: A total of 219 PET/CTs were performed in 80 patients with cytology-proven, node-positive disease at baseline (PET/CT1, n = 80) and twice during NAC (PET/CT2 n = 62, PET/CT3, n = 77). The relative changes in maximum standardized uptake value (SUVmax) of axillary nodes were examined for their ability to assess pathological response. All patients underwent ALND after chemotherapy, and complete axillary response (pCR), defined as absence of isolated tumor cells and of micro- and macrometastases, served as the reference standard. Results: A total of 32 (40 %) patients experienced axillary pCR. The relative decrease in SUVmax was significantly higher in patients with pCR than in those without, both on PET/CT2 (p < 0.001) and PET/CT3 (p = 0.025). The area under the receiver operating characteristic curve values for PET/CT2 and PET/CT3 were 0.80 (95 % confidence interval 0.68–0.92) and 0.65 (95 % confidence interval 0.52–0.79), respectively. A relative decrease of ≥60 % on PET/CT2 had an excellent specificity (35 of 37, 95 %), a high positive predictive value (12 of 14, 86 %), and a sensitivity of 48 %—that is, it accurately identified histologic pCR in 12 of 25 patients with disease that responded to therapy. Conclusions: 18F-FDG PET/CT early during NAC is useful for axillary response monitoring in cytology-proven node-positive breast cancer because it identifies pathological response, thus permitting ALND to be spared.Bas B. Koolen, Renato A. Valdés Olmos, Jelle Wesseling, Wouter V. Vogel, Andrew D. Vincent, Kenneth G. A. Gilhuijs, Sjoerd Rodenhuis, Emiel J. Th. Rutgers, Marie-Jeanne T. F. D. Vrancken Peeter