Early and delayed prediction of axillary lymph node neoadjuvant response by (18)F-FDG PET/CT in patients with locally advanced breast cancer

To determine the utility of (18)F-FDG (FDG) PET/CT performed in an early and delayed phase during neoadjuvant chemotherapy in the prediction of lymph node histopathological response in patients with locally advanced breast cancer. FDG PET/CT studies performed in 76 patients (mean age 53 years) at baseline (PET-1), after the second course of chemotherapy (PET-2) and after the last course of chemotherapy (PET-3) were prospectively analysed. Inclusion criteria were lymph node involvement detected by PET/CT and non-sentinel node biopsy before or after the baseline PET/CT scan. Following the recommendations of the 12th International Breast Conference (St. Gallen), the patients were divided into five subgroups in relation to biological prognostic factors by immunohistochemistry. For diagnosis visual and semiquantitative analyses was performed. Absence of detectable lymph node uptake on the PET-2 or PET-3 scan with respect to the PET-1 scan was considered metabolic complete response (mCR). Lymph nodes were histopathologically classified according the lymph node regression grade and in response groups as pathological complete response (pCR) or not pCR (type A/D or B/C of the Smith grading system, respectively). ROC analysis was performed to determine a cut-off value of Δ% SUV1-2 and SUV1-3 for prediction of nodal status after chemotherapy. An association between mCR and pCR was found (Cohen's kappa analysis), and associations between phenotypes and metabolic behaviour and the final histopathological status were also found. Lymph node pCR was seen in 34 patients. The sensitivity, specificity, and positive and negative predictive values of PET-2 and PET-3 in establishing the final status of the axilla after chemotherapy were 52 %, 45 %, 50 % and 47 %, and 33 %, 84 %, 67 % and 56 %, respectively. No significant relationship was observed between mCR on PET-2 and PET-3 and pCR (p = 0.31 and 0.99, respectively). Lymph node metabolism on PET-1 was not able to predict the final histopathological status, whereas basal carcinomas showed a higher rate of pCR (70.6 %) than the other groups (p = 0.03). FDG PET/CT seems to have limitations in both the early and delayed evaluation of lymph node status after chemotherapy, with reduced predictive values.