Prediction of pathological response using 18F FDG PET/CT derived metabolic parameters in locally advanced breast cancer patients.

Abstract

This study aims to assess the value of flurodeoxyglucose (FDG)-PET derived metabolic parameters for prediction of pathologic response in LABC postneoadjuvant therapy. Totally 47 patients with LABC underwent initial and postneoadjuvant therapy PET scans. ΔSUVmax%, ΔTLG% and ΔMTV% were calculated. Post-therapy histopathologic therapeutic response was assessed. In total 91.5% of patients had invasive duct carcinoma and the remaining (8.5%) had invasive lobular carcinoma. Postneoadjuvant PET/CT was able to detect 91.7% of patients with pathologically proven complete response in primary tumor, 69% of those with Pathologic partial response and 88.3% of those with pathological no response (P value <0.001). However, 40 out of the 47 patients had regional nodal metastases. PET/CT was able to predict 57.1% of the patients with pathologically nonresponding nodal deposits and 93.9% of those revealed pathologic therapeutic effect (P value <0.001). Receiver operating characteristic curve (ROC) curve marked Δ1ry SUVmax of 26.25% (P value 0.003), Δ1ry TLG of 48.5% (P value 0.018). PET and pathological response correlated well with ΔSUVmax%, and Δ1ry TLG% correlated well with PET, pathologic response and expression of HER II receptors (P value <0.001, 0.003 and 0.037 respectively). ROC curve marked ΔLN SUVmax% of 80.15% (P value 0.012), ΔLN TLG% of 86.6% (P value 0.002), whereas for ΔLN MTV% cut off point of 55% (P value 0.003). ΔSUVmax%, ΔTLG % and ΔMTV% for regional nodal metastases, were significantly correlated with PET (P values <0.001, <0.001 and 0.003, respectively) and pathologic (P values 0.018, 0.001 and 0.002, respectively) response. FDG-PET is a useful tool for monitoring the neoadjuvant therapeutic effect for primary and regional nodes in patients with LABC.