Impact of biomarker changes during neoadjuvant chemotherapy for clinical response in patients with residual breast cancers.

Abstract

Residual cancer burden or Ki67 expression levels in residual tumors reportedly provided significant prognostic information for a non-pathological complete response subset after neoadjuvant chemotherapy (NAC). However, the significance of Ki67 reduction for clinical response during chemotherapy in each subtype or menopausal status is yet to be determined. A total of 183 breast cancers surgically removed after chemotherapy were recruited for this study. Expression levels of estrogen receptor (ER), progesterone receptor (PgR), and Ki67 were determined immunohistochemically for semiquantitative measurement and these biomarkers were compared in pre- and post-NAC samples from pathological non-responders (n = 125). Responses to chemotherapy were evaluated both clinically and pathologically. Ki67 expression levels after NAC (median 5 %, range 0-70 %) were significantly reduced compared with before NAC (25, 1-80 %, P < 0.0001), but only in patients who attained clinical response. This significant suppression of Ki67 in clinical responders was consistently observed in breast cancers from the ER-positive subset, but not the ER-negative subset in the total test set (n = 120). These observations were also made in the validation set (n = 63). Among premenopausal, but not postmenopausal patients, a significant decrease in PgR expression levels was detected in breast cancers of patients who attained clinical response (pre-NAC 50, 0-100 %, post-NAC 5, 0-20 %; P = 0.0003). The impact of Ki67 suppression on clinical response seems to be restricted to ER-positive breast cancers. Since PgR expression levels of premenopausal ER-positive cancers were significantly reduced in clinical responders, inhibition of estrogen signaling due to chemotherapy-induced amenorrhea may be involved in this association.