Use of Ultrasound and Ki-67 Proliferation Index to Predict Breast Cancer Tumor Response to Neoadjuvant Endocrine Therapy

Background: Prediction of tumor shrinkage and pattern of treatment response following neoadjuvant endocrine therapy (NET) for estrogen receptor positive (ER+), Her2 negative (Her2-) breast cancers have had limited assessment. We examined if ultrasound (US) and Ki-67 could predict the pathologic response to treatment with NET and how the pattern of response may impact surgical planning. Methods: A total of 103 postmenopausal women with ER+, HER2- breast cancer enrolled on the FELINE trial had Ki-67 obtained at baseline, day 14, and surgical pathology. A total of 70 patients had an US at baseline and at the end of treatment (EOT). A total of 48 patients had residual tumor bed cellularity (RTBC) assessed. The US response was defined as complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). CR or PR on imaging and <= 70% residual tumor bed cellularity (RTBC) defined a contracted response pattern. Results: A decrease in Ki-67 at day 14 was not predictive of EOT US response or RTBC. A contracted response pattern was identified in one patient with CR and in sixteen patients (33%) with PR on US. Although 26 patients (54%) had SD on imaging, 22 (85%) had RTBC <= 70%, suggesting a non-contracted response pattern of the tumor bed. The remaining four (15%) with SD and five with PD had no response. Conclusion: Ki-67 does not predict a change in tumor size or RTBC. NET does not uniformly result in a contracted response pattern of the tumor bed. Caution should be taken when using NET for the purpose of downstaging tumor size or converting borderline mastectomy/lumpectomy patients.