Early assessment of response to aromatase inhibitor (AI) therapy

Abstract

11075 Background: Early decline in proliferative index (Ki-67) predicts response to endocrine therapy. For estrogen receptor positive (ER+) tumors, a low Ki-67 two weeks into endocrine therapy also predicts response (Dowsett JNCI 2007). FDG PET is promising for evaluating early response to systemic therapy and has the advantage of being non-invasive and of providing simultaneous quantitative measures at multiple tumor sites in vivo. We tested the association between FDG PET SUV and Ki-67 after two weeks of aromatase inhibitor (AI) therapy, and their prediction of response. Methods: Patients undergoing AI therapy were eligible for enrollment in the imaging study, and underwent a two week “run-in” of AI monotherapy prior to continued monotherapy or the addition of chemotherapy. Patients underwent baseline FDG PET, and Ki-67 was measured in tumor biopsy specimens. Following two weeks of therapy FDG PET and biopsy were again performed. Patients on AI therapy only were followed for clinical response. Results: A total of 21 patients enrolled in the study. 16/21 (76%) of patients had a greater than 20% decline in FDG SUV, including 11 with analyzable serial biopsy, all with 2-week Ki67 ≤5%. Of the 5 patients who did not show greater than 20% decline in FDG SUV, all 3 with serial biopsy had 2-week Ki67 of ≥20%. Subsequent response to AI was evaluable at 4–6 months in 8 patients who underwent AI monotherapy (5 R, 3 PD). All 5 responders showed decline of at least 20% in FDG SUV (30% - 60% decline), and 2/3 with PD did not (no change, 3% decline, 29% decline). Conclusions: We observed substantial changes in Ki-67 and FDG SUV following 2 W of targeted therapy. Decline in FDG SUV in patients on single agent AI predicted response in almost all cases. Serial FDG PET appears promising as an early indicator of response to AI therapy. No significant financial relationships to disclose.