Purpose: To assess whether the early metabolic response evaluated by (18)F-fluorodeoxy-glucose positron emission combined with computed tomography (FDG PET/CT) predicts the morphological, pathological, and cell-cycle responses to neoadjuvant endocrine therapy of hormone receptor-positive primary breast cancer.
Study design: Eleven patients (12 tumors) with estrogen receptor-positive (Allred score 7 or 8) primary breast cancer were enrolled. All patients received a daily dose (2.5 mg) of letrozole for 12 weeks followed by surgery. Sequential FDG PET/CT scans were performed before treatment (baseline), at 4 weeks after the initiation of endocrine therapy (PET2), and prior to surgery (PET3). Tumors showing a 40% or more reduction and those showing a less than 40% reduction in the standardized uptake value maximum (SUV(max)) at PET2 compared with the baseline PET were defined as metabolic responders and metabolic nonresponders, respectively. Change in tumor size as measured by ultrasound (morphological response), pathological response, and change in the Ki67 labeling index in tumor tissue (cell-cycle response) during the neoadjuvant letrozole therapy were compared between the metabolic responders and nonresponders.
Results: The average decreases in SUV(max) at PET2 compared with the baseline PET in the metabolic responders (n = 6) and the metabolic nonresponders (n = 6) were 60.9% (±21.3 SD) and 14.2% (±12.0 SD), respectively. At PET3 compared with the baseline PET, the metabolic responders showed a significantly higher decrease of 64.5% (±18.7 SD) (p = 0.0004), whereas the nonresponders showed a nonsignificant decrease of 16.7% (±14.1 SD) (p = 0.06). The morphological and pathological responses after letrozole therapy did not differ between the metabolic responders and nonresponders. The metabolic responders showed a marked decrease in the Ki67 labeling index at 2 weeks after the initiation of treatment (62.9%, ±35.9 SD, p = 0.04) and at surgery (91.7%, ±10.7 SD, p = 0.03) compared with the baseline values. In contrast, metabolic nonresponders showed no significant change in the Ki67 index either after 2 weeks of therapy or at surgery.
Conclusion: Cell-cycle response monitored by the Ki67 labeling index correlates with metabolic response monitored by tumor SUV(max). Monitoring of tumor SUV(max) using FDG PET/CT may be feasible to predict cell-cycle response to neoadjuvant endocrine therapy of primary breast cancer.