P5-11-12: Correlation of Ki67 Expression between Initial Biopsy and Surgical Specimen in Untreated Breast Cancer Patients. Does Menstrual Cycle Matter?

Abstract

Abstract Background: The measure of proliferative index by Ki67 expression is being increasingly used to identify breast cancer subtype and to guide treatment decisions. Findings on initial tumor biopsy are in some cases, such as for primary systemic treatment, the only data available prior to treatment. In addition, results from window studies to develop novel anti-tumor agents depends on initial breast cancer biopsy findings. This study aims to compare the Ki67 expression on initial biopsy with matched primary breast cancer surgical specimens. In addtion, we aimed to explore the possible effect of hormonal levels — as reflected by menstrual cycle phase — on discordant Ki67 cases. Material and Methods: A total of 50 random cases were analyzed by 2 blinded pathologists. Ki67 expression was collected as a continuous variable. Tumors were then classified as high — or low-proliferative according to a Ki67 value of > or = 15% or < 15%, respectively. Sperman Rank test was used to study the correlation between initial and surgical biopsies. Data on menstrual cycle phase at the time of biopsy and surgery were retrived from medical history of all premenopausal patients. Results: Baseline patients characteristics are as follows: median age 48 (35-56 years), premenopausal 27 (54%), postmenopausal 23 (46%), invasive ductal carcinoma 38 (76%). A differential expression of Ki67 was found in 36 patients (72%), 20 pre- and 16 post-menopausal. Spearman's Rank Order correlation did not show any significant difference between biopsy and matched surgical specimen in terms of Ki67 expression. Indeed, there was a strong positive correlation between initial and surgical biopsy, which was statistically significant across all patients (rs = .556, P < 0.001), and regardless the menopausal status. However, 15% of pre-menopausal patients transformed from Ki67 low to high. Therefore, data on pre-menopausal patients were analyzed according to the phase of menstrual cycle at the timing of initial and surgical biopsies. The Spearman's Rank Order gave no correlation among 14 premenopausal patients with differential menstrual cycle phase at the timing of initial biopsy and surgery (rs = .411, P = 0.145). Conclusions: Initial breast cancer biopsy can be used with high confidence for Ki67 determination. However, it should be noted that menstrual cycle phase could affect Ki67 expression as women with distinct menstrual cycle phase at the time of initial biopsy and final surgery do not show concordant results. Prospective evaluation of Ki67 expression in premenopausal patients is being planned. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P5-11-12.