Predictive Marker (PM) Discordance between Primary and Metastatic Breast Cancer: The Role of Decalcification and Targeted Therapy
Gary Tozbikian, Amy Ziober, Paul Zhang. Hospital of the University of Pennsylvania, Philadelphia, PA
Background: Discordance in PM between primary (PBC) and metastatic breast cancer (MBC) is well documented and impacts clinical management and survival. Clonal selection due to targeted therapy (TT), sample size, fixation, decalcification and low PM status are possible causes. In this study we evaluated these potential causes for PM discordance in MBC.
Design: A 3 year retrospective evaluation of MBC identified 80 patients with 88 MBC with PM available for comparison with that of the PBC. History of TT treatment, sample size (biopsy vs. excision), decalcification or not and MBC site were collected. All HR and HER2 tests were performed using DAKO PharmDx kits and pairwise comparison made between PBC and MBC. Discordance was defined by gain or loss of HR (Allred score 3-8 to 0-2) or HER2 (3+ to 0-1+), and alteration of HR defined by (6-8 to 3-5) and HER2 (3+ or 1+ to 2+).
Results: Of 80 PBC, 56% (46) were HR+, 16% (13) HER2+, and 26% (21) triple negative. MBC sites included lung (24), bone (21), liver (18), brain (8), and others (17). 13% (12) of MBC had discordance including loss of ER (10), HER2 (1) and gain of ER (1); and 24% (21) had alteration in ER (8), HER2 (12), and both (1). 17 of the bone MBC were decalcified, 24% (4) of which were discordant and 47% (8) altered. Decalcification was significantly correlated with loss or decline of ER (P=0.002) but not HER2. 55 patients had hormone therapy (HRT) and 14 trastuzumab. Trastuzumab was not correlated with discordance or alteration of HER2 (P=0.157). HRT was significantly correlated with loss or decline of ER (P=0.027) but not when decalcified cases were excluded (P=0.13). Discordance was not correlated with sample type/size (P=0.44). Low ER positivity was seen in 14% of PBC and not significantly correlated with HR discordance or alteration in MBC (P=0.08, P=0.29).
Conclusions: 13% of MBC had discordant PM, mostly as loss of ER. Decalcification is significantly associated with discordance or decline in HR status on MBC, but not with HER2 status. HRT has a significant correlation with HR decline, but not when effect of decalcification/bone MBC were eliminated in the analysis. This finding concurs with prior studies comparing HR in decalcified and non-decalcified tissue of the same tumor and further emphasizes the adverse effect of decalcification in HR tests in MBC. Our study also showed sample size and low level PM status were not significantly correlated with discordance. The role of TT in discordant or altered HR and HER2 expression in MBC needs to be further studied in larger series.
Monday, March 19, 2012 9:30 AM
Poster Session I Stowell-Orbison/Surgical Pathology/Autopsy Awards Poster Session # 18, Monday Morning