Chen, B

Abstract Background: Epidermal Growth Factor Receptor (EGFR, HER1) is known to heterodimerize with HER2 and may impact the effectiveness of trastuzumab. Historically, EGFR has been difficult to measure; thus there is no evidence regarding its effects on trastuzumab clinical therapy. Here we use a new, standardized, quantitative immunofluorescence assay and a novel EGFR antibody to evaluate the correlation between EGFR expression and clinical outcome in a cohort from the North Central Cancer Treatment Group (NCCTG, Alliance) N9831 trial. Methods: Of the 3505 women in the N9831 trial, tissue microarrays were constructed that allowed analysis of 1,444 patients randomly assigned to receive chemotherapy alone (Arm A), sequential trastuzumab after chemotherapy (Arm B) or concurrent trastuzumab with chemotherapy (Arm C). Fluorescence-based, standardized measurement of EGFR was done using the rabbit monoclonal EGFR antibody D38B1 on the AQUA platform (HistoRx). EGFR expression levels were analyzed as a continuous variable and dichotomized using recursive partitioning to define an optimal cut point. The cut point was validated on an independent retrospective cohort comprised of 130 patients with HER2-positive metastatic breast cancer who received trastuzumab. Results: EGFR assessed as a continuous variable shows an association with disease free survival (DFS) in Arm C (p = 0.002) but not in Arms A or B. Dichotomizing EGFR expression shows that high level expression is associated with worse outcome (HR = 2.2; 95% CI 1.33–3.59, p = 0.002) in arm C. Five year DFS within the low level EGFR expression group was 71%, 74% and 90% for Arms A, B and C, respectively (log-rank p-value=0.02) and within the high level EGFR expression (n = 213) group was 62%, 76% and 74% respectively (log-rank p-value 0.22). As validation, the same cut-point was used to define high EGFR expression in a retrospective metastatic breast cancer cohort. The median progression free survival difference between EGFR high group and EGFR low group was 6.1 months (p = 0.0063) and the hazard ratio of high EGFR was 1.92 (p = 0.0073). Conclusions: High expression of EGFR by the AQUA platform appears to be associated with decreased benefit from adjuvant concurrent (not sequential) trastuzumab and shorter PFS in the metastatic setting. The biological underpinning for these observations is being considered. Since there are other treatment options for HER2-driven tumors, this marker, once validated, could help select patients for alternative or additive therapy. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr S5-4.

Abstract Background: Unstained recuts from formalin-fixed paraffin-embedded tissues are commonly collected for cooperative group studies. There is concern among pathologists that improper storage conditions can lead to antigen degradation. In an effort to quantify this effect, we compared the expression of HER1 and HER2 on two sets of identical cohort tissue microarrays (TMAs) from the N9831 HER2+ adjuvant phase III trial (NCT00005970; www.clinicaltrials.gov); one freshly cut set (cut April 18, 2011) and a second set stored at 4 degrees for over two years (cut between Nov, 2007 and Jan, 2008). Methods: The two sets of TMA slides containing 1580 tumor samples from the N9831 cohort were treated identically using the AQUA method of quantitative immunofluorescence. HER1 was tested with D38B1 (rabbit monoclonal, Cell Signaling Technology, Inc.) and HER2 with CB11 (mouse monoclonal, Biocare, Inc.) on tumors from 695 patients (712 specimens) in the fresh TMAs and 779 patients (800 specimens) in the old TMAs in up to three-fold redundancy per specimen. Results: Frequency distributions of the expression of HER2 revealed bimodality in the fresh TMAs compared to an attenuated distribution of the old cases. The average score of the entire cohort was significantly lower in old TMAs compared to fresh cuts (paired t-test, p<0.0001). Linear regression of the average HER2 scores from new TMAs versus the average scores from old TMAs showed a slope term of 0.52, which is statistically significantly different from the hypothetical value of 1 (p<0.0001). Regressions between any two fresh slides showed slopes close to 1.0. Similar results were seen for HER1, but fewer positive cases made the changes less dramatic. Conclusions: The storage condition of tissue slides is a critical pre-analytical variable that can dramatically lower the score of HER1 and HER2, artificially. Thus, studies done on inadequately stored slides, either whole sections or TMAs, must be interpreted with caution. Tissue collection and analysis of biomarkers for cooperative group studies should not include unstained recuts, but rather, entire blocks or large cores from tissue blocks. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr PD05-04.