Evaluation of Epidermal Growth Factor Receptor Expression and Gene Amplification in Colorectal Carcinoma: An Immunohistochemical and Chromogenic In Situ Hybridization Study Using Tissue Microarray
RS Saad, C Rowsell, MA Khalifa. Sunnybrook Health Sciences Center/University of Toronto, Toronto, ON, Canada
Background: Recent data has shown that immunohistochemistry (IHC) for detection of EGFR protein is not useful for prediction of response of colorectal cancer to anti-EGFR drugs. Chromogenic in situ hybridization (CISH) is shown to be a potential alternative to FISH. In this study, we assess EGFR gene amplification in colorectal adenocarcinoma using CISH in comparison to protein expression detection by IHC in a tissue microarray (TMA).
Design: One hundred and twenty five (n=125) paraffin-embedded tissue blocks of histologically-confirmed primary colorectal adenocarcinoma were cored twice at a diameter of 1.5 mm, using TMA technology (TMArrayer 100). We also included 50 cases of metastatic colorectal carcinoma to the liver. The cores were re-embedded into the final recipient block. IHC was performed using anti-EGFR monoclonal antibody (31G7). CISH protocol was applied based on the use of EGFR gene. EGFR IHC was scored in 4-tiered system based on the percentage of positive cells: 0 no stain, 1+ (1-10%), 2+ (11-50%) and 3+ (> 50%).
Results: With IHC, EGFR was detected in 86/175 (49%) cases: strong membrane staining (3+) in 31/175 (18%) cases, moderate (2+) in 21/175 (12%) and weak staining (1+) in 34/175 (19%). The remaining cases, 89/175 (51%), were negative. CISH demonstrated gene amplification (>4 copies/nucleus) in 46/175 (26%): 19/31 (61%) of (3+) IHC cases, 5/21 (24%) of (2+) IHC, 7/34 (21%) of (1+) IHC cases, and 15/89 (17%) of negative cases. High level gene amplification (>10 copies/nucleus) was only seen in 3/175 cases. There was a significant correlation between IHC and CISH staining (r=0.28, P< 0.05). IHC (3+) predicted a higher percentage of CISH amplification than other grades of staining with a statistically significant difference. In metastatic carcinoma, CISH showed gene amplification in 26/50 (54%) cases, in comparison to 20/125 (16%) in primary colorectal carcinoma, with a significant difference (P< 0.05).
Conclusions: Strong IHC staining (3+) is predictive of EGFR gene amplification, which may still be clinically significant. In some cases gene amplification was only focal offering a potential explanation for poor response to targeted therapy in EGFR-positive tumors. CISH EGFR may represent a more effective tool for patient selection for targeted treatment. This study also shows that only a small fraction of EGFR-positive colorectal carcinomas detected by IHC truly have gene amplification.
Tuesday, March 23, 2010 1:00 PM
Poster Session IV # 82, Tuesday Afternoon