It has become important to accurately evaluate the status of HER-2/neu in invasive breast cancer, especially when one is considering the use of anti-HER-2 monoclonal antibody therapy (Trastuzumab). Almost one third of invasive breast carcinomas overexpress the HER-2/neu protein, so the use of the anti-HER-2/neu monoclonal antibody Herceptin (trastuzumab) to block the protein has become important in the management of and in prolonging the survival for patients with metastatic breast cancer. The effectiveness of this therapy is dependent on accurately evaluating the HER-2 status in these tumors, which can be done either by studying the expression of HER-2 protein by immunohistochemistry (IHC) or by evaluating HER-2 gene amplification by fluorescent in situ hybridization (FISH). Since interobserver variability may occur in manually grading HER-2 protein expression by IHC, the aim of this study was to compare the HER-2/neu expression by IHC using a computer-based image analysis system with that of the gene amplification by FISH. Formalin-fixed paraffin-embedded archival tissue from 108 primary infiltrating ductal carcinomas were immunostained using the HercepTest (DAKO). To reduce interobserver variability, membrane staining was evaluated using the Automated Cellular Imaging System (ACIS) by ChromaVision, and the cases were divided into four groups: group 1 (n=23) with HER-2/neu expression ACIS score less than or equal to 1.5; group 2 (n=17) with a score ranging from 1.6 to 1.9; group 3 (n=46) with a score 2.0 to 2.5; and group 4 (n=22) with a score greater than or equal to 2.6. FISH was performed on all of the 108 cases using the PathVysion HER-2/neu DNA probe kit from Vysis Inc. All cases were also manually reviewed and graded as negative, 1+, 2+, and 3+ according to the DAKO HercepTest grading scheme. Cases with negative and 1+immunostaining were considered as HER-2 not overexpressed, and cases with 2+ and 3+ staining were classified as showing HER-2 overexpression. In group 1, 1 of 23 (4%), in group 2, 2 of 17 (12%), in group 3, 5 of 46 (11%), and in group 4, 19 of 22 (86%) cases showed gene amplification by FISH. Furthermore, in group 4 all 15 (100%) cases with an ACIS score of 3 or greater were FISH positive. Correlation with manual IHC score and FISH showed that 2 of the 23 (9%) IHC negative (0 and 1+) cases and 25 of the 85 (29%) IHC positive (2+ and 3+) cases showed gene amplification by FISH. This study shows that the amplification of the HER-2/neu gene correlates better with overexpression of the HER-2/neu protein by IHC when the score is either less than 1.5 or greater than 2.6 by ACIS. Therefore, FISH may be useful to better evaluate HER-2/neu status in breast cancer in cases where the ACIS score by immunohistochemistry is 1.6 to 2.5, and since the correlation is so good, FISH may not be needed for HER-2 evaluation in cases with ACIS scores less than 1.5 and greater than 2.6.