International Journal of Surgical Pathology 18(4):343-951, 2005, Role of Chromogenic in Situ Hybridization (CISH™) in the Evaluation of HER2 Status in Breast Carcinoma: Comparison with Immunohistochemistry and FISH Elsa Li-Ning-T, MD, Ruben Ronchetti, MD, Carlos Torres-Cabala, MD, and Maria J. Merino, MD We report our experience with Chromogenic in Sine Hybridization (CISH™) for the evaluation of HER? amplification on 35 cases of formalin-fixed, paraifin-embedded invasive breast carcinomas of different histology. Al the results were corrected for chromosome 17 ancusomy and compared with immunohistochemistry (THC); a subset of cases was compared to FISH. Thirty-one of 32 cases in which FISH and CISH™ were performed yielded the same results. CISH™ and IHC showed a good concordance in the 0/1+ and 3+ category, while a poor agreement with weakly pro- tein overexpression was confirmed. Chromosome 17 analysis was necessary in cases with a low number of HER2 gene copies. CISH™ is a useful tool to evaluate breast cancer HER? status that can be easily implemented In a laboratory of surgi- cal pathology. Int J Surg Pathol 13(4):343-351, 2005 Key words: Chromogenic in Site Hybridization (CISH™), Fluorescence in Situ Hybridization (FISH), immunobistochemistry, breast carcinoma, HER2, ERBB2. ERBB2 or HER2/neu (HERZ), an oncogene located on chromosome 17q21, belongs to the human epi- ermal growth factor family, and encodes a 185 kDa transmembrane protein with tyrosine kinase activ- ity, which is involved in signal transduction [1,2}. In about 30% of breast cancers [3] there is an increase in the expression of the HER2 protein, This increase is associated with cancer progression (4). In the ma- jority of those cases, HER2 overexpression is caused Laboratory of Pathology, National Cancer Institute, Bethesda, MD. Reprint requests: Mara J. Merino, MD, National institutes of Health, NCU/Laboratory of Pathalogy, 9000 Rockville Pike, ide Y0/Room 2N212, Bethesds, MD 20892, by HER2 amplification [3,5]. The assessment of HER? protein overexpression or HER? amplification {sa prognostic marker of breast cancer and a predic- tor of responsiveness to chemotherapy treatment [4]. Determination of HER2 status can be achieved by using several methods, including 2 immunobis- tochemistry (IHC) assays for HER2 overexpression {DAKO Cytomation, Carpinteria, CA, and Ventana Medical Systems, Inc, Tucson, AZ) and 2 fluores- cence in situ hybridization (FISH) assays for HER2 gene amplification (Ventana Medical Systems, Inc, Tucson, AZ and Wysis, Inc, Downers Grove, IL) [6] that have been approved by the United States Food and Drug Administration (FDA), Recently, the assessment of HER2 status has be- come important in the selection of patients that ‘would benefit from a new tailored therapy. The new‘344 International Journal of Surgical Pathology Vol. 13 No. 4 October 2005 drug named trastuzumab (Herceptin®; Genentech Inc, San Francisco, CA) [7,8] consists of a humanized ‘monoclonal antibody directed against the extracellu- lar domain of the HER2 protein. So far, both IHC as- says and the Vysis FISH method have been approved by the FDA to be used for this specific purpose, IHC is the more common method used to ascer- tain the HER2 status in breast cancer specimens (9). However, THC results on formalin-fixed, paraffin- embedded (FFPE) tissues have been discrepant be- cause IHC for HER2 is a qualitative assay subject to high interobserver variability and can be affected by many technical factors. ‘On the other hand, the evaluation of gene ampli- fication as assessed by FISH is more reproducible [10,11]. Furthermore, it seems to be more accurate in determining expression levels of HER2 (11] anda better predictor of outcome than IHC [12]. FISH is therefore considered the most sensitive technique for the evaluation of HER2 status in breast cancer [5]. Additionally, FISH can be performed as a dual- probe test allowing the simultaneous enumeration of HER2 and chromosome 17, which is necessary in order to distinguish true HER? amplification from an increase in HER2 copies due to chromosomal aneusomy [13]. However, FISH is a laborious tech- nique that requires expensive equipment unavail- able in most pathology laboratories, such as fluores- cence microscopy and digital cameras to record the results. Moreover, the evaluation of histology on FISH slides requires a particular expertise and a par- allel comparison with H6E-stained sections. Chromogenic int situ hybridization (CISH™*) is a method introduced by Tanner etal [14] that utilizes a HER2 DNA probe generated by Subtraction Probe Technology (SPT™) (zymed Inc,, South San Fran- cisco, CA) [15]. Unlike FISH, the probe can be visu- alized by using an ordinary microscope because it uses a chromogen and a peroxidase reaction instead of fluorescence. CISH™ offers the possibility of a straightforward microscopic evaluation of gene am- plification, and the potential for concomitant mor- phologiec examination, However, CISH™ is an experimental technique that still requires extensive testing. We report here our experience with CISH™ in the determination of HER? status in breast cancer with an emphasis on its potential role in surgical pathology. The results were correlated with IHC and FISH. Mater is and Methods Fifty-five cases of breast cancer were randomly obtained from the archives of the Laboratory of Pathology at the National Cancer Institute, Bethesda, MD. The tumors were morphologically dlassified as ductal (n = 42), lobular (n= 5), pleo- morphic lobular (n = 2), and micropapillary (n = 6) carcinomas. Blank slides were obtained to perform CISH™, THC, or FISH, Each assay was evaluated blindly for the other 2. cisH™ CISH™ was performed on 4-ym-thick tissue sec- tions. The sections were baked for 2 hours at 60°C, and deparaffinized twice in xylene (10 minutes each), and 3 times in alcohol 100% (5 minutes each). The slides were air-dried and then incubated in pretreatment buffer in a microwave oven at 92°C for 15 minutes, with use of a Spot-Light FFPE reagent kit (Zymed Laboratories, Inc, San Francisco, CA), Then, they were cooled down for 20 minutes and washed with phosphate-buffered saline (PBS). Enzymatic digestion was done by applying 200 uL. of prewarmed 37°C FFPE digestion enzyme onto slides (3 minutes at 37°C). The slides were washed with PBS and dehydrated with graded alcohols (70%, 80%, 95%, and 100%, 2 minutes each). The ready-to-use digoxigenin-labeled HER-2/new probe (Zymed) or biotin-labeled chromosome 17 cen- tromere probe (Zymed) were applied onto slides, which were covered by CISH™ UnderCover Slips (Zymed); the slides were denatured on a hot plate (95°C) for 5 minutes, and hybridization was per- formed at 37°C for 18 hours. After hybridization, the slides were washed with 0.5x standard saline citrate for 5 minutes at 80°C, followed by 3 washes in PBS/0.25% Tween 20 (2 minutes each) at room temperature, Then, quenching was done with 3% H,0, diluted in methanol (10 minutes at room tem- perature}, followed by 3 washes in PBS (2 minutes each). Unspecific hydrophobic interactions were blocked with a nonspecific blocking solution (CAS- Block®™, Zymed) (10 minutes at room temperature). For some of the cases, additional slides were ob- tained and alkaline phosphatase/Fast Red was used. instead of peroxidase/DAB (Diaminobenzidine) as chromogen/substrate system, because it allows a much better contrast and visualization of the sig- nals. Thus, 2 different systems were tested for HER2/new probe detection. The first one consisted of sequential incubations with mouse antidigoxigenin, (45 minutes), PBS/0.25% Tween 20 (3 times 2 min- utes cach), polymerized HRP-goat antimouse (45 minutes), PBS/0.25% Tween 20 (3 times 2 minutes each), DAB (25 minutes), and wash with deionized ‘water (3 minutes), at room temperature, Then, they were counterstained with hematoxylin, rinsed inCISH™ for HER2 Status in Breast Cancer @ LiNing-Tetal. 845 deionized water, dehydrated, and mounted (Histo: ‘mount, Zymed). The second one was done with se quential incubations with AP-antidigoxigenin (45 minutes), PBS/0.25% Tween 20 (3 times 2 minutes each), Fast red chromogen (40 minutes, changed 4 times), rinsed in deionized water for 3 minutes, at room temperature, Then they were counterstained with hematoxylin, rinsed in deionized water and cover with Clearmount (Zymed), then incubated at 60°C for 2 hours, then dip in xylene 3 times and mounted with Histomount (Zymed) The chromosome 17 centromere probe was de- tected with sequential incubation with HRP-conju- gated streptavidin (45 minutes), PBS/0.25% Tween 20 (3 times 2 minutes each), and DAB (30 minutes) Slides were then washed (deionized water, 3 min- utes), counterstained with hematoxylin, dehy- drated, and mounted (Histomount, Zymed). CISH™ results were evaluated following Tanner's criteria [14]: unaltered gene copy number (1 to 5 signals per nucleus), low-level amplification (6 to 10 signals per nucleus in more than 50% of cancer cells), and high-level amplification (more than 10 copies or copy clusters in more than 50% of cancer cells) (Fig. 14-C). Chromosome 17 CISH™ was in- terpreted as disomy or aneusomy (Fig. 24, B) and copy numbers were compared to HER2 to define true amplification, as previously recommended [16] Ie TAC for HER2 protein was done utilizing the anti- body A0485 (DAKO, dilution 1:100) in an auto- mated method (Ventana). The IHC was evaluated according to the criteria specified by DAKO, as fol- lows: strongly positive (3+) if a strong complete membrane staining was observed in more than 10% of tumor cells; weakly positive (2+) if more than 10% of tumor cells showed weak to moderate complete membrane staining; and negative (0/1+) forall the other patterns (Fig. 3). Fig. 1. HER2 gene amplification. (A) Low-level amplification by CISH. Nuclel showing 6 to 10 brown signals, Magnification X60, (B) High-level amplification by CISH. Clusters of brown signals are seen on breast carcinoma nudlel in the center of the image. Magnification x60. (C) Dual-color FISH microphotograph displaying HER2 gene am- plification (red signals) and chromosome 17 (green signals). Magnification x100,‘346 International Journal of Surgical Pathology Vol. 13 No, 4 October 2005, Fig. 3. HER2 overexpression by IHC, (A) Detail of a 2+ score showing a complete membranous pattern of moderate FISH In total, FISH was performed in 32 of 55 cases. Sixteen cases were performed using Vysis Pathvys- ion™ HER-2 DNA Probe (Vysis; Downers Grove, IL) in an outside Laboratory (American Medical Labo- ratories, INC., Chantilly, VA}. The other 16 cases were performed in our laboratory on 4-1m-thick tissue sections with use of the Vysis PathVision™ pretreatment and DNA probe kit. The slides were preheated at 58°C overnight, deparaffinized in Cit- Intensity, (B) detail of a 3+ score showing a complete membranous Pattern of strong intensity. Magnification x40, risoly (Fisherbrand, Decon Laboratories, Inc., Bryn ‘Mawr, PA), rehydrated in ethanol, and immersed in 4 0.2N HCI solution for 20 minutes, The slides were then immersed in Vysis Pretreatment Reagent for 30 minutes at 80°C in a waterbath. After rinsing in ‘wash buffer (Vysis) the slides were digested with a protease solution (Vysis) at 37°C for 10 to 20 min- utes, Enzymatic digestion was assessed as previ- ously reported [17]. After rinsing again, 10 pl. of ‘Wysis PathVysion LSI-HER2/CEP-17 probe was ap- plied. The slides were denatured for 5 minutes atCISH™ for HER2 Status in Broast Cancer @ LiNing-Tet al. 947 72°C in a 70% formamide/2XSSC solution, then covered with Zymed UnderCoverslips and hy- bridized overnight at 37°C. Posthybridization wash- {ing was performed in a 2XSCC solution with 0.3% NP4O at 72°C in a waterbath, The slides were then air-dried in the dark and counterstained with 20 pL. of DAPI (4,6-diamidino-2-phenylindole). The FISH signals were visualized by using a Nikon Eclipse E1000 Fluorescent Microscope and pictures were captured with a Photometrics Cool SNAP-FX CCD camera (Photometries, Roper Scientific Inc., Tucson, AZ). Enumeration was done following the manu- facturer’s guidelines, HER2 and Chromosome 17 signals were counted in at least 60 nuclei; a LSI- HER2/CEP-17 ratio > 2.0 was considered positive for HER2/neu gene amplification. Results cIsH Fourteen of 55 (25.45%) cases showed HER2 gene high-level amplification (clusters or more than 10 copies). These cases included 3 micropapillary, 10 ductal, and 1 pleomorphic lobular carcinomas. Six cases (10.91%) were classified as low HER2 gene amplification (between 6 to 10 copies). Histologic varieties in this group were ductal (5) and mi- cropapillary (1) carcinomas. The remaining 35 cases were negative (Table 1). Seven (12.72%) cases showed Chromosome 17 aneusomy. The rest of the cases (48) were diploid for chromosome 17. Four of 6 cases displaying HER2 gene low ampli- fication showed chromosome 17 aneusomy (polysomy). These tumors were ductal (3) and mi- cropapillary (1) types. Therefore, they were not considered to be true gene amplification (also called pseudoamplification) (Table 1) Immunohistochemistry Fourteen (25.45%) cases were classified as 0/1+ staining score. These included ductal (9), lobular (3), pleomorphic lobular (1), and micropapillary (1) tumors. Twenty-eight (50.9%) cases displayed a 2+ score. They were lobular (2), ductal (21), pleomor- phic lobular (1), and micropapillary (4) carcinomas. A 3+ score was assigned to 13 (23.6%) cases. These cases were ductal (12) and micropapillary (1) carci- nomas (Table 2) FISH Of the 32 cases analyzed by FISH, 1 was lobular, another micropapillary, and the rest were ductal breast carcinomas. Twenty-two of 32 cases were considered negative since no amplification was seen, The IHC analysis of these cases showed 0/1+ (8 cases) and 2+ (14 cases} scores. Ten cases were positive for FISH. Nine showed strong positivity Table 1, HER-2 and Chromosome 17 Analysis by Chromogenic in Situ Hybridization (CISH) HERD ‘Chromosome 17 Histologic Type Unalteed Low-Level Amplification High-Level Amplification Disomy-Ancusomy Labular (5) 5 ° ° 4 t Ductal (42), 7 co to a7 ea Pleomorphie lobular (2) 1 ° 1 2 ° Micropapillary (6) 2 & 3 3 a *Cases displaying low-level amplification and aneusomy were not considered true amplfication. Two cases showed high-Ievelampll- lication and aneusomy, they were considered true amplifications Table 2. HER-2 Chromogenic in Situ Hybridization (CISH) and Immunohistochemistry (IHC) in Breast Cancer HER? (HC) HER-2 (CIS) Histologic Type one 2 » Amplification No Amplification Lobular (3) 2 0 5 Ductal (42), a1 2 2 30 Pleomorphic lobular (2) Mieropapillary (6) 1 4 t 1 3 3