ORIGINAL ARTICLE Assessment of HER-2 Status in Pancreatic Adenocarcinoma Correlation of Immunohistochemistry, Quantitative Real-Time RT-PCR, and FISH With Aneuploidy and Survival Alex J. Saxby, MB BChir,* Aiqun Nielsen, MSc,* Christopher J. Scarlett, BSc (Hons),* Adele Clarkson, BSc,† Adrienne Morey, FRCPA,‡ Anthony Gill, FRCPA,† and Ross C. Smith, FRACS* Abstract: HER-2 is a transmembrane growth factor receptor recognized in overexpression as an independent adverse prognostic factor in several cancers. This study measured HER-2 overexpression in pancreatic adenocarcinoma at the genetic, transcriptional, and translational level. Expression was gauged with regard to stage, grade, and survival. Pancreatic adenocarcinoma samples (n = 30) were analyzed with immunohistochemical labeling for HER-2 protein, Quantitative real-time reverse transcriptase polymerase chain reaction (Q-RT-PCR) measurement of HER-2 mRNA and fluores- cence in situ hybridization (FISH) analysis of HER-2 gene expres- sion. HER-2 expression in benign pancreatic lesions (n = 10) provided a control. Five (17%) of the pancreatic adenocarcinomas scored maximal 3+ immunohistochemistry (IHC) labeling, seven (23%) had significantly increased expression of HER-2 mRNA, while only one (3%) exhibited low level HER-2 gene amplification. Ten (33%) tumors demonstrated aneuploidy. In general, concordance be- tween methodologies was poor, but the best agreement was seen between FISH aneuploidy status and Q-RT-PCR mRNA over- expression (80% agreement), followed by IHC and Q-RT-PCR (73% agreement). The least agreement was seen between IHC and FISH aneuploidy status (67% agreement). Tumor stage was positively as- sociated with HER-2 mRNA and protein expression, but tumor grade and other patient characteristics did not reach statistical significance. A poor survival outcome was demonstrated with positive HER-2 status in all three measures of overexpression (Kaplan-Meier log-rank score; P , 0.01 [IHC], P = 0.05 [Q-RT-PCR], P = 0.02 [FISH]). Discordance in expression at the nuclear, cytoplasmic, and cell surface levels highlights the limitations of immunohistochemical evaluation alone and stresses the need for further evaluation of response to anti-HER-2 targeted therapies in tumors displaying overexpression in gene copy, mRNA, and receptor protein. Key Words: pancreatic adenocarcinoma, HER-2, quantitative real- time RT-PCR, fluorescence in situ hybridization, survival (Am J Surg Pathol 2005;29:1125–1134) P ancreatic cancer remains the fourth leading cause of cancer- related deaths in the United States with around 31,800 pancreatic cancer deaths predicted this year. 11 Of the 10 most common cancers, pancreatic cancer has the lowest 5-year survival rate, quoted as 4% in the United States. 10 Growth factors play a key role in the cancer process. They are important in cell proliferation, adhesion, and survival and therefore influence processes such as tumor growth, inva- sion, and metastasis. Human epidermal growth factor receptor 2 (HER-2) belongs to a family of four transmembrane tyrosine kinase receptor proteins known collectively as the HER or erbB family. Signal transduction involves activation of the tyrosine kinase causing a cascade of intracellular enzyme regulated pathways. Such activation requires dimerization with a second HER receptor. Huge signaling variety is permitted by multiple dimer combinations and a range of different activating ligands. 23 HER-2 appears pivotal in the signaling pathway of the HER family for several reasons. Not only is it the preferred dimeri- zation partner for the other three family members, but the dimers formed are more stable and the signals more intense. 7 HER-2 is overexpressed in a number of different cancer types, 34 in particular, breast cancer where the expression of HER-2 has important prognostic significance. 28 A recent review of 81 series, including 27,161 breast cancer patients, found 90% of studies demonstrated a positive correlation between HER-2 overexpression and negative outcome. 23 Reports of over- expression in pancreatic cancer vary considerably with values quoted as low as 0% 21 and as high as 82%. 5 Most of these studies assessed expression by immunohistochemical labeling alone, and few have correlated this with other methodologies. Such huge variability puts into question the use of HER-2- targeted therapy in pancreatic cancer unless a more repro- ducible way of assessing true HER-2 expression can be found. To better understand the expression of HER-2 in pancreatic cancer, this study applied three different measures: assessing gene amplification, cytoplasmic mRNA, and re- ceptor labeling. It is the largest published study to date to From the *University of Sydney, Department of Surgery, Royal North Shore Hospital; †Department of Anatomical Pathology, Royal North Shore Hospital; and ‡Department of Anatomical Pathology, St. Vincents Hospital, Sydney, Australia. Supported by the Cancer Surgery Research Foundation, CanSur (Research Support Grant) and Roche Pharmaceuticals (contributed to costs of FISH testing only). Reprints: Ross C. Smith, FRACS, University of Sydney, Department of Surgery, Royal North Shore Hospital, St. Leonards, New South Wales, Australia, NSW 2065 (e-mail: rsmith@med.usyd.edu.au). Copyright Ó 2005 by Lippincott Williams & Wilkins Am J Surg Pathol  Volume 29, Number 9, September 2005 1125