Advances in Brief Predictive Value of the EphA2 Receptor Tyrosine Kinase in Lung Cancer Recurrence and Survival

Purpose: Underestimation of disease severity is a major problem confronting the successful clinical management of non-small cell lung cancer. Recent advances in molecular biological substaging may provide an opportunity to identify those patients with the most aggressive forms of the disease, but there is a continuing need for accurate markers of disease relapse and survival. Experimental Design: In our present study, immunohistochemical analyses of a retrospective database of pathologic specimens were used to demonstrate that the EphA2 receptor kinase is frequently overexpressed in NSCLC. Results: Initial presentation with high levels of EphA2 predicts subsequent survival, overall relapse, and site of relapse. Specifically, high levels of EphA2 in the primary tumor predict brain metastases, whereas low levels of EphA2 relate to disease-free survival or contralateral lung metastasis. Conclusions: These data suggest that EphA2 may provide a molecular marker to identify and predict patients who have isolated brain metastases. Moreover, the high levels of EphA2 in lung cancer may provide an opportunity for therapeutic targeting. Introduction NSCLC is the leading cause of cancer mortality in the United States (1). Approximately 177,000 cases of new lung cancer will be diagnosed this year and 160,000 people will die of the disease. The current standard of care for clinical stage I (pathologic stage I, II, and IIIA) NSCLC is surgical resection of the primary tumor followed by observation (2). However, patients with clinical stage I NSCLC who have undergone a resection have an estimated 50% recurrence and death rate within 5 years, representing one of the poorest outcomes for stage I disease among all tumor types (3–5). One of the main contributing factors to such a poor survival rate is that a significant number of patients are understaged at the time of resection (4, 6). There is an urgent need for a detection method that is more sensitive and reliable than current approaches (4, 7). In particular, many of the most dangerous forms of lung cancer arise when a population of tumor cells gains the ability to colonize the brain (8, 9). Brain metastases comprise greater than one-quarter of all recurrences in patients with resected NSCLC and are a major source of lung cancer morbidity and mortality. When symptomatic, the median survival of patients with brain lesions is 4 months (5, 6, 10). Thus, it is critical to identify and treat those patients who are at risk for brain metastases before the onset of symptoms. Recent investigation has sought to identify factors that predict poor prognosis. Unfortunately, the use of conventional histopathological variables (performance status, subtype, size, differentiation, and mitotic rate) to construct a risk model has been limited by low prevalence and the discontinuous nature of individual variables (4, 11, 12). These limitations have underscored recent efforts to ask if molecular markers might improve risk stratification (4). A series of genetic and epigenetic changes facilitate the transition from normal bronchial epithelium to carcinoma and examples include mutation or overexpression of Ras oncogenes (K-Ras, H-Ras, and N-Ras), receptor tyrosine kinases (epidermal growth factor receptor and HER2) and cell cycle proteins (retinoblastoma and p53; Ref. 4). However, none of these markers has been linked with the recurrence of brain metastases. Our laboratories have sought to identify the causes of metastatic cell behavior, with emphasis upon the identification of markers that predict metastatic recurrence in the clinical setting. Our most recent studies have focused upon the EphA2 receptor tyrosine kinase, which is overexpressed and functionally altered in a variety of different solid tumor types (13–18). EphA2 has been linked to the regulation of cellular behaviors that promote a metastatic phenotype and appears to provide an exciting new opportunity for therapeutic targeting (13, 15). In this study, we demonstrate that EphA2 provides a new and independent marker of lung cancer recurrence and survival. Importantly, high levels of EphA2 identify those patients that are at risk of lung cancer metastasis to the brain. Thus, EphA2 provides a new and independent marker of aggressive behavior with potential clinical utility in cancer diagnosis and treatment. Materials and Methods Database and Staining Procedures. Formalin-fixed, paraffin-embedded blocks of pathologic specimens were used from a retrospective database consisting of patients who had undergone a complete resection for NSCLC as detailed previously (4). Histological examination of the specimens was performed by board-certified pathologists at the Duke University Medical Center at the time of resection. Clinical grading of the patients Received 7/15/02; revised 10/7/02; accepted 10/11/02. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 To whom requests for reprints should be addressed, at MedImmune, Inc., 35 West Watkins Mill Road, Gaithersburg, MD 20878. Phone: (240) 632-4639; E-mail: [email protected]. 2 The abbreviation used is: NSCLC, non-small cell lung cancer. 613 Vol. 9, 613–618, February 2003 Clinical Cancer Research Research. on April 20, 2017. © 2003 American Association for Cancer clincancerres.aacrjournals.org Downloaded from was performed by Dr. Harpole, a board certified thoracic surgeon. Sections (4–6 M) sections were obtained from each tissue block by microtome sectioning and deparaffinized as detailed previously (4). The samples were then stained with EphA2-specific polyclonal (catalogue no. SC-924; 1:100; Santa Cruz Biotechnology, Santa Cruz CA) or monoclonal antibodies (catalogue no. 05-480; 1:500; Upstate Biotechnology, Inc., Lake Placid, NY) as detailed previously (16). Powerblock Universal Blocking Reagent (catalogue no. HK085-5K; BioGenex Laboratories, San Ramon, CA) was used to block nonspecific binding. The slides were developed using diaminobenzidine as the chromogen and counterstained with hematoxylin. Known positive tumors and normal lung tissue were used as positive and negative controls, respectively. Data Interpretation and Analysis. All slides were read independently by two experienced observers who were blinded as to the tissue source. Differences in immunohistochemical scores were rare and resolved by consensus. There were no instances of one-third discordance (2 point differences). Each sample was scored using a 0–3 scale, with 0 denoting no staining. In general, the tumor cells were found to stain uniformly across the sample, thus the scoring system did not consider the fraction of tumor cells with positive staining. All statistical analyses were performed using statistical software (Microsoft Excel). A two-tailed, homoscedastic Student’s t test was used for comparisons among primary tumor samples. A two-tailed, paired analysis was used to compare EphA2 immunoreactivity in matched pairs of primary tumor and brain metastases. All analyses defined P 0.05 as significant.