[CANCER RESEARCH 57. 5505-5508. December 15. 1997] Mitosin (a New Proliferation Marker) Correlates with Clinical Outcome in Node-negative Breast Cancer1 Gary M. Clark,2 D. Craig Allred, Susan G. Hilsenbeck, Gary C. Chamness, C. Kent Osborne, Diane Jones, and Wen-Hwa Lee Department of Medicine, Division of Medical Oncology [G. M. C.. S. G. H.. G. C. C.. C. K. OJ. Department of Pathology /D. C. A.], and Department of Molecular Medicine, Institute of Biotechnology ¡D.J., W.-H. L], University of Texas Health Science Center, San Antonio, Texas 78284 ABSTRACT Tumor proliferation rate is an important prognostic factor in breast cancer, and S-phase fraction (SPF), as measured by flow cytometry, is the most clinically validated of several methods for measuring it. However, flow cytometry is not well suited to evaluating the formalin-fixed, paraf fin-embedded tumors that are routinely available or to the increasing number of small breast cancers. These and other limitations have moti vated research into alternative methods for measuring proliferation, in cluding immunohistochemistry (IHC) against cell cycle-related antigens, which are better suited for the evaluation of small archival tissue samples. Mitosin is a recently described 350 kl) nuclear phosphoprotein that is expressed in the late G,, S, G2, and M phases of the cell cycle but not in G0. Using a new monoclonal antibody (14C10), this pilot study evaluated mitosin expression by IHC in a series of 386 node-negative, formalin-fixed, archival breast cancers and correlated the results with several prognostic factors and clinical outcome (median follow-up, 78 months; range 3-214 months). The median and range of mitosin positive cells were 7% and 1-47%, respectively. There was a strong positive correlation between mitosin and SPF (r = 0.57; /' = 0.0001), and there were significant negative correlations with estrogen receptor, progesterone receptor, and patient age. Mitosin was not related to overall survival in this pilot study. However, in a univariate cutpoint analysis of disease-free survival (DFS), patients with high levels of mitosin (>')'< positive cells) had significantly worse DFS than did patients with lower levels (68% versus 84% at 5 years, respectively). In a multivariate analysis of DFS, large tumor size (>2 cm) and high mitosin were the only independently significant predictors of recurrence (relative risks = 2.47 and 1.72, respectively) in a model con taining the additional factors estrogen receptor, progesterone receptor, patient age, and SPF. These preliminary results suggest that mitosin as assessed by IHC may be superior to SPF as a prognostic factor in node-negative breast cancer, but additional studies are necessary to vali date these promising findings. INTRODUCTION Alterations in the control of cell proliferation are important in the development and progression of many types of human cancer, and the overall rate of proliferation is indirectly related to clinical outcome. There are several techniques for estimating or measuring proliferation, including light microscopic mitotic index, thymidine labeling index, bromodeoxyuridine uptake, and DNA flow cytometry. Each of these methods has strengths and weaknesses, but none are ideal for meas uring proliferation in formalin-fixed, paraffin-embedded tissue, which has become the standard method of processing and storing clinical Received 2/19/97; accepted 10/14/97. 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. 1This work was supported by NIH Grants POI CA30195. P50 CA58I83. and P30 CA54174. 2 To whom requests for reprints should be addressed, at Department of Medicine. Division of Medical Oncology. University of Texas Health Science Center. 7703 Floyd Curl Drive. San Antonio. TX 78284-7884. Phone: (210) 567-4749: Fax: (210) 567-6687: E-mail: gary@oncology.uthscsa.edu. samples. IHC3 against cell cycle-related antigens is a more recent method for measuring proliferation that is better suited for evaluating routine archival tissue, and antibodies have been developed to a few antigens, such as KÃOE67, which have been used in preliminary studies with varying success. The ultimate success of IHC will depend on how specific the antigens are for dividing cells and how sensitive the antibodies against them are in archival tissue. So far, none of them is optimal. Mitosin is a recently identified 350 kD nuclear phosphoprotein that is involved in cell division and, therefore, may be a suitable target for evaluating proliferation by IHC. It was identified through its binding to purified retinoblastoma protein, and it is expressed in the late G,, S, G2, and M phases of the cell cycle but is absent in G0 (1). It associates with the centromere, spindle, and midbody of the mitotic apparatus during M and completely degrades following cytokinesis (1). The carboxy terminus of the molecule has been shown to be essential for the localization of mitosin in the nucleus (2). We have developed several monoclonal antibodies to mitosin that are suitable for IHC in archival tissue. This pilot study used one of these antibodies (14C10) to measure mitosin expression by IHC in 386 archival node-negative human breast cancers. The results were compared to SPF, as determined by flow cytometry, several other clinical/pathological features, and patient outcome. MATERIALS AND METHODS Antibody Production and Characteristics. Balb/c mice were immunized with bacterially expressed fusion protein GST lOBgl (containing amino acids 1759-2093 of mitosin). Injections of 100 /*g of protein were given s.c., the first in Complete Freund's Adjuvant, followed by three boosts in Incomplete Freund's Adjuvant at 4-6-week intervals. A final boost of GST lOBgl in sterile PBS was given i.v. 3 days prior to the fusion. Isolated immune splenocytes were fused with NSI mouse myeloma cells using 50% polyethyleneglycol essentially as described by Oi and Herzenberg (3). Successful hybridomas were initially screened by ELISA against GST and GST lOBgl. Several clones that only reacted to GST lOBgl were expanded, single-cell cloned, and retested by ELISA to confirm specificity. Five clones (2G8, 13E7, 14C10, 16G6, and 18E1) with strong GST lOBgl activity were then injected into the peritoneum of Balb/c mice to form ascites. IgG was purified from the ascites using protein G Sepharose, dialyzed against PBS. and stored long-term at -80°C. In a previous study using normal mouse kidney CV1 cells synchronized by different drugs to obtain uniform populations at different phases of the cell cycle, mitosin was shown to be expressed in late G,, S, G2, and M but not in G0 (1). The specificity of the mitosin antibody used in this study was con firmed by immunoprecipitation against whole-cell lysates from the normal human breast epithelial cell line HBL-IOO. To obtain sufficient quantity of cells expressing mitosin, they were first arrested and synchronized at the G,-S boundary for 24 h with hydroxyurea and released for 4 h prior to harvesting. The lysates were precipitated with each of the five purified antibodies de scribed above, the products were separated on a polyacrylamide gel. Western blotted, and probed with antibody 14C10. Each antibody showed single-band 1The abbreviations used are: IHC, immunohistochemistry: SPF. S-phase fraction: GST. glutathione 5-transferase; ER. estrogen receptor; PgR. progesterone receptor; DFS. disease-free survival; OS. overall survival. 5505 Research. on December 3, 2015. © 1997 American Association for Cancer cancerres.aacrjournals.org Downloaded from