1. ¹Department of Surgery II, School of Medicine, Tokyo Women's Medical University, 8-1 Kawadacho, Shinjuku-ku, Tokyo 162-8666, Japan
2. ²Cancer Research UK Tumor Pathology Group, Nuffield Department of Clinical Laboratory Sciences, University of Oxford, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
3. ³MRC Human Immunology Unit, Weatherall Institute of Molecular Medicine, John Radcliffe Hospital, Headington, Oxford OX3 9DS, UK
4. ⁴Department of Surgical Pathology, School of Medicine, Tokyo Women's Medical University, 8-1 Kawadacho, Shinjuku-ku, Tokyo 162-8666, Japan
5. ⁵Department of Pathology, School of Medicine, Tokyo Women's Medical University, 8-1 Kawadacho, Shinjuku-ku, Tokyo 162-8666, Japan
6. ⁶Cancer Research UK Molecular Oncology Laboratory, Institute of Molecular Medicine, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK

Correspondence: Dr T Kato, Onuma Hospital, 2-10-2, Higashimizumoto, Katsushika-ku, Tokyo 125-0033, Japan. E-mail: t-kato@bd5.so-net.ne.jp

Received 20 July 2005; Revised 8 September 2005; Accepted 28 September 2005; Published online 25 October 2005.

Abstract

This study was undertaken to determine the highly sensitive method for detecting tumour lymphatic vessels in all the fields of each slide (LV), lymphatic microvessel density (LMVD) and lymphatic vessel invasion (LVI) and to compare them with other prognostic parameters using immunohistochemical staining with polyclonal (PCAB) and monoclonal antibodies (MCAB) to the lymphatic vessel endothelial hyaluronan receptor-1 (LYVE-1), and the pan-endothelial marker factorVIII in a series of 67 human breast cancers. In all LYVE-1-stained sections, LV (some of which contained red blood cells) were frequently found localised in extralobular stroma, dermis, connective tissue stroma and adjacent to artery and vein, but were rare within the intralobular stroma or the tumour body (3/67 cases) or areas of widespread invasion. In contrast small blood vessels were observed in intra- and extralobular stroma in the factor VIII-stained sections. Quantitation of vessel numbers revealed that LYVE-1/PCAB detected a significantly larger number of LV than either H&E or LYVE-1/MCAB (P<0.0001). LYVE-1/PCAB detected LVI in 25/67 cases (37.3%) and their presence was significantly associated with both lymph node metastasis (²=4.698, P=0.0248) and unfavourable overall survival (OS) (P=0.0453), while not relapse- free survival (RFS) (P=0.2948). LMVD had no influence for RFS and OS (P=0.4879, P=0.1463, respectively). Our study demonstrates that immunohistochemistry with LYVE-1/PCAB is a highly sensitive method for detecting tumour LV/LVI in breast cancer and LVI is a useful prognostic indicator for lymphatic tumour dissemination.