Influence of investigator experience and microscopic field size on microvessel density in node-negative breast...

In this study on the determination of intratumoral microvessel density (MVD) in breast cancer, we have investigated the influence of the observer experience and the microscopic field size. We have used the sample set reported on earlier in the J Natl Cancer Inst 87: 1797-1798, 1995. This case-control study has shown a positive association of high MVD and unfavorable outcome when comparing node-negative pT1-2 breast carcinoma (NNBC) patients with a disease-free period of over ten years with those with an early distant relapse. Tumor sections of both outcome groups (favorable: n = 19; unfavorable: n = 19) were immunostained for factor VIII related-antigen (FVIII r-Ag). Microvessels were counted in the areas of most intense vascularization ('hot spots'), both at magnification x 200 (field size of 0.61 square mm) and x 400 (field size of 0.15 square mm), by one inexperienced and three experienced observers. Microphotographs of individual vascular hot spots were analyzed using overlays resembling the two field sizes. The main results obtained are: i) a confirmation of the prognostic value of microvessel density in the case-control sample set (n = 38) was established by all experienced but not by the unexperienced investigator; ii) both at x 200 and x 400 magnification, angiogenesis quantification in vascular hot spots contained prognostic information. The results of this study indicate that the selection of vascular hot spots in tumor sections immunostained for an antigen expressed on endothelial cells is more prone to inter-observer variability and more dependent on training than the counting of the microvessels within predefined hot spots itself. The microscopic magnification and resulting field size do not influence the prognostic significance of MVD in NNBC. This information validates the development of more objective methods of measuring the amount of angiogenesis within malignant tissue. This will allow more accurate implementation of the angiogenesis parameter in multiparametric and prospective prognostic factor studies in NNBC.