Background: Prior studies have suggested that microvessel density is an important prognostic factor in invasive breast cancer. However, the extent and distribution of microvessels in association with ductal carcinoma in situ (DCIS) have not been well defined. Purpose: Our goal was to determine the density and distribution of stromal microvessels in DCIS and to investigate the relationships among microvessel density, histopathologic features, HER2/neu oncogene expression, and tumor proliferation rate. Methods: Of 61 consecutive cases of DCIS identified from hospital pathology reports, 55 cases were evaluated. Breast biopsy specimens had been preserved in paraffin blocks for each DCIS case. Histologic sections of formalin-fixed, paraffin-embedded tissue were stained with hematoxylin-eosin and immunostained for factor VIII-related antigen, the HER2/neu oncoprotein, and the proliferative-associated antigen detected by the Ki-S1 antibody. Factor VIII-stained sections from each case were independently examined by two pathologists and overall tumor-associated stromal microvessel density was scored semiquantitatively on a 1+ to 3+ scale by each observer. Quantitative microvessel counts of DCIS-associated stromal microvessel density were performed. The presence or absence of a cuff of microvessels in immediate apposition to the basement membrane of involved spaces was also evaluated. Results: A variable number of microvessels were found to be present in a diffuse pattern surrounding spaces involved with DCIS. Semiquantitative microvessel scores were 2+ in the majority of cases (53%); 22% of cases were 1+, and 25% were 3+. Quantitative microvessel counts ranged from 17 to 80 vessels per 100x field (0.45 mm(2)), with a mean +/- SD of 42.9 +/- 16.6. Comedo-type lesions were significantly (P = .004) more often associated with 3+ microvessel density than non-comedo-type lesions by semiquantitative assessment. As determined by both semiquantitative and quantitative analysis, respectively, the presence of prominent microvessel density was significantly associated with marked stromal desmoplasia (P = .05 and P = .04), HER2/neu expression (P = .03 and P = .0002), and high Ki-S1 proliferation index (P = .05 and P = .01). Vascular cuffing around involved spaces was identified in 21 of the 55 cases (38%) and was not significantly associated with histologic features, HER2/neu expression, or Ki-S1 proliferation index. Conclusions: DCIS of the breast is characterized by two patterns of stromal microvessels. The first pattern is a diffuse increase in stromal microvessels surrounding involved spaces. This pattern is particularly prominent in comedo-type lesions with marked stromal desmoplasia. The second pattern is microvessel cuffing of involved spaces that is present in only a minority of cases and appears unrelated to histologic features evaluated, including DCIS subtype.
