The study aimed to evaluate the prognostic value of angiogenesis by vascula
r grading of primary breast tumours, and to evaluate the prognostic impact
of adding the vascular grade to the Nottingham Prognostic Index (NPI). The
investigation included 836 patients. The median follow-up time was 11 years
and 4 months. The microvessels were immunohistochemically stained by antib
odies against CD34. Angiogenesis was graded semiquantitatively by subjectiv
e scoring into three groups according to the expected number of microvessel
s in the most vascular tumour area. The vascular grading between observers
was moderately reproduced (kappa = 0.59). Vascular grade was significantly
associated with axillary node involvement, tumour size, malignancy grade, o
estrogen receptor status and histological type. In univariate analyses vasc
ular grade significantly predicted recurrence free survival and overall sur
vival for all patients (P < 0.0001), node-negative patients (P < 0.0001) an
d node-positive patients (P < 0.0001). Cox multivariate regression analysis
showed that vascular grading contributed with independent prognostic value
in all patients (P < 0.0001). A prognostic index including the vascular gr
ade had clinical impact for 24% of the patients, who had a shift in prognos
tic group, as compared to NPI, and implied a better prognostic disseminatio
n. We concluded that the angiogenesis determined by vascular grading has in
dependent prognostic value of clinical relevance for patients with breast c
ancer. (C) 2000 Cancer Research Campaign.