Relation of microvessel density with microvascular invasion, metastasis and prognosis in renal cell carcinoma

To clarify the significance of microvessel density (MVD) in a retrospective investigation the relationship between the pattern of MVD (reflecting angiogenesis), and tumour stage, grade, size, and occurrence of microvessel invasion (MVI), metastasis, and cancer-specific survival (CSS) in patients who had surgery for renal cell carcinoma (RCC). Vessels were labelled in sections of formalin-fixed, paraffin-embedded tissues from 54 RCCs by CD34 immunohistochemistry. The mean MVD, expressed as the number of vessels per 10 high-power fields (HPF, x400) were measured for each case. In addition, all pathological slides were reviewed for the presence and absence of MVI. The prognostic value of MVD and MVI was then evaluated, and correlated with the usual prognostic variables, tumour metastasis and CSS. In a univariate analysis of CSS, the MDV tended to be lower as stage increased from pT1 to pT3, and as grade increased from G1 to G4, although it was statistically significant only for stage (P < 0.001 and 0.050, respectively). The mean MVD was higher in 42 nonmetastatic than in 12 metastatic tumours, and in 33 tumours associated with MVI than in 21 with no MVI (P < 0.001). The mean MVD was also lower and significantly different for 28 large than 26 small tumours (P = 0.005). The survival rate of patients with tumours that were small, low-stage, of higher MVD, with no MVI and metastasis was significantly higher than that of patients with large, high-stage, low MVD, with MVI and metastatic tumours (all P < 0.001). MVI was significantly more common with a decreasing trend in MVD and the presence of metastasis (Spearman rank correlation r(s) = -0.68, P = 0.01, and r(s) = 0.39, P = 0.01, respectively). Independent prognostic factors in a multivariate analysis were: in all patients with RCC, tumour stage (P = 0.013) and metastasis (P = 0.028); in those with low MVD, MVI (P = 0.004) and metastases (P = 0.016); in those with no MVI, stage (P = 0.020); in those with MVI, MVD (P = 0.001); in those with no metastases, stage (P = 0.045); and in those with metastases, MVD (P < 0.001). No independent predictor was identified in patients with high MVD. In patients with no metastases there was a significantly shorter median CSS time in RCCs with low MVD and with MVI (P = 0.004 for both). Similarly, patients who had grade 3-4 tumours, vs those with lower MVD and with MVI, had a significantly shorter median CSS (P = 0.020 for MVD, and 0.01 for MVI). This study suggested that MVD in RCC was inversely associated with MVI, tumour metastasis, patient survival and tumour diameter and stage, from the usual prognostic variables, but MVD was not an independent prognostic factor in multivariate analysis for all patients with RCC. Low MVD and the presence of MVI appears to be a marker for identifying patients with an adverse prognosis.