Vascular invasion is underrecognized in colorectal cancer using conventional hematoxylin and eosin staining

Purpose: This study was designed to test the hypothesis that highlighting vascular spaces with histochemical or immunohistochemical stains facilitates the identification of extramural and intramural vascular invasion in resected colorectal cancer specimens compared with routine hematoxylin and eosin staining. Methods: Archival tumor sections from 50 resected colorectal cancers, in which extramural vascular invasion was not seen within the original tissue sections, were stained with hematoxylin and eosin, elastic van gieson histochemistry, and immunohistochemistry for CD31 and CD34. Two observers assessed the stained sections and the agreed incidence of vascular invasion using the four staining methods was compared. Results: Vascular invasion was more commonly identified in Dukes C (pTanyN1/2) (vascular invasion seen in 24 of 25 cases by at least 1 method) than Dukes B tumors (pT3/4N0) (vascular invasion seen in 14 of 25 cases by at least 1 method). Vascular invasion was identified in significantly more cases using elastic van gieson (24 cases; P=0.0001), CD31 (18 cases; P=0.0064), and CD34 (21 cases; P < 0.0001) than with hematoxylin and eosin alone (5 cases). Conclusions: This study was novel in that it compared both histochemical and immunohistochemical methods for identifying vascular invasion in cases of colorectal cancer in which vascular invasion had not been identified during initial reporting. Highlighting of endothelium significantly increases the observed incidence of vascular invasion in colorectal cancer compared with hematoxylin and eosin alone. Elastic van gieson seemed sensitive for the presence of vascular invasion but with uncertain specificity. The possibility that these immunohistochemical methods may identify a subset of patients with colorectal cancer who may benefit from chemotherapy warrants further study.