Histologic criteria and pitfalls in the diagnosis of lymphovascular invasion in radical prostatectomy specimens

Lymphovascular invasion is a known independent prognostic factor in prostate cancer. The objective of this study is to describe reliable morphologic features for identification of lymphovascular invasion in prostatectomy specimens and avoid misinterpretation of its mimickers. A total of 364 foci of lymphovascular invasion were analyzed in 264 slides from 170 prostatectomies. The average tumor volume was 25.5%. Tumor emboli were seen inside the tumor (8%), at the front edge of the tumor (30%), separated from the tumor (32%), and distant from the tumor (30%). Tumor emboli were more frequent per case and more often in an extraprostatic location in lymph node-positive cases (P<0.05). One hundred thirty-four emboli were in a single thin-walled vessel, 227 were in a thin-walled vessel next to an artery, and 3 were seen inside an artery. Twenty-eight tumor emboli were attached to a vessel wall, 18 had proteinaceous material in the vascular lumen, and 14 were surrounded by erythrocytes. The following mimickers were seen: retraction artifact and perineural invasion-all cases; cancer impinging upon vascular space-45 foci; tangential sections of endothelium-10 foci; displacement of benign and collapsed malignant glands-16 and 27 foci, respectively; retraction with erythrocytes-3 cases; intravascular degenerating tumor cells-3 foci; malignant glands in atrophic ducts-4 foci; and myofibroblastic proliferation in thrombosed vessels-2 foci. In 50 stained blocks, CD31 and D2-40 immunostaining studies confirmed all lymphovascular invasions diagnosed by hematoxylin and eosin staining and demonstrated emboli in 47 lymphatic and 16 blood vessels. In summary, the current study identifies features of true lymphovascular invasion and how to distinguish them from mimickers on routine hematoxylin and eosin sections.