Clinicopathologic Significance of Cathepsin B and Urokinase-type Plasminogen Activator Immunostaining in Colorectal Adenocarcinoma

Acetone-fixed cryostat sections of 52 colorectal adenocarcinomas (26 stage B, 26 stage C) were immunostained with antibodies for two “invasion-associated” proteolytic enzymes—cathepsin B and urokinase-type plasminogen activator. Cathepsin B immu nostaining of neoplastic cells was observed in 21 tumors (40%). It was generally accompanied by staining among peritumoral host cells. There was no correlation be tween neoplastic cell cathepsin B staining and tumor stage (stage B, 42% positive vs stage C, 38% positive) however staining was more frequent in tumors that recurred (cathepsin B negative, 23% recurred vs cathepsin B positive, 48% recurred, P = .05-.06,25 months follow-up). Significant neoplastic cell urokinase-type plasminogen activator staining was present in only six cases (12%); however, peritumoral spindle and inflammatory cells exhibited positivity in 35% of tumors. There was no correlation between host cell derived urokinase-type plasminogen activator staining and either node metastasis or patient outcome. Tumors that stained for both neoplastic cell ca thepsin B and stromal cell urokinase-type plasminogen activator (12%) characterized a morphologically and clinically aggressive subset, compared to cases that stained for only one (51%) or neither (37%) enzyme (both positive, 83% recurred, 83% poorly differentiated vs neither positive, 26% recurred, 5% poorly differentiated). These data imply that invasion-associated proteases are derived from heterogeneous cellular sources in colorectal tumors. Further, synergistic protease activity may promote aggres sive clinical behavior accounting, in part, for the adverse prognostic significance of poor differentiation. Int J Surg Pathol 1(4):227-234, 1994