Fig. 3. Effect of different colloid-treatments on neutrophil (PMN) adhesion to cytokine-activated endothelium. Firm adhesion is presented as ratios (colloid/control) and upper 95% confidence intervals, calculated from means of six experiments (absolute mean of cytokine-activated control, 460 PMNs/mm²; 95% confidence interval, ±35). Starting 2 h before addition of tumor necrosis factor α (0.5 ng/ml), colloid-pretreated endothelial cells were incubated with the indicated colloids at 10 mg/ml throughout the complete period of cytokine activation (4 h). In the group containing colloid-pretreated endothelium and untreated PMNs (  black columns ), only the endothelial cell culture medium was supplemented with the indicated colloids. In the group containing colloid-pretreated endothelium and cotreated PMNs (  gray columns ), the colloids were also added to the PMN suspension. The cytokine-activated controls, necessary to calculate the ratios (colloid/control), were activated with tumor necrosis factor α in medium alone. Firm adhesion was quantified from means of five different fields of view for each experimental condition as detailed in the Materials and Methods. Effects of colloid treatment were analyzed by an analysis of covariance ( *P < 0.05  vs. control,  post hoc t test). Colloid preparations of different molecular weights are presented with their molecular weight in kilodaltons (  e.g. , hydroxyethyl starch as HES 200).