Assessment of the optimal therapeutic range of lepirudin in HIT patients with thrombosis.

Risk ratios (RR) and 95% confidence intervals for the combined endpoint of new TEC, limb amputation, and death (black bars and solid line) and for bleeding complications (white bars and dotted line) are given for the lepirudin-treated patients according to aPTT classes in comparison with the historical control population. In this Cox proportional hazards model, the aPTT ratio was considered a time-dependent covariate. Medium aPTT ratios (1.5-2.5) were associated with a clinically and statistically significant reduction in the incidence of the combined endpoint compared with the historical control group (RR = 0.42; 95% CI, 0.22-0.80; P = .009). Low aPTT ratios (1.0-1.5) were associated with an insignificant effect (RR = 0.86; 95% CI, 0.38-1.94; P = .72), whereas high aPTT ratios (greater than 2.5) demonstrated no additional improvement compared with medium aPTT ratios (RR = 0.70; 95% CI, 0.21-2.32; P = .56). However, there was a marked increase in bleedings with high aPTT ratios (RR = 6.03; 95% CI, 2.34-15.54; P = .0002). We, therefore, consider aPTT ratios of 1.5 to 2.5 as the therapeutic window in patients with HIT and acute thrombosis.