The contribution of ICAM-1 to neutrophil-SEC interaction is under controversy. In a recent issue, we reported that ICAM-1 plays a crucial role in neutrophil adhesion to and migration through the SEC monolayer in vitro (1). Jaeschke et al. (2,3) previously reported that anti-ICAM-1 had no significant effect on hepatic accumulation of neutrophils in endotoxic liver injury, and speculated that neutrophil accumulation is caused by SEC and Kupffer cell swelling and the stiffness of neutrophil cell membrane when exposed to endotoxin. In this letter, they stated that the inhibition of ICAM-1 had only a limited effect in our in vitro study, and that the data obtained under our experimental conditions could not be applied to the behavior of neutrophils in hepatic sinusoid in vivo. However, in contrast to their observation, a few experiments in vivo showed that ICAM-1 was involved in liver injury during endotoxemia (4,5), which is consistent with our findings. We can propose two reasons why their data in vivo differs from ours in vitro. First, they investigated the role of ICAM-1 in hepatic infiltration of neutrophils only 90 min after LPS exposure in mice. As we stated in the issue, this time frame would be too early for them to estimate the effect of anti-ICAM-1 antibody on neutrophil-SEC interaction because overexpression of ICAM-1 on SEC was not observed until 6 to 8 h after LPS exposure in vivo (6,7). In fact, Xu et al. (4) reported that hepatic infiltration of neutrophils was significantly inhibited 24 h after LPS injection in ICAM-1-deficient mice, but was not inhibited 2 h afterward. This in vivo data clearly showed that ICAM-1 was involved in neutrophil accumulation in ICAM-1-overexressing liver. We can agree with the data of Jaeschke that hepatic neutrophil accumulation was not mediated by ICAM-1 90 min after LPS exposure when ICAM-1 was not enough upregulated on SEC. However, our experiments suggest that ICAM-1 plays an important role in hepatic infiltration in a later stage, when the expression of ICAM-1 was enhanced on SEC. The next reason is that they estimated neutrophil accumulation by histochemical analysis. To investigate neutrophil-SEC interaction, we used an in vitro flow system (IVFS) (1,8), which allows of the dynamic analysis of cell-to-cell interaction (9). IVFS enables us to investigate the each step of infiltration and to distinguish adhered cells from floating or rolling cells on SEC monolayers. In other words, we can count the actually adhered or migrated cells by using our device, but by means of their methods, it is impossible to distinguish strictly neutrophil adhesion from leukostasis, which was secondarily induced by circulatory disturbance. Vollmar et al. (10) reported that the number of nonperfused sinusoids correlated with the number of neutrophils in sinusoids during endotoxemia. Indeed, our in vitro data showed that once a neutrophil adhered to SEC, the next adhesive reaction occurred around the adhered neutrophil at first. This finding may reveal that the adhered cells change the flow rate and make turbulent flow, which enhances additional adhesive reaction (preliminary data). Besides, neutrophil-SEC interaction mediated by ICAM-1 induces SEC injury by activated neutrophils (8). We do not deny the possibility that neutrophil accumulation is caused by passive trapping due to SEC swelling or active vasoconstriction; however, our findings revealed that anti-ICAM-1, in part but at least significantly, inhibited an initial adhesive reaction that induces the secondary leukostasis and SEC injury in hepatic sinusoids in vivo. Considering these points, even if anti-ICAM-1 inhibits the adhesion of only 50%–60% of neutrophils, we can conclude that ICAM-1 plays a major role in the hepatic infiltration of neutrophils during endotoxic liver injury. Dr. Takeshi Okanoue Dr. Shinichi Sakamoto