TO THE EDITORS:We read with great interest the article by Freitas and colleagues, 1 who reported the important role of Janus kinase 2 (JAK2)/signal transducer and activator of transcription 1 signaling in the pathophysiology of hepatic ischemia/reperfusion injury (IRI). However, there are some methodological concerns about the study that do not allow firm conclusions to be drawn about the data. Using a standard model of warm hepatic IRI, the authors demonstrated that C57BL mice, treated with the selective JAK2 inhibitor AG490, showed reduced hepatocellular injury and decreased macrophage and neutrophil infiltration.The authors used mouse hepatoma cells in some of their in vitro experiments. They showed that AG490 decreased lipopolysaccharide (LPS)-mediated mouse hepatoma cell apoptosis. However, previous studies have shown that human hepatoma cells do not accurately reflect the biological responses of primary human hepatocytes.2 This presents 2 problems for the interpretation of the study. First, there are no data within the study showing that mouse hepatoma cells respond to LPS in a way similar to that of primary mouse hepatocytes. Previous studies have also shown that mouse hepatoma cells differ phenotypically from primary mouse hepatocytes.3 Moreover, hepatoma cells differ functionally from primary hepatocyte cell cultures.2,4 Second, the authors did not use a suitable in vitro model of warm IRI to adequately demonstrate the effects of AG490. This is surprising because of the number of previously published studies that have used such models. 5,6 Indeed, previous work has illustrated the different responses of the human hepatocyte cell line HL-7702 to the tumor necrosis factor superfamily ligand and tumor necrosis factor-related apoptosis-inducing ligand, during hypoxia and with hypoxia/reoxygenation. 5 Therefore, the effects of JAK2 inhibition should be investigated during hypoxia and hypoxia/reoxygenation because cells will be exposed to oxygen tensions akin to those experienced in vivo. Experiments conducted with this type of in vitro model are more relevant to hepatic IRI than investigations of the effects of LPS during normoxia. This model would generate the accumulation of intracellular reactive oxygen species, which are crucial to part of hepatic IRI.7 Certainly, if JAK2 were to inhibit cellular apoptosis in this in vitro model of warm hepatic IRI, a stronger, more convincing argument could be made about the potential therapeutic benefit of AG490.We believe that the in vivo data of this study by Freitas et al.1 do show some clinical promise with respect to the amelioration of hepatic IRI by JAK2 inhibition. However, the in vitro data do not provide a clear explanation of the effects of JAK2 inhibition on hepatocyte cell death. We encourage the use of primary mouse hepatocytes in a suitable in vitro model of warm IRI to further delineate the effects of JAK2 on hepatocyte function. At present, no clear conclusion can be drawn about the benefits of JAK2 inhibition during hepatic IRI.