Combined Treatment with Ruxolitinib and Dexamethasone Curtails Inflammation and Lessens Disease in Preclinical Studies of Hemophagocytic Lymphohistiocytosis

The Hemophagocytic lymphohistiocytoses (HLH) comprise a heterogeneous group of disorders characterized by abnormal and severely damaging immune responses. Patients with HLH experience episodes of inflammation associated with the hyper-activation of CD8+ T cells and macrophages, which overproduce pro-inflammatory cytokines including interferon (IFN)-γ, interleukins (IL)-12 and -18, -6 and Tumor Necrosis Factor-α. Current treatments include immunosuppressive medications (corticosteroids, anti-thymocyte globulin) with or without cytotoxic chemotherapeutic agents (etoposide). Despite the use of these drugs, >50% of HLH patients die from unbridled inflammation. To develop more effective treatments for HLH, our laboratory has focused on blocking the effects of cytokines, which drive inflammation and mediate morbidity and mortality in HLH. We recently demonstrated that targeting the Janus Kinases (JAK1/2) with the JAK inhibitor ruxolitinib (INCB018424) significantly ameliorates the clinical and laboratory manifestations of HLH in preclinical mouse models (R. Das et al., Blood, 2016). Despite its beneficial effects, ruxolitinib did not completely abrogate disease. We therefore sought to understand its mechanisms of action and further improve upon these results. In mice, IFN-γ is central to disease pathogenesis, as its neutralization reduces anemia and prolongs survival. To examine whether ruxolitinib might be acting primarily through the inhibition of this cytokine, we modeled HLH in perforin-deficient (Prf1^-/-) mice that were infected with lymphocytic choriomeningitis virus (LCMV). Beginning on day 4 post infection (p.i.), mice were treated or not with ruxolitinib (90 mg/kg by oral gavage twice daily) or with an IFN-γ blocking antibody (0.5 mg intraperitoneally [i.p.] on days 4 and 7). On day 9 p.i., the end point for these studies, mice were examined for the manifestations of HLH. Compared to untreated LCMV-infected mice, which lost on average 16% of starting body weight, mice treated with ruxolitinib or IFN-γ blockade exhibited only 7% weight loss. Ruxolitinib and IFN-γ blockade also ameliorated LCMV-induced anemia and thrombocytopenia, with platelet counts with ruxolitinib treatment reaching levels comparable to those observed in uninfected animals. Of note, ruxolitinib treatment decreased splenomegaly by over 50% (p=0.0001), while IFN-γ blockade had no effect. When compared to mice that received either no treatment or IFN-γ blockade, mice treated with ruxolitinib also exhibited a reduction in inflammation, which was manifested as a significantly lower area of liver encompassed by inflammatory foci, as well as a marked reduction in the absolute number of total and gp33-tetramer reactive splenic and intrahepatic CD8+ T cells. Together, these results suggest that ruxolitinib exerts its effects through mechanisms that are in part independent of the inhibition of IFN-γ signaling. We next examined whether combining ruxolitinib with dexamethasone, a gold standard medication in HLH treatment, might further diminish inflammation. To this end, LCMV-infected Prf1^-/- mice were treated or not with ruxolitinib (as above), dexamethasone (15mg/kg i.p. daily) or ruxolitinib and dexamethasone. All treatments resulted in a similar reduction in weight loss. Thrombocytopenia was abrogated with ruxolitinib treatment, either alone or in combination with dexamethasone. Mice treated with both drugs, however, did not develop splenomegaly and were less anemic than mice treated with either single agent. Strikingly, when compared to the livers of infected untreated mice, livers from mice treated with combination therapy exhibited a 99.7% decrease in the area encompassed by inflammatory foci. This was in contrast to mice treated with ruxolitinib or dexamethasone, which showed a 66.8% or 87% reduction, respectively. Consistent with these findings, combination therapy was also more potent in decreasing the numbers and to a lesser extent the frequencies of total and antigen-specific CD8+ T cells. These findings suggest that treatment with ruxolitinib and dexamethasone further limits inflammation and lessens the manifestations of HLH, which represents a serious clinical problem for which there has been limited progress in treatment. The results of these studies support the incorporation of ruxolitinib and dexamethasone into future clinical trials to improve the cure rate for HLH.