Continuous blockade of IFN-γ signaling is necessary for prevention of HLH and IFN-γ mediated pathologies. (A) Survival and global clinical score of perforin-deficient (prf^−/−) mice after LCMV infection and ruxolitinib treatment. All agents were started 5 days after infection, before onset of HLH, as preemptive treatments (as indicated by clinical score¹³). Two dose levels of ruxolitinib (1 or 90 mg/kg per dose, oral gavage [PO] twice daily, as previously reported^7,8) and 2 routes of administration (oral or intraperitoneal [IP]) were tested. For comparison, wild-type (WT) mice and prf^−/− mice given control treatments (carrier gavage) are shown. (B) Kinetic assessment of IFN-γ signaling blockade in vivo by ruxolitinib. Mice were treated with ruxolitinib (90 mg/kg, IP) or anti-IFN-γ antibody (40 mg/kg XMG1.2) at the indicated times before injection with IFN-γ (300 ng, IP). Twenty minutes later, animals were euthanized and peritoneal macrophages were harvested via peritoneal lavage with 1% paraformaldehyde. After methanol permeabilization, phospho-STAT1 was measured in F4/80⁺ cells by flow cytometry. (C) Development of splenomegaly and anemia after IFN-γ infusion. Uninfected WT mice were infused with IFN-γ (or phosphate-buffered saline) via osmotic pumps (as described previously¹⁴) and spleen weights and blood hemoglobin levels were assessed 6 days later. Mice were simultaneously treated with saline, ruxolitinib (90 mg/kg, twice daily), or anti-IFN-γ antibody as indicated. (D) Survival and clinical score after LCMV infection of prf^−/− mice treated with ruxolitinib administered continuously in the drinking water (0.5 mg/mL, delivering 45 mg/kg per day). (E) Assessment of IFN-γ signaling blockade in vivo by continuously administered ruxolitinib. Mice received ruxolitinib in the drinking water (0.5 mg/mL) for 12 hours and were assessed as described in B. N = 5-39 mice/group (39 mice total treated with ruxolitinib in panel A) combined from 3 to 12 independent experiments. *P < .05, **P < .01, ***P < .001.