Portal de Periódicos da CAPES

Ruxolitinib as first-line treatment in secondary hemophagocytic lymphohistiocytosis: A second experience

2018; Wiley; Volume: 93; Issue: 5 Linguagem: Inglês

10.1002/ajh.25063

1096-8652

Inuk Zandvakili, Caitlin B. Conboy, Ayed O. Ayed, Elizabeth Cathcart-Rake, Ayalew Tefferi,

Parvovirus B19 Infection Studies

We report here a second case of a patient with hemophagocytic lymphohistiocytosis (HLH) treated with the JAK inhibitor ruxolitinib as a first-line treatment. HLH is a syndrome of extreme but ineffective immune response driven by hyperactivation of CD8 T-cells and macrophages, resulting in immune infiltration of tissues and hemophagocytosis, and subsequent multi-organ failure. The liver, spleen, and bone marrow are primarily affected resulting in cytopenias. HLH is divided broadly into primary and secondary forms. The primary form is a heterogeneous group of genetic disorders associated with various immunodeficiencies, and commonly presents in infancy. Secondary HLH is typically triggered by malignancy, infection, or by autoimmune and rheumatological conditions. When resulting from a rheumatological condition such as adult onset Still's disease, it is termed macrophage activation syndrome. In all cases, mortality from HLH is very high, and in adult patients is estimated to be ∼66% by two years, with mean survival of 2 months.1 The pathophysiology of HLH is intimately related to macrophage activation, regulation, cytokine release, and phagocytosis. When macrophages are activated by a stimulus such as infection there is interaction with, and regulation by, natural killer (NK) or cytotoxic lymphocytes (CTL). An important form of macrophage regulation occurs via destruction of macrophages via perforation of the cell membrane by perforin and injection of cytotoxic substances such as granzyme B protein. The perforin-dependent cytotoxicity pathway is a locus of many of the genetic abnormalities resulting in primary HLH in humans. Perforin-deficient transgenic mice serve as models of HLH and can be induced into an HLH-like state by challenge with murine virus. Study of these mice demonstrated that interferon gamma (IFN-γ) activation was key to HLH induction in mice and that blockade of the IFN-γ is sufficient to block immune hyperactivation. Since IFN-γ and various other cytokines implicated in HLH in humans [primarily interleukins (IL) 2, 6, 10 and 18] signal via the JAK-STAT pathway, it was hypothesized that JAK-STAT inhibitors may be effective in treating HLH. This was successfully demonstrated in several mouse models of HLH using the JAK inhibitor, ruxolitinib.2, 3 Previously, there have been case reports of ruxolitinib use in HLH after failure of established therapies, primarily dexamethasone, and etoposide.4, 5 Clearly, use of cytotoxic agents in critically ill patients with secondary HLH can be fraught with challenges and at times may not be feasible. Our group first reported first-line use of ruxolitinib in a patient with secondary HLH from disseminated histoplasmosis.6 The patient was immunosuppressed from methotrexate and hydroxychloroquine use for rheumatoid arthritis. The patient's clinical parameters improved dramatically with a combination of antifungals and ruxolitinib. Specifically, improved parameters included septic shock, cytopenias, liver function, and inflammatory markers including elevated ferritin. Here, we present a second case of ruxolitinib as first-line treatment for secondary HLH. A 72-year-old female with a history of inflammatory bowel disease and seropositive destructive rheumatoid arthritis on chronic methotrexate and infliximab presented to our hospital for one month of fatigue and generalized weakness. On admission, she had altered mental status, and vitals were notable for fevers up to 39.5°C with intermittent tachycardia and hypotension. Moreover, she was hypoxemic and had a new oxygen requirement of 4 liters per nasal cannula. She denied fevers, lymphadenopathy, or night sweats, but endorsed 10 pounds of weight loss over the past few weeks, along with decreased appetite. She denied any rash, sinus congestion, dyspnea or pain, and there were no sick contacts or travel. Except for mental status, her examination was unremarkable. Initial laboratory results were notable for worsening of chronic normocytic anemia from 12 down to 10 g/dL, and new thrombocytopenia to approximately 80 K/mcL. She had no recent heparin exposure and ADAMSTS13 activity level was not reduced. d-Dimer was elevated and fibrinogen was low, but activated partial thromboplastin time (aPTT) and international normalized ratio (INR) were normal and peripheral blood smear morphology was unremarkable, specifically there were no schistocytes. Moreover, she also had hyponatremia, normal stable renal function, and elevated liver function enzymes in a cholestatic pattern. Inflammatory markers were elevated with an initial CRP of 87.1 mg/L. Due to her oxygen requirement and symptoms a CT chest was obtained which demonstrated impressive diffuse bilateral ground-glass opacities with sparse focal infiltrates. Given the fevers, cytopenia, hypofibrinogenemia, and elevated liver enzymes, we were concerned for HLH and therefore obtained ferritin, triglycerides, soluble IL-2 receptor alpha, immune phenotyping of the peripheral lymphocytes by flow cytometry, and bone marrow biopsy on hospital day 2. Empiric antibiotics had been started for possible community acquired pneumonia in an immunocompromised host and she underwent a bronchioalveolar lavage with a broad infectious work-up on hospital day 3. Unfortunately, due to hypoxemia and hypotension during bronchoscopy, the patient was intubated, mechanically ventilated, and started on vasoactive medications. The following day, on the evening of hospital day 4, histoplasma polymerase chain reaction (PCR) from the bronchioalveolar lavage returned positive and she was started on liposomal amphotericin B. On hospital day 4, preliminary hematopathology of her bone marrow examination demonstrated hemophagocytic macrophages. Her ferritin returned at 1541 mcg/L, along with other supporting evidence of HLH including, triglycerides 274 mg/dL and soluble IL-2R (CD25) 13840 pg/mL. NK cell activity was not assayed. The spleen was not palpable, nor was it elongated on abdominal ultrasound, but it did appear enlarged on CT. Therefore, the patient fulfilled 7 of 8 diagnostic criteria for HLH.7 In light of her critical and tenuous clinical condition, we elected to target her HLH with ruxolitinib, which was started at a dose of 10 mg PO BID. Upon starting antifungal and ruxolitinib treatment, various parameters improved. Clinical indicators of septic shock and hypoxic respiratory failure improved, and she was extubated and weaned off vasoactive medications within 24 and 48 hours, respectively. As outlined in Figure 1, her cytopenias, particularly thrombocytopenia, started improving within 48 hours, with a more rapid phase of improvement 72 hours after starting ruxolitinib (Figure 1). With ongoing clinical improvement, amphotericin B transition to itraconazole began on hospital day 7. Itraconazole can increase serum ruxolitinib levels as an inhibitor of multiple cytochrome P450 enzymes, and therefore we reduced ruxolitinib beginning hospital day 9, with plans to taper over a 3-week period. The patient steadily improved, was discharged from the hospital to a skilled nursing facility, and made a full recovery. She will continue on itraconazole for one year. Her laboratory findings improved to near normal after 6 weeks (Supporting Information Table 1). Hematologic and laboratory parameters after start of ruxolitinib for HLH. In summary, we have found further evidence that early intervention of secondary HLH with ruxolitinib is feasible, well-tolerated and may benefit the patient. Current ongoing trials investigating the use of ruxolitinib in HLH (NCT02400463) should provide definitive answers as whether early intervention has a clinically significant effect on the morbidity and mortality of this deadly condition. Nothing to report. Additional Supporting Information may be found online in the supporting information tab for this article. Supporting Information Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.