Ruxolitinib Might Be an Effective Hormone Reduction and Replacement Drug in Children With Secondary Hemophagocytic Lymphohistiocytosis

Background: To analyze the effects of ruxolitinib on children with secondary hemophagocytic lymphohistiocytosis (HLH). Methods: Eleven pediatric patients diagnosed with HLH and treated with ruxolitinib (ruxolitinib group: group R) between November 2017 and August 2018 were retrospectively analyzed. Eleven age-matched pediatric patients with HLH undergoing conventional treatment (control group: group C) were also analyzed. Results: In group R, three patients who did not respond to conventional treatment (dexamethasone and etoposide) were treated with Ruxolitinib and their temperature decreased to normal levels. Four patients had normal temperature after conventional treatment, but they had severe organ involvement, including obvious yellowing of the skin, increased liver enzyme levels and neuropsychiatric symptoms, and they were all ameliorated with ruxolitinib treatment. Four patients were relieved with ruxolitinib therapy alone. In group C, the body temperature of eleven patients all decreased to normal levels after conventional treatment. The body temperature of group R patients decreased to normal levels more rapidly than that of group C patients. The hormone dosage in group R was signicantly lower than that in group C. Both groups were followed up for 2–2.5 years. No obvious adverse drug reactions of ruxolitinib were observed during treatment and follow-up. Conclusion: Ruxolitinib might be an effective drug in controlling body temperature and reducing inammation indicators. It might be a potential replacement for hormone therapy for HLH treatment in children, thereby reducing or avoiding hormone-related adverse reactions.


Background
Hemophagocytic syndrome is also known as hemophagocytic lymphohistiocytosis (HLH). This rare lifethreatening syndrome is characterized by excessive proliferation and activation of lymphocytes caused by cytokine storms and severe systemic in ammatory responses. According to the accepted doctrine, the pathogenesis of HLH is closely associated with the "cytokine storm". 1 Depending on the etiology, there are two forms of HLH: primary autosomal recessive inheritance, also known as familial hemophagocytic lymphohistiocytosis, and secondary HLH, which develops because of strong immune activation. Primary HLH is mostly caused by genetic defects leading to immune system dysfunction. Infection, connective tissue disease, and malignancy are considered as common causes of secondary HLH. [2][3][4] Epstein-Barr virus (EBV), a DNA virus and member of the Herpesviridae family, has been consistently associated with HLH. [5][6][7][8][9][10][11][12][13][14][15] HLH in children is a rare disease with a high fatality rate. Studies have shown that most cytokines related to HLH are activated through activation of the Janus Kinase (JAK)/signal transducer and activator of transcription signaling pathway, which not only regulates the biological activity of cytokines, but also affects the differentiation of primary T cells into T helper (TH) cell families, TH1, TH2, TH17, and regulatory T cells. 16 According to the HLH-2004 diagnostic criteria, 2,17 a diagnosis of HLH can be made if ve of the following eight criteria are ful lled: (1) fever, (2) splenomegaly, (3) cytopenias (affecting ≥2 of three lineages in peripheral blood; hemoglobin <90 g/L, platelets <100 × 10 /L, neutrophils <1.0 × 10 /L), (4) hypertriglyceridemia and/or hypo brinogenemia, (5) hemophagocytosis in the bone marrow or spleen or lymph nodes (no evidence of malignancy), (6) low or absent natural killer (NK) cell activity (according to local laboratory reference), (7) ferritin ≥500 μg/L, and (8) soluble cluster of differentiation 25 (i.e., soluble interleukin-2 (IL-2) receptor (IL-2R)) ≥2400 U/mL.
The above clinical and laboratory ndings are related to the pathophysiology of HLH. High interleukin levels cause fever. Elevated ferritin >10,000 µg/L has been demonstrated to be 90% sensitive and 96% speci c for HLH. 18-21 Activation of lymphocytes can result in high concentrations of soluble IL-2 receptor. 22 Most clinicians still adopt the HLH-04 protocol recommended by the Histiocyte Society for the treatment of HLH. 23 Conventional treatment options typically include three phases of induction therapy, maintenance therapy, and/or hematopoietic stem cell transplantation. The drugs for HLH include dexamethasone, cyclosporine A, and etoposide. However, a multi-center study in 2016 revealed no signi cant bene t from cyclosporine and intrathecal injections, and macrophage activation syndrome secondary to connective tissue disease is not always treated with the HLH-04 regimen. The ve-year survival rate for secondary HLH in adults worldwide is approximately 54% and the survival rate reported in China is even lower, ranging from 31.7-56.1%. [24][25] Hence, less toxic, more effective, and better targeted immunosuppressive treatments in HLH are urgently needed.
In recent years, JAK inhibitors have been the focus of research on new small molecule targeted therapies and can be used for the treatment of in ammatory diseases such as hematological diseases, tumors, rheumatoid arthritis, ulcerative colitis, and other autoimmune diseases. 26 Ruxolitinib is a Janusassociated kinase 1/2 (JAK1/2) inhibitor that impedes downstream signaling pathways of cytokines such as interferon-γ, IL-2, and IL-6 to reduce in ammatory responses triggered by these cytokines, which play important roles in HLH. Ruxolitinib was found to control in ammatory storms and prolong survival in secondary HLH model mice; this treatment was effective in 10 cases of HLH, 27-32 including one child from abroad. 30 Herein, we present 11 cases of children treated with ruxolitinib. We compared the effects of ruxolitinib administration in children with HLH to the effects of conventional therapy in a control group of children with HLH; both groups were followed up for 2-2.5 years.

Materials And Methods
2.1 Patients: A study was performed on 11 children diagnosed with HLH and treated with ruxolitinib (group R) and 11 children with HLH who were age-matched and were not administered ruxolitinib during the same period (group C). The diagnosis was made between November 2017 and August 2018 with a follow-up endpoint in February 2020. This study was reviewed by the Hospital Ethics Committee; informed consent was obtained from the patients' parents, who had signed a written instrument, prior to the use of ruxolitinib and specimen collection.
2.2 Inclusion criteria: Compliance with HLH-2004 diagnostic criteria 2 was the inclusion criterion for this study.
2.3 Exclusion criteria: Children underwent puri ed protein derivative skin test, chest radiography, highresolution computed tomography if necessary, and T-SPOT (T-cell enzyme immuno-spotting) to con rm the absence of tuberculosis infection, which was the exclusion criterion for this study.
2.4 Etiological analysis: All children were tested for bacteria, viruses, fungi, and parasites. Whole exome sequencing was performed for all cases in group R and no known pathogenic genes were found.
2.5 Treatment: Four patients in group R were treated with ruxolitinib alone and ve patients were treated with ruxolitinib following the failure of previous therapy. The dosage used in this study was the same as that used to treat graft-versus-host disease: 2.5 mg/dose orally twice daily for those with body weight ≤25 kg and 5 mg/dose orally twice daily for those with body weight >25 kg 33 . Children with viral infections were also treated with antiviral therapy. Group C was treated with the conventional treatment (dexamethasone and etoposide).
2.6 Follow-up: Monitoring of clinical symptoms, such as body temperature and hepatosplenomegaly, and determination of blood routine, blood biochemistry, coagulation function, C-reactive protein, erythrocyte sedimentation rate, ferritin, cytokines, and NK cell activity were performed. The children were followed-up in outpatient clinics to record their clinical symptoms, treatment status, and outcomes once every month for the rst, second, and third months, and then once every 3 months, and once every 6 months after 1 year, for a total of 2-2.5 years.
2.7 Safety evaluation: Symptoms of discomfort after taking medication were recorded. Liver and kidney functions, as well as whether the patients had co-infections were monitored.
2.8 Statistical analysis: SPSS Statistics 24 (SPSS, Inc., Chicago, IL, USA) was used for statistical analysis. For continuous data, normal distribution was expressed as the`x ± s, and the independent sample t-test was used. Data that did not meet the normal distribution were expressed as M (P25, P75), and the Wilcoxon rank-sum test was used. For continuous data from ≥3 measurements, the repeatedmeasures design data analysis of variance was used. Categorical variables were represented by N (%), and chi-square test was used. P < 0.05 was considered as statistically signi cant.

Results
3.1 General information: Group R consisted of 11 children: six girls (54.5%) and ve boys (45.5%); aged 1-6 years, with a median age of 3.3 years. The course of HLH prior to ruxolitinib administration varied from 4 days to 2 months. Group C consisted of 11 children: ve girls (45.5%) and 6 boys (54.5%); aged 1-8 years, with a median age of 3.8 years. There was no signi cant difference in the age and sex of the children in the two groups. Information on clinical features, underlying diseases, and therapy prior to ruxolitinib administration of group R is shown in Table 1. 3.4 E cacy of ruxolitinib therapy in group R: The body temperature of R1-R3 did not decrease with conventional treatment but returned to normal levels following administration of ruxolitinib. The temperature of R3 rose again after 3 days of ruxolitinib therapy but stabilized once the dose was increased from 2.5 mg bid (twice daily) to 3.75 mg bid (Figure 1). Although the body temperature of R4-R7 decreased to normal levels after conventional treatment, but they had severe organ involvement, including obvious liver damage and neuropsychiatric symptoms. They were all ameliorated by ruxolitinib treatment. R8-R11 were treated with ruxolitinib immediately after diagnosis. All children had normal body temperature at 3 days after treatment ( Figure 2).
3.5 Changes in body temperature between two groups: There were no signi cant differences in body temperature between the two groups before treatment (P = 0.24). In group R, the children's body temperature decreased to normal levels after 2 days of treatment and remained stable. In group C, the body temperature of eight patients decreased to normal levels after 3 days of treatment, whereas the body temperature of the other three cases returned to normal levels after 5-6 days of treatment. However, fever recurred in eight patients in group C at 2-7 days after the body temperature had normalized and had to be controlled with immunosuppressants. Figure 3 shows the rapid temperature decrease in group R patients compared to in group C patients: the temperature of group R patients was signi cantly lower than that of group C patients on days 2-3 and 7-9 after treatment (Pd2 = 0.022, Pd3 = 0.014, Pd7 = 0.003, Pd8 = 0.020, Pd9 = 0.031). (Table 2): Following treatment with ruxolitinib, the white blood cell (WBC) count and brinogen levels gradually increased, whereas ferritin and IL-2R levels gradually decreased. The differences in WBC, brinogen, ferritin, and IL-2R levels of two groups were signi cant compared to prior treatment (P values were 0.002, <0.001, <0.001, and 0.036, respectively). One week and one month after treatment, the WBC levels in group R patients showed signi cantly rapid improvement compared to those in group C patients (P1w = 0.037, P1m = 0.002). There was no signi cant difference in the ferritin levels of the two groups of patients (P1w = 0.398, P1m = 0.064). Although there were no signi cant differences in the brinogen and IL-2R levels of the two groups after 1 week (P1w distribution is 0.74, 0.062), these levels showed signi cantly rapid improvement in R group patients compared to those in group C patients after one month (P1m distribution is 0.035, 0.041). In group R, ve cases of EBV infection, one case of HBV infection, one case of cytomegalovirus infection, and one case of in uenza virus infection were treated with a combination of ruxolitinib and antiviral drugs, after which these patients tested negative for antiviral antibodies and their viral DNA copy numbers had decreased. In group C, there were 2 cases with Epstein-Barr virus infection and 2 cases with parain uenza virus infection. After traditional therapy combined with antiviral drugs, the clinical symptoms improved, the virus antibody turned negative, and the DNA copy number decreased.  (Figure 4): One and two months after discharge, the oral doses of methylprednisolone in both groups were signi cantly lower than those at discharge (F = 60.536 P < 0.001). Compared with group C, the average dose of methylprednisolone in children in group R was signi cantly reduced (F = 29.756, P < 0.001).

Changes in laboratory values
3.8 Safety evaluation of ruxolitinib: No headache, dizziness, or other discomfort; no gastrointestinal reactions such as nausea and vomiting; no obvious infection; no thrombocytopenia, neutropenia, or anemia; no impairment of liver or kidney function; no increase in triglyceride and cholesterol levels; and no other adverse drug reactions were observed after treatment with ruxolitinib.
3.9 Follow-up: The patients were followed up for 2-2.5 years (average of 2.4 years). In group R, the dosage of the hormone was rst reduced, and the used of ruxolitinib was discontinued after 3 months of administration. Follow-up was continued for 21-27 months. One child presented with recurrent fever, accompanied by joint swelling and pain, and was diagnosed with juvenile idiopathic arthritis; the child improved after treatment with tocilizumab. In group C, three children with juvenile idiopathic arthritis experienced recurrent symptoms after hormone reduction, presenting as fever with joint swelling and pain; they also improved after treatment with tocilizumab.

Discussion
In this study, we found that administration of ruxolitinib to children with HLH was effective for controlling their body temperature, improving in ammatory indices (ferritin, IL-2R), and ameliorating symptoms of CNS involvement. Combining hormones and antiviral agents resulted in the resolution of viral infection and reduction in the dosage of hormones.
Hermans et al. found that administration JAK inhibitors signi cantly inhibited the degranulation of mast cells and reduced the production of cytokines in an in vitro study of lymphocytes. 34 Ruxolitinib also reversibly improved the killing and degranulation of NK cells and ameliorated organ damage in HLH animal models. 16, [35][36] In 2016, Das et al. used lymphocytic choriomeningitis virus to infect perforinde cient mice and construct a model of secondary HLH. A large dose of ruxolitinib (90 mg/kg) not only improved the disease symptoms and decreased cytokine levels in HLH model mice, but also increased the survival rates in mice. [37][38] A small dose of ruxolitinib (1 mg/kg) also signi cantly improved long-term survival and clinical symptoms and promoted liver tissue regeneration. 2 Subsequently, a case of an 11year-old child with refractory HLH who was treated with a combination of dexamethasone, etoposide, ruxolitinib (2.5 mg), and alemtuzumab was reported,. In ammatory factor levels rapidly decreased, organ function was restored, and no HLH relapse was observed even after etoposide treatment was discontinued. 30 Ruxolitinib was used as rst-line treatment with dexamethasone in a 71-year-old patient with HLH. Administration of ruxolitinib (10 mg/dose, twice daily) was started on the 8th day of hospitalization; the patient's condition and laboratory values improved on the 15th day of hospitalization. 31 A 38-year-old patient was treated with dexamethasone, immunoglobulin, etoposide, and rituximab; after administration of ruxolitinib (20 mg/dose, twice daily), there was an improvement in the patient's phagocytic indicators such as serum ferritin, lactate dehydrogenase, and brinogen levels, but the patient eventually died of intracranial bleeding and multiple organ failure. 32 In this study, refractory cases of HLH were treated with ruxolitinib, resulting in the regulation of body temperature and in ammatory factors and improvement of CNS involvement. In addition, administration of ruxolitinib reduced the hormone dose, which is important for the growth and development of children.
Five cases (R1-R5) in group R had recurrent fever and showed no improvement in clinical features or in ammatory indicators such as IL-2R, ferritin, and C-reactive protein levels despite hormonal, immunosuppressive, and immunoglobulin therapies; hence, ruxolitinib was added to their treatment regimen. Their body temperature subsequently dropped rapidly and in ammatory indicators improved, demonstrating that ruxolitinib controlled the body temperature and inhibited the in ammatory response.
Two cases (R6, R7) with CNS involvement showed no improvement in neuropsychiatric symptoms after termination of the HLH-04 regimen, but gradually recovered with administration of ruxolitinib, suggesting a positive effect with CNS involvement. However, other studies have shown unsatisfactory therapeutic effects of ruxolitinib in patients with HLH combined with CNS involvement, attributing the lack of e cacy to its large molecular weight which prevents its penetration across the blood-brain barrier to act on the CNS. 39 This contradictory evidence warrants further study.
Four cases (R8-R11) were treated with ruxolitinib directly without using the HLH-04 regimen. One patient was diagnosed with HLH and HBV and was reported in the journal of Pediatric). As immunosuppressants and hormones may aggravate HBV infection, the child was treated with ruxolitinib combined with antiviral therapy using entecavir. All disease indicators improved signi cantly. During follow-up after more than 1 year, the body temperature of the child remained stable, and HBV-DNA was undetectable because of the treatment with antiviral drugs. This is the rst study to use ruxolitinib alone to treat HLH. The other three children also had normal body temperature after treatment with ruxolitinib, and all indicators showed improvement.
Compared with group C, the dosage of hormones in R group children was apparently reduced throughout the treatment, suggesting that ruxolitinib can reduce the dosage of hormones or even replace hormone therapy, thereby reducing or avoiding hormone-related adverse reactions.
In children with co-infections, including EBV and HBV, there was no exacerbation of existing infections during ruxolitinib administration combined with antiviral therapy.
During the 2-2.5 years of follow-up, no children in group R had any co-infections or recent adverse reactions. One patient developed fever again following discontinuation of ruxolitinib and had joint swelling and pain. That patient was diagnosed with juvenile idiopathic arthritis, and the patient's condition improved after treatment with tocilizumab.

Conclusion
To date, there have been a few reports of individual cases being treated with ruxolitinib. In this study, we present 11 cases of children treated with ruxolitinib, demonstrating that ruxolitinib might be effective for treating HLH and is convenient to administer. It may be used as a rst-line treatment for HLH with hormone reduction. Treatment with ruxolitinib also improved the symptoms of in ammatory factors and CNS involvement in refractory HLH. A limitation of this study was the small sample size. In future studies, a large sample cohort should be used to con rm the safety, optimal dosage, treatment duration, withdrawal criteria, long-term effects, and e cacy of ruxolitinib in HLH caused by different etiologies. The ndings of this study indicate that ruxolitinib can partially replace hormones and become a potential rstline drug for the treatment of HLH, which is important for the growth and development of children. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
The study was reviewed and approved by the Hospital Ethics Committee of Capital Institute of Pediatrics.

Consent for publication
Written consent was obtained.   Mean body temperature of two groups: Body temperature returned to normal after 3 days of ruxolitinib treatment. The mean body temperature of group R patients was lower than that of group C patients.

Figure 4
Comparison of MP oral doses between both groups after discharge. Compared with group C, the average dose of MP in children in group R was signi cantly reduced. Abbreviations: MP, methylprednisolone; group R, ruxolitinib-treated group; group C, control group treated with conventional therapy.