Hepatic Involvement in Systemic Juvenile Idiopathic Arthritis and Macrophage Activation Syndrome: Clinical and Therapeutic Insights From 2 Cases

To The Editor:

We read with interest the manuscript entitled “Hepatic Manifestations in Systemic Juvenile Idiopathic Arthritis and Macrophage Activation Syndrome” by Eloseily et al,¹ describing 4 patients with systemic juvenile idiopathic arthritis (sJIA) complicated by macrophage activation syndrome (MAS) and hepatic involvement and treated with emapalumab. We report our experience with this subgroup of patients, describing 2 cases treated at Meyer Children’s Hospital in which patients experienced multiple hepatic relapses requiring a combined treatment approach.

Patient 1, a 9-year-old male individual, was initially evaluated at another hospital for sJIA complicated with MAS at onset (Table) and was treated with glucocorticoids (GCs) and anakinra. During GC tapering, he experienced a disease flare, which led to a switch to canakinumab. Nevertheless, he developed MAS, with marked elevation of transaminases and otherwise preserved liver function. Due to persistent systemic inflammation and worsening hepatic involvement despite treatment with methylprednisolone pulses, high-dose intravenous (IV) anakinra, and cyclosporine, emapalumab was initiated and administered for 4 weeks, resulting in clinical and laboratory improvement. Approximately 6 months after the last infusion of emapalumab, he experienced another flare complicated by MAS and hypertransaminasemia, leading to treatment intensification with baricitinib. At the last follow-up (12 months after baricitinib initiation), the patient was taking anakinra (100 mg twice daily), baricitinib (4 mg/day), and low-dose prednisone (< 0.1 mg/kg/day); liver function was normal, and no further flares have occurred.

Patient 2, an 8-year-old female individual, was diagnosed with sJIA (Table), complicated by MAS at onset with predominant hepatic involvement (hypertransaminasemia with preserved liver function). Initial treatment consisted of methylprednisolone pulses and anakinra. After 12 months, she experienced 2 sJIA relapses characterized by marked transaminase elevation, requiring IV GCs and prompting the addition of baricitinib. Since starting baricitinib (16 months at last follow-up), she experienced 1 additional subclinical MAS flare, characterized by hypertransaminasemia. At the last follow-up, she was receiving anakinra (100 mg twice daily), baricitinib (4 mg/day), and low-dose prednisone (< 0.1 mg/kg/day); transaminase levels remained within the normal range, and no disease flares have occurred over the past 10 months.

In both flares experienced by patient 1 and in 1 of 3 flares experienced by patient 2, hepatic manifestations were associated with concomitant elevation of interleukin (IL)-18 and C-X-C motif chemokine ligand 9 (CXCL9), indicating subclinical MAS, whereas isolated IL-18 elevation in the other 2 flares experienced by patient 2 were consistent with sJIA flare (Figure). IL-18 remained persistently elevated (median 108,306 pg/mL and 132,254 pg/mL in patients 1 and 2, respectively), even during apparent clinical and laboratory remission. Recent evidence indicates that high IL-18 levels at baseline predict a chronic, persistent disease course. Additionally, persistently elevated IL-18 levels over time, rather than at a single timepoint measurement, reflect ongoing subclinical inflammation despite normalization of clinical and laboratory variables.² These findings underscore the value of IL-18 as a dynamic biomarker of subclinical disease activity.²

• Download figure
• Open in new tab
• Download powerpoint

Figure.

Serum levels of IL-18 and CXCL9 in (A,B) patient 1 and (C,D) patient 2. ★: macrophage activation syndrome episode; ●: systemic juvenile idiopathic arthritis flare. Ba: baricitinib; CXCL9: C-X-C motif chemokine ligand 9; Em: emapalumab; IL-18: interleukin 18.

Our 2 cases share both similarities and differences with those reported by Eloseily et al.¹ Our experience confirms that hepatic involvement may occur at any point during the disease course and typically coincides with disease flares but is not necessarily accompanied by MAS. In all patients in the study by Prencipe et al, the elevated levels of transaminases were associated with increased IL-18 and CXCL9 levels, supporting the role of interferon (IFN)-γ as a potential underlying driver of liver dysfunction.³ Although concomitant immunosuppressive therapy may attenuate MAS manifestations, hepatic involvement may represent the only overt expression of disease activity. Our patients exhibited hypertransaminasemia as the sole manifestation of hepatic injury; conversely, those reported by Eloseily et al¹ also showed elevated cholestatic indices and, in 1 case, jaundice. Further, in our patients, liver biopsy revealed mild inflammatory infiltrates, whereas the cases reported by Eloseily et al¹ demonstrated more severe histopathological features, including hepatocellular necrosis and architectural distortion.

All reported patients in the study by Eloseily et al¹ showed persistently high IL-18 levels (> 40,000 pg/mL) and an inadequate response to anti–IL-1 therapy. In this subset of patients with sJIA, highly targeted biologics may be insufficient, as selective cytokine blockade can reshape the inflammatory milieu without providing adequate global immunosuppression.⁴ These patients may therefore benefit from combined strategies. These include IL-1 and IL-6 inhibition with GCs and/or T-cell–targeting agents; therapies that directly interfere with IFN-γ signaling during MAS; and bispecific agents targeting both IL-18 and IL-1 during sJIA flares.⁵

Eloseily et al¹ reported a rapid clinical and laboratory response to anti–IFN-γ therapy with emapalumab, without providing data on long-term outcomes. In our series, emapalumab was administered to patient 1, resulting in an initial clinical response, followed by disease relapse 6 months after treatment discontinuation, which prompted initiation of baricitinib. Janus kinase inhibitors such as baricitinib interfere with the IFN-γ signaling and may represent a valuable therapeutic option in selected patients.^6,7 Given that emapalumab is not widely available in Europe, off-label treatment with baricitinib may be considered, taking into account the advantage of oral administration.

In conclusion, our observations further illustrate the heterogeneity of sJIA-MAS with hepatic involvement. This subset of patients warrants further investigation to clarify the central role of IFN-γ–driven liver pathology and to guide future strategies for individualized treatments.

• Copyright © 2026 by the Journal of Rheumatology