Background/Purpose:

Persistently and extremely elevated serum IL-18 has been associated with Macrophage Activation Syndrome (MAS). Chronic IL-18 is hypothesized to contribute to excessive interferon (IFN)-g and MAS-like inflammation. However, NK cells in systemic juvenile idiopathic arthritis (sJIA) fail to increase IFN-g production in response to IL-18. In contrast to IL-18 insensitivity in sJIA, acutely blocking IL-18 successfully treated a patient with NLRC4-MAS.[1] Due to the complex role of IL-18 in MAS and related diseases, we proposed to 1) characterize IL-18 in hyperferritinemic diseases, and 2) characterize NK and other IL-18 responsive cells in murine models of chronic IL-18 exposure.

Methods:

Serum from patients with active sJIA, sJIA-MAS, Infection-Associated Hemophagocytic Lymphohistiocytosis or familial HLH (fHLH) was assayed by ELISA for total IL-18 and related cytokines. Transcriptional and flow cytometric assessment of lymphocytes from WT, Nlrc4 mutant (N4-TS), and IL-18 transgenic (Il18tg) mice was also performed.

Results:

Disease activity markers were comparable across the human cohort. Total IL-18 was substantially higher in sJIA-MAS than fHLH. IL-18 Binding Protein (IL-18BP) was not significantly different between hyperferritinemic syndromes, resulting in substantial free IL-18 in sJIA-MAS. By contrast, CXCL9 (a marker of IFN-g activity) was somewhat higher in fHLH. The ratio of total IL-18 to CXCL9 provided a basis for distinguishing sJIA-MAS from fHLH (ROC area under the curve = 0.93). A ratio > 2.3 provided 83% sensitivity and 100% specificity in distinguishing MAS from fHLH. (Figure 1) 

Though we have observed increased serum IL-18 in N4-TS mice, we detected free IL-18 only in Il18tg mice. Consistent with this finding, we identified downregulation of the IL-18 receptor (IL-18R) in NK cells of Il18tg but not WT or N4-TS mice. Il18tg NKs had increased transcription of genes encoding perforin, IL-12R, and STAT1 but not IFN-g, suggesting a poised state of activation. The pattern of IL-18R expression in T cells differed from NKs in that we observed cell surface upregulation of IL-18R on activated tissue-resident CD8+ T cells in both N4-TS and Il18tg mice, but not WT.

Conclusion:

Delays in precise diagnosis in hyperferritinemic syndromes may contribute to increased morbidity and mortality. We found that a diagnostic approach balancing assessment of inflammasome activity (represented by IL-18) with IFN-g activity (represented by CXCL9) impressively distinguished MAS from fHLH in this retrospective cohort. Whereas high IL-18 induces insensitivity in NKs, our data suggest it may promote exuberant responses by CD8+ T cells. Though the downstream effects of increased IL-18 on T cells requires further analysis, our data raise the possibility that the pathogenic effects of IL-18 in MAS are mediated through activated T cells.

Reference:

[1] Canna SW et al., JACI, 2017