HLH: it is all about communication Free

In this issue of Blood, Frank-Kamenetskii and colleagues, using a refinement of an in vitro murine cellular model of hemophagocytic lymphohistiocytosis (HLH), were able to dissect differential mechanisms of HLH-type hyperinflammation at the level of the immune synapse (IS) communication node.¹ With this study, we get a better idea of the puzzle between different HLH origins and common pathophysiological end tracks. Additionally, they describe a new potential target for therapy: receptor-interacting protein kinase 1 (RIPK1).

The mouse model used in this study was based on the work of Jenkins et al.² Frank-Kamenetskii et al were able to identify and analyze simultaneously 3 different parameters of cellular interaction. The first area explored is impaired IS termination between the cytotoxic T lymphocyte (CTL) and the target cell. The prolonged interaction through IS is caused by a lack of a termination signal, which is given by the target cell entering apoptosis. This kind of interaction is analogous to our day-to-day business life: prolonged business meetings do not necessarily achieve better results as they are often caused by weak or unrealistic meeting goals. Instead of searching for another target cell, the defective CTL remains attached to the first one—like unlucky employees stuck at an endless meeting. At some point, they start texting other coworkers, thus creating information chaos in the company. CTLs produce the chaos, stimulating cytokines causing HLH. Cytokine concentrations were the second parameter analyzed by the authors. Although generally positively correlating with the IS time, those concentrations were also dependent on the T-cell receptor (TCR) affinity to the target cell. It turns out that the more engaged the employees are, the shorter the meeting can be (IS time), but more text messages (cytokines) may be produced. Thus, both prolonged IS interaction and high TCR affinity to the target cells can increase cytokine production. The third parameter analyzed was the mode of cell death (MOCD). Regular cytotoxic activity (mediated by perforin and granzymes) induces apoptosis, which has an immunosuppressive effect on adjacent cells. Here, the authors made a serendipitous discovery that in the absence of perforin activity, interleukin 18 (IL-18)–induced necroptosis may occur. This MOCD increases inflammation in its vicinity. This finding helps to explain the HLH predisposition in states with elevated IL-18 (X-linked inhibitor of apoptosis protein deficiency, and macrophage activation syndrome associated with NLR family CARD domain-containing (NLRC4) and Still disease) but without cytotoxic granule pathway defect.

The finding of IL-18–dependent necroptosis in HLH inspired further experiments. Suppression of this MOCD via an RIPK1 inhibitor was able to ameliorate the hyperinflammation symptoms in IL-18–dependent, and to a lesser extent, in the perforin-dependent murine HLH models. This makes RIPK1 a potential treatment target in HLH. Interestingly, elevated RIPK1 activity was described in a rare disorder called cleavage-resistant RIPK1-induced autoinflammatory syndrome. Cleavage by caspase 8 is one of the RIPK1 downregulation mechanisms. Mutant protein resistant to this cleavage has prolonged activity and leads to an autoinflammatory disorder with some symptoms resembling HLH: recurrent high fevers and splenomegaly with response to steroids and tocilizumab.³ A RIPK1 inhibitor was tested in humans without significant adverse effects, but it was not active in randomized trials comparing it as monotherapy to placebo in patients with ulcerative colitis and rheumatoid arthritis.^4,5 Another RIPK1 inhibitor developed using an artificial intelligence–driven approach has been tested in an HLH model with promising results.⁶ 

Interpretation of the extent to which an HLH-associated mutation in an individual patient causes HLH is difficult to assess. This is in particular true in cases of heterozygosity or newly described mutations with residual activity, as derived from in silico analysis. Thus, in patients with HLH gene mutations, it has to be weighed if genetic predisposition makes HLH relapse inevitable despite successful induction, so that risks associated with allogeneic stem cell transplantation are acceptable. If the pipeline for testing the effects of different mutations were standardized and optimized for easy use, in the future it may be used to better understand HLH not only at a single-cell level, but also at a single patient level—ideally for every doubtful case. This future is especially appealing as the model presented by the authors is able to detect abnormalities in heterozygous perforin deficiency.

Besides ex vivo functional assessment of the central communication node of HLH pathophysiology, the presented model of cellular communication can help in testing new drug candidates. Additionally, the model may also be useful for translational analysis of chimeric antigen receptor T-cell therapy for both efficacy and for the risk of hyperinflammatory toxicity, commonly known as immune effector cell–associated hemophagocytic lymphohistiocytosis-like syndrome.⁷ The current model is limited by its murine origin. In future development, humanized models will be needed to validate that the reported observations apply to human HLH. And finally, an array of genes other than perforin will need to be tested.⁸ 

The presented results have a great potential to influence clinical practice. Effective communication between the basic scientists and clinicians can bring these findings of cell communication to the next level of directly impacting patient care.

Conflict-of-interest disclosure: The authors declare no competing financial interests.