Allogeneic stem cell transplantation (ASCT) is a potentially curative therapeutic modality used primarily in the treatment of hematological malignancies and bone marrow failure syndromes (1, 2). Despite better supportive care and increased understanding of the underlying pathophysiology processes, the majority of patients will experience either acute or chronic graft-versus-host disease (GVHD) (3).

Graft-versus-host disease is a complex biological process, which includes severe immune dysregulation with sustained inflammation (4). Several studies have identified interleukin-6 (IL-6) and as an important factor in the pathophysiology of GVHD (5, 6). IL-6 is an important cytokine involved in both acute and chronic inflammation, and it plays a major role in several autoimmune disorders (7–9). It also mediates strong anti-inflammatory effects important for immunoregulation, metabolic control, and tissue regeneration (10). During the last 15 years the pleiotropic effects of IL-6 in different organ systems have been reveled, and anti-IL-6 treatment has been approved for autoimmune and neoplastic disorders (11). With increasing insight into the complex signaling events induced by IL-6, more specific blockade of the anti-inflammatory functions of IL-6 is under development.

In current ASCT protocols an initial conditioning treatment with chemotherapy and/or total body irradiation is followed by infusion of a stem cell graft derived from either umbilical cord blood, bone marrow, or peripheral blood of a healthy donor. Severe GVHD occurs in 20–30% of adult allotransplant recipients and is an important cause of transplant-related mortality (12, 13). GVHD is classified as either acute or chronic GVHD based on clinical characteristics. According to the clinical definitions, classic acute GVHD (aGVHD) develops within 100 days posttransplant with distinctive clinical features in the skin, gastrointestinal (GI) tract, or liver. However, manifestations of aGVHD can reoccur, persist, or present after day 100 posttransplant and is then classified as recurrent, persistent, or late onset aGVHD. Classical chronic GVHD is not limited to specific organ system and can present at any time after transplantation. The diagnosis of chronic GVHD is based on consensus criteria for each organ system involved. While some clinical features are diagnostic (e.g., bronchiolitis obliterans and sclerotic features of the skin), other manifestations require additional clinical and/or histological criteria to be fulfilled. Some patients may also present as an overlap syndrome with clinical features of both acute and chronic GVHD.

Scoring systems have been developed to grade the severity of aGVHD based on the involvement of skin, GI-tract, and liver; the Glucksberg scale (grading aGVHD from I to IV) or the International Bone Marrow Transplant Registry (grading from A to D) are most widely used (14). First-line treatment for grade II–IV aGVHD is high-dose steroids (15), and a complete response is observed in 20–40% of patients (16, 17). There is no consensus regarding the preferred therapy for steroid-refractory aGVHD; commonly utilized options include TNFα blockade, anti-IL-6 antibodies, mTOR inhibitors, mycophenolate mofetil, and extracorporeal photopheresis (18). Despite several therapeutic alternatives, patients with grade IV steroid-refractory aGVHD have a dismal long-term prognosis (15).

Donor T cells play major roles in the development of aGVHD, but multiple factors contribute to the overall risk of severe aGVHD. The development of aGVHD is believed to be a three-step process each involving different subsets of immune cells (Figure 1) (19, 20). The first step is characterized by activation of the innate immune system; this results in a state of systemic inflammation prior to the introduction of donor T cells. The chemotherapy and/or radiation used in the conditioning therapy invoke tissue damage with release of pro-inflammatory cytokines, e.g., IL-6, TNFα, IFN-γ, and IL-1. Damage to the GI barrier allows increased translocation of microorganisms from the gut microbiome, with subsequent increased levels of circulating lipopolysaccharide as well as molecules with pathogen- and damage-associated molecular patterns (PAMPs/DAMPs) (21). As a consequence, antigen-presenting cells (APC) secrete pro-inflammatory cytokines and present host as well as pathogen peptides on their major histocompatibility complex molecules. A reduction of pretransplant inflammation (e.g., through the use of reduced intensity conditioning, reduction of gut microbiome levels) and the presence of genetic factors associated with reduced inflammatory responses are associated with a reduced risk of GVHD and treatment-related mortality (22–24). Several other factors also influence the pretransplant pro-inflammatory state, e.g., pro-inflammatory metabolites, soluble adhesion molecules, and factors associated with altered endothelial cell functions with increased vascular permeability (6, 25–27).

The second step is characterized by alloreactive T-cell activation, proliferation, and differentiation as a response to the presentation of host antigens by host APC in a proinflammatory context. Stimulation of specific T helper cell subsets is thought to be important for the initiation of later cytotoxic T-cell-mediated tissue damage (19, 20). Th1 cells release interferon-γ (IFN-γ) at high levels and express the transcription factors STAT4 and STAT1/T-bet (28, 29). This Th1 polarization seems to be important for the development of aGVHD, especially in the GI tract (30). Additionally, Th2 cells characterized by expression of the transcription factor GATA-3 and secretion of anti-inflammatory cytokines IL-4, IL-10, and IL-13 (31) may also play a role in GVHD pathogenesis. The available studies of Th2 cells in GVHD have given conflicting results, but some studies have described an association between Th2 differentiation and pulmonary as well as skin involvement in aGVHD (30, 32, 33). Th17 cells are characterized by the expression of the transcription factor RORγt and IL-17 secretion (34), and IL-17-secreting T cells are thought to be important for aGVHD severity (35) as well as for early transplant-related severe lung injury (36). Finally, T regulatory cells are a specialized Th subset characterized by expression of the transcription factor FOXP3 and high IL-2 receptor (CD25) expression. These cells play an important regulatory role by actively suppressing immune responses through their release of the anti-inflammatory cytokines IL-10, TGF-β, and IL-35 (37), and through direct interaction with other T cell subsets. Tregs are suppressed during GVHD, and resolution of GVHD is associated with restoration of the Treg function (38). Of particular interest, given the current topic of discussion, a majority of early posttransplant circulating TCRαβ⁺ CD4⁺ and CD8⁺ T cells release IL-6 at relatively high levels together with classical proinflammatory cytokines such as IFN-γ and TNFα (39).

The third step of GVHD development is characterized by local action of cytotoxic CD8⁺ T cells triggered by the local release of chemokines during the two first steps; these cells mediate direct cytotoxic effects upon target cell recognition, including secretion of perforin/granzyme and FAS-ligand. Other T-cell subsets contribute to the organ-specific manifestations of GVHD through polarization of macrophages toward a pro-inflammatory (M1) phenotype that further increase tissue damage through the release of oxidants and pro-inflammatory cytokines including TNF-α and IL-6 (40, 41).

The acute phase response represents a physiological increase in the levels of certain serum proteins due to increased production and release especially by the liver, and this includes C-reactive protein (CRP), serum amyloid P, ferritin, mannose binding protein, and fibrinogen (62). IL-6 is the main driver of the response, but other cytokines (IL-1, IL-8/CXCL8, and TNFα) also contribute (62). The levels of several acute phase proteins (e.g., CRP) is strongly correlated with IL-6 levels, and the IL-6 levels are again often correlated with the extent of tissue damage. In some malignancies there is evidence of chronic inflammation with a persisting acute phase response; but this response may also be due to constitutive IL-6 release by malignant cells (63, 64).

Both IL-6 and CRP levels are elevated in most patients prior to ASCT, and this may be due to chronic fungal or bacterial infections, GI barrier break, or residual malignant disease (65, 66). The impact of pretransplant CRP levels after allotransplantation has been investigated in several studies (6, 65–71). Elevated CRP levels seem then to independently entail a higher transplant-related morality without increasing the risk of acute or chronic GVHD. The effect of both CRP and IL-6 levels has only been investigated in two studies (6, 66). Although CRP and IL-6 serum levels are highly correlated, CRP seems to independently influence the TRM, while the pretransplant IL-6 level does not seem to have a similar impact on TRM or GVHD. A possible explanation for this could be that the CRP levels are influenced by several comorbidities and not only IL-6.

Interleukin-6 has both direct and indirect effects on immunocompetent cells involved in the development of GVHD. Several of these effects are summarized in Table 2. First, IL-6 is essential for maturation, proliferation, differentiation, and maintenance of B cells/plasma cells and several proinflammatory T-cell subsets. IL-6 seems to enhance the development of pro-inflammatory Th17 and Th2 cell and to inhibit the development T regulatory cells. Second, IL-6-signaling is crucial for trafficking of immune cells to inflamed tissues and lymphoid organs. This is caused both by altered expression of adhesion molecules by endothelial cells and by expression of their ligands by immunocompetent cells. Third, IL-6 has important functions in GVHD target organs, and there may therefore be a risk of combined injury during GVHD (e.g., GVHD-induced immune-mediated damage, pharmacological toxicity, and IL-6 inhibition). Finally, IL-6 together with TNF-α released from macrophages has been reported to directly/independently contribute to tissue damage in GVHD (40).

Interleukin-6 is important for the regulation of stem cells and tissue regeneration in several organs. This has been best demonstrated for hematopoiesis, liver cells, GI mucosa, and muscle cells. A complete overview of these effects is beyond the scope of this article. A brief overview of the major effects of IL-6 in these different organ systems is given in Table 3. Impaired IL-6 function in these organs is associated with reduced regeneration after injury, and IL-6 dysregulation during chronic inflammation can contribute to organ dysfunction.

The role of IL-6 in acute and chronic GVHD has been investigated in several mouse models. Givon et al. (110) examined the effect of IL-6 on bone marrow reconstitution after syngeneic, semi-allogenic, and ASCT and found that posttransplant treatment with subcutaneous recombinant IL-6 significantly supported white blood count reconstitution and improved survival in syngeneic as well as allogenic transplantation. This effect was only significant for animals transplanted with a low stem cell dose. In contrast, mice receiving IL-6 showed increases in both the severity of and mortality from GVHD.

Chen et al. (111) observed increased systemic IL-6 and IL-6R levels early after both syngeneic and allogeneic transplantation. These levels returned to baseline over time in the syngeneic group, whereas IL-6 levels remained high in mice developing GVHD. Both IL-6 and IL-6R expression increased in the liver and colon (but were stable in the spleen), and the highest IL-6R mRNA levels were observed in these two organs. Selective IL-6 knockout in neither recipient nor donor cells was sufficient to protect from GVHD. However, GVHD treatment with anti-IL-6 resulted in significantly lower weight loss, less histopathological evidence of damage to colon, liver, and lungs, and significantly increased Treg levels in the spleen. The increased Treg levels were not dependent on an intact thymus; rather the IL-6 blockade increased peripheral generation of Treg cells and reduced the levels of Th1 and Th17 cells. Similar results were shown by Noguchi et al. (112); treatment with an anti-IL-6 antibody reduced liver enzyme levels as well as the occurrence of organ failure and was associated with reduced infiltration of Th1 and Th17 cells, increased Treg cells, and improved survival.

Tawara et al. (113) investigated the effect of IL-6 derived from donor T cells. In contrast to the findings by Chen et al., selective IL-6 knockout in donor/graft T cells was associated with less severe GVHD and prolonged survival. Pretransplant anti-IL-6 treatment also significantly improved survival and clinical as well as histopathological severity of GVHD, but the Treg levels were not altered. Importantly, the GVL effect was maintained despite the reduction in GVHD. The systemic cytokine levels and the levels of circulating cells were not altered by the treatment, and selective ablation of IL-6 in recipient bone marrow cells did not reduce the occurrence or severity of GVHD.

Organ-specific effects of IL-6 in GVHD have been investigated in two different mouse models. Varelias et al. (36) examined the role of IL-6 in idiopathic pulmonary syndrome (IPS) after ASCT and demonstrated that local secretion of IL-6 induced Th17 cell differentiation that was necessary for disease development. IPS development could be prevented by IL-17 knockout or by the use of neutralizing antibodies against IL-17. In a second study, Le Huu et al. (114) investigated the role of IL-6 in a sclerodermous cGVHD model and observed that IL-6 levels increased during disease progression. Treatment with IL-6-neutralizing antibodies prior to manifestations of scleroderma significantly decreased disease severity, while no reduction was observed when IL-6 neutralization was started after the onset of cGVHD. Treatment with anti-IL-6 was associated with a significant increase in the number of splenic Treg cells, whereas the expression of IFN-γ, TNFα, IL-6, IL-18, TGF-β1, CCL2, CCL3, and CCL5 in affected skin was significantly reduced.

The role of STAT3 in the regulation of activation and differentiation of Tregs and Th17 cells after allotransplantation has been investigated in several mouse models (i.e., STAT3-knockout mice) and in vitro models. Emerging evidence also suggests that activation of STATs in B cells is also important for GVHD (115). A complete review of this scientific field is beyond the scope of this article, but an overview of important observations is given in Table 4 (116–120). In summary, these reports suggest that targeting of IL-6/JAK2/STAT3 signaling can be effective against the development of GVHD. Treatment can be given either by graft engineering or as posttransplant therapy, the function of several immunocompetent cells will then be affected, and reduced GVHD severity as well as improved survival has been observed. Based on these findings, treatment with the JAK1/2 inhibitor ruxolitinib has been tested in patients with manifest chronic and aGVHD, and it was reported to improve GVHD and reduce serum cytokine levels (121, 122). Patients included in these two studies are heterogeneous with respect to GVHD manifestations, previously GVHD prophylaxis and treatment. Further well-designed studies are definitely needed to determine the efficacy and toxicity of JAK2/STAT3 blockade in allotransplant recipients.

Graft-versus-host disease is a complex, multi-systemic disease in which local and systemic factors play a role in orchestrating inflammation, tissue repair, and regeneration. The impact of SNPs in the IL-6 gene on GVHD risk indicates that IL-6 plays a role during early GVHD development. Blockade of the pro-inflammatory effects of IL-6 seems to be a possible therapeutic strategy since IL-6 and STAT3 activation are closely linked to the development of both Th17 and Treg cells, and since early STAT3 phosphorylation posttransplant seems to precede development of GVHD (118). However, currently available clinical data indicate that IL-6 blockade is most effective when used as GVHD prophylaxis, whereas manifest GVHD might be less susceptible to IL-6 blockade. An unselective systemic blockade of IL-6 is probably not without caveats, given the physiological importance of IL-6 in tissue regeneration and homeostasis in both the liver and GI tract, two organs commonly affected both in acute and chronic GVHD.

Most of the pro-inflammatory effects of IL-6 seem to be caused by trans-signaling. A fusion protein consisting of the immunoglobulin Fc region and gp130 can be used for specific blocking of trans-signaling (56); this protein binds IL-6 in complex with sIL-6R but cannot bind free IL-6. Treatment with this selective inhibitor inhibits inflammation and preserves the regenerative effects of IL-6 in mouse models of a wide variety of diseases, e.g., ulcerative colitis, rheumatoid arthritis, and acute pancreatitis-associated lung injury (56). All experimental models of GVHD explored so far have utilized either genetic or pharmacological blocking of IL-6, and little is therefore known about the relative contributions of trans- versus cis-/classical IL-6 signaling in these settings. Further models should therefore address the use of selective blockade of IL-6 trans-signaling after ASCT.

TT, EE, AT, and ØB contributed equally to this work.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.