The role of inflammation in MPN and its therapeutic management

Myeloproliferative neoplasms (MPN) are a group of hematological malignancies characterized by aberrant clonal expansion of hematopoietic stem cells (HSCs) in the bone marrow (BM). This results in the overproduction of one or more blood cell lineages, including white blood cells, red blood cells, and/or platelets.¹ In Philadelphia chromosome (Ph)-negative MPN, which include polycythemia vera (PV), primary myelofibrosis (PMF) and essential thrombocythemia, the majority of patients harbor mutations in the Janus kinase 2 (JAK2), calreticulin (CALR), or thrombopoietin receptor (MPL), which drive the pathogenic HSC clonal expansion.¹

Despite the genetic component of MPN, inflammatory processes that modulate the BM microenvironment and affect both the malignant and physiological HSCs seem to contribute to disease pathogenesis and progression. In a recent review by Nicolas Chatain, MPN Hub Steering Committee member Steffen Koschmieder, and Edgar Jost in Cancers,¹ the contributing role of inflammation to Ph-negative MPN pathogenesis and how it gets modulated following hematopoietic stem cell transplantation (HSCT) was discussed. We hereby provide a summary of the key messages presented by the authors.

Cytokines involved in MPN and their sources¹

MPN can be characterized as a state of chronic inflammation in the BM hematopoietic niche, which is maintained not only by the malignant cells but also by BM resident cells, like HSCs, mesenchymal stromal cells (MSCs), and endothelial cells.¹ More specifically, the malignant HSC clones that usually harbor a driver mutation in JAK2, CALR, or MPL secrete multiple inflammatory cytokines and differentiate into various progenitor cells, including megakaryocytes, monocytes, dendritic cells, and granulocytes. All these cells, in turn, signal to neighbouring BM cells to secrete more inflammatory cytokines, creating a vicious inflammation cycle. Moreover, MSCs are recruited to the BM through the secretion of the chemokine CXCL4 from mutated megakaryocytes and differentiate into myofibroblasts, thus contributing to collagen deposition and fibrosis in the BM, a classic symptom of MPN. All this aberrant inflammation impairs normal hematopoiesis and leads to the reduction of non-malignant blood cells, resulting in anemia, thrombocytopenia, and neutropenia.¹

Multiple cytokines have been shown to contribute to this MPN inflammatory BM profile. More specifically, studies have identified 19 upregulated cytokines in patients with PMF, among which interleukin (IL)-8, IL-12, IL-15, soluble IL-2 receptor (sIL-2R), and the inducible protein 10 (IP-10) were independent prognostic markers for poor survival outcomes. In patients with PV, the macrophage inflammatory protein 1b (MIP1b) was correlated with a shortened survival. The same study also reported significant associations between the levels of interferon (IFN)-a and IFN-γ with the clinical sign of thrombocytosis, while increased levels of IL-1β, IL-2, Il-7, fibroblast growth factor β (FGF-β), and hepatocyte growth factor (HGF) were linked to leukocytosis. Last but not least, changes in hematocrit that are observed in patients with PV were correlated with increased IL-12 expression.¹

Other cytokines that have been extensively studied in MPN are tumor necrosis factor α (TNFα), IL-6, and transforming growth factor-β (TGF-β). One preclinical study reported that TNFα was necessary for colony stimulation of JAK2-mutated MPN cells and thus the generation of the MPN phenotype in those mouse models. Interestingly, TNFα expression did not only lead to the expansion of the mutated clones, but also inhibited the growth of JAK2 wild type cells in the same model, something that has also been observed with the adipokine, lipocalin-2 (LCN2), which is also upregulated in MPN.¹

A recent gene expression study reported an upregulation of inflammatory genes in patients with myelofibrosis (MF), particularly with excessive IFN, TNFα and mature dendritic cell  signatures. Importantly though, they showed that the genetic inflammatory signature was different for each patient, suggesting that the same disease phenotype can arise from the activation of different inflammatory pathways.¹

Except for cytokines, reactive oxygen species (ROS) have also been implicated in disease progression in MPN, albeit with some controversy. Normal ROS levels or ROS detoxification have both been shown to favor malignant clonal expansion. Nevertheless, other pre-clinical studies have reported increased levels of ROS in MPN JAK2-mutated mouse models and a reduction in splenomegaly, clone cell infiltration, and DNA damage following treatment with the antioxidant N-acetylcysteine (NAC). NAC was also recently shown to reduce the risk of thrombosis in another JAK2-mutated mouse model.¹ The precise role of ROS in MPN progression is yet to be clarified.

HSCT modulates inflammation in MPN¹

HSCT is currently the only curative treatment for MPN. With the use of HSCT, the malignant cells are replaced by normal donor HSC progenitors and other immature and mature immune donor cells, which become activated in the host. Since the main graft source for HSCT is peripheral blood-derived stem cells, no MSCs, endothelial cells, or other BM stromal cells are transfused into the host. This means that following HSCT, the healthy donor-derived HSCs will solely interact with the surrounding BM milieu without the presence of the malignant clones and their aberrant inflammatory stimuli.

As mentioned above, the malignant clone is believed to initiate the inflammatory processes and cell differentiations that finally lead to BM fibrosis in MPN. Until recently, BM fibrosis was considered to be an irreversible process. Nevertheless, we now know that it is a dynamic process that can be reversed in the absence of the inflammatory signals produced by the malignant clones, upon HSCT. This graft-versus-MF effect seems to be mediated by T cells and can take from one to several months.¹ The precise factors involved in this BM remodelling following HSCT are largely unknown. Preliminary data have correlated a decrease in TIMP metallopeptidase inhibitor 1 (TIMP1) and platelet-derived growth factor subunit A (PDGFA) with BM remodelling in patients with MF, who achieved remission following HSCT. Except for the reversal of BM fibrosis, the elimination of the malignant clone following HSCT largely leads to a sustained decrease in BM inflammation and the normalization of the BM stromal cells, thus improving patient outcomes. Nevertheless, in some cases of MF, the initial BM stromal dysfunction does not seem to completely reverse following HSCT and a higher rate of graft failure has been observed. This is true especially for some patients with high- or intermediate-risk MF, or those with concomitant myelodysplastic syndrome and BM fibrosis.¹ Thus, certain abnormalities induced by the initial malignant clones may be irreversible in some patients even following HSCT (learn more about the curative role of HSCT here).

Despite the great therapeutic potential of allogeneic HSCT for MPN, they do confer the risk of developing graft-versus-host disease (GvHD). GvHD occurs due to the deleterious interactions of alloreactive CD4⁺ T cells with the donor HSCs and can lead to the ‘re-activation’ of the BM inflammatory state. Another consideration is the accompanied long-term bone resorption that can occur following allogeneic HSCT. In fact, preclinical studies have shown that in mouse models of GvHD, along with HSCs, osteoblasts are also targeted by alloreactive CD4⁺ T cells, resulting in bone loss. Thus, anti-CD4 therapies might present a good option for preventing HSCT-mediated bone loss in patients with MPN.¹

Therapeutics that reduce inflammation and prevent graft failure¹

JAK inhibitors in HSCT

JAK inhibitors are the therapeutic core of MPN clinical management. Except for the inhibition of JAK2 MPN-related mutations, they also block a wide range of inflammatory signals and pathways. The JAK1/2 inhibitor, ruxolitinib, has been approved by the U.S. Food and Drug Administration (FDA) for the treatment of patients with MF. Ruxolitinib is considered as the standard of care for patients with MPN 2–3 months prior to HSCT but has also shown efficacy post-HSCT. The activity of ruxolitinib pre- and post-HSCT seems to be due to its immunosuppressive action that leads to a reduced inflammatory state. However, ruxolitinib is hematotoxic, and thus needs to be withdrawn usually one day prior to the start of HSCT conditioning and during aplasia after HSCT. In some patients, this abrupt discontinuation of ruxolitinib leads to ‘ruxolitinib withdrawal syndrome’ that is characterized by acute relapse, massive cytokine upregulation, and loss of clinical benefit. Therefore, it is advised that ruxolitinib discontinuation prior to HSCT is carefully tapered and combined with a good conditioning regimen that will help minimise the risk of such complications (for more information, see here).

In one study, ruxolitinib was successfully continued together with HSCT conditioning, in an attempt to protect patients from developing GvHD.¹ The superiority of ruxolitinib over other investigator-elected treatments for the management of steroid-refractory GvHD has recently been proven by the REACH2 phase III clinical trial.² For more information on the activity of ruxolitinib against GvHD, click here. In patients with MF, ruxolitinib acts by suppressing inflammatory signals but doesn’t act on the malignant clone itself, therefore patient-specific evaluation is advised when prescribing ruxolitinib, since its mediated immunosuppression may also favor relapse and lead to serious adverse events.¹

ROS and iron overload targeting

Targeting the ROS and iron pathways may provide a useful approach for reducing BM inflammation and fibrosis in patients with MPN and could potentially aid with HSCT engraftment. As mentioned above, preclinical data have suggested that the antioxidant NAC is prophylactic in MPN mouse models and promotes the recovery of hematopoiesis. Interestingly, in patients with poor graft function, incomplete hematopoiesis recovery, or who develop GvHD, significantly increased ROS levels have been identified. Moreover, in mouse models of MPN, treatment with the proinflammatory cytokine TNFα triggers significant upregulation of ROS, which impairs the engraftment of primary and secondary transplants. NAC pre-treatment was able to reduce this TNFα-mediated ROS elevation and normalize graft function. Promising activity on graft improvement and reduction in GvHD risk has also been seen with the anti-inflammatory and immunomodulatory ROS scavenger, dimethyl fumarate. Nevertheless, some caution is warranted with the use of dimethyl fumarate as it can lead to the activation of redox-sensitive transcription factors, which have been previously described in the pathogenesis of MPN.¹

Iron overload is a common issue in patients with MPN, and excessive free iron can lead to increased formation of ROS intermediates. Increased serum ferritin and hepcidin have been linked to poor survival in patients with PMF, independent of cytokine levels (inflammatory state). Preclinical data further indicate that increased levels of iron in the BM leads to a significant engraftment and hematopoiesis recovery delay, suggesting a direct harmful effect of excessive iron on HSCs in the BM niche. Interestingly, this was reversed following NAC detoxification, indicating once again the close association between iron and ROS levels.¹ Despite most of the data on the role of ROS and iron levels in MPN being of preclinical nature, they provide a good starting point for the clinical consideration of antioxidant therapies in combination with JAK inhibition and HSCT for MPN.

Conclusion

BM inflammation plays an important role in MPN pathogenesis and disease progression. The chronic BM inflammatory state observed in MPN originates mainly from the malignant clone and is further sustained by clonal and non-clonal HSCs and neighbouring cells. The removal of the malignant clone by HSCT seems to effectively restore BM stromal cells, reduce inflammation, and reverse BM fibrosis in most cases. Nevertheless, in some cases, graft failure rate and incomplete regeneration are still being observed following HSCT. Despite HSCT being the only curative option for MPN, the risk of GvHD, graft failure, and bone loss remain important therapeutic limitations. It is certain that the correct treatment combinations before and after HSCT (i.e., JAK inhibitors, ROS scavengers) are key, not only for HSCT success but also for improving outcome in patients with MPN.