Tipping the GVH/GVL balance by targeting HIF1α

Summary

Promoting GVL activity while eliminating GVHD is the utmost goal to treat hematological malignancies with allo-HCT. Bailey et al. demonstrate that targeting HIF1α can favor GVL activity while limiting GVHD after allo-HCT even in combination with immune checkpoint inhibition.¹

Promoting GVL activity while eliminating GVHD is the utmost goal to treat hematological malignancies with allo-HCT. Bailey et al. demonstrate that targeting HIF1α can favor GVL activity while limiting GVHD after allo-HCT even in combination with immune checkpoint inhibition.¹

Main text

Allogeneic hematopoietic cell transplantation (allo-HCT) is a valid therapeutic option that can cure various hematological malignancies through the graft-versus-leukemia/lymphoma (GVL) effect. However, graft-versus-host disease (GVHD) remains a major cause of morbidity and mortality that limits its success. Allogeneic T cells are a key driving force to induce GVHD, including acute and chronic types.

In the clinic, the HLA-matched related and unrelated donor are the two most commonly used approaches in allo-HCT. The standard regimens for GVHD prophylaxis are a calcineurin inhibitor (CNI) plus methotrexate, CNI plus mycophenolate mofetil (MMF), or a combination of post-transplant cyclophosphamide (PTCy), tacrolimus (Tac), and MMF. The three-drug combination was reported as the most promising regimen, which resulted in an adjusted 52.7% GVHD-free, relapse-free rate of survival. However, 53.8% of patients still developed grade II-IV acute GVHD (aGVHD) by day 100 and 21.9% chronic GVHD (cGVHD) by 12 months after HCT.² In late 2021, the FDA approved abatacept (CTLA4-Ig) as a GVHD prophylaxis regimen for matched unrelated donor HCT. With an abatacept/CNI/methotrexate combination, the 100-day incidence of grade II-IV aGVHD was 43.1% and the 1-year rate of moderate-severe cGVHD was 44.6%.³ The Jak1/2 inhibitor ruxolitinib was FDA approved in 2019 for treatment of steroid-refractory aGVHD, and in 2021, for treatment of cGVHD after failure of one or two lines of systemic therapy; it is currently under investigation as a GVHD prophylaxis regimen. Despite recent progress, new agents with high efficacy and low toxicity are still warranted.

Immunotherapy via targeting PD-1/PD-L1, CTLA-4, or both, known as immune checkpoint inhibition (ICI), is an important breakthrough in cancer therapy for treatment of hematological and non-hematological malignancies.⁴ Although ICI reduced malignancy relapse, augmented alloresponse and exacerbated GVHD hindered the success of ICI in patients following allo-HCT.⁵ Thus, new and better approaches are necessary to enhance the GVL effect without jeopardizing the GVHD.

HIF1α is a common transcription factor (TF) that mediates cellular responses to limited oxygen. HIF1α also critically regulates functions of many types of immune cells, including T cells, dendritic cells, macrophages, and neutrophils. In T cells, HIF1α plays an important role in T cell differentiation, especially in Th17/Treg balance, follicular regulatory/helper T cell balance, and T cell glycolytic metabolism and cytotoxic activity.^6,7 HIF1α inhibitor echinomycin was shown to attenuate aGVHD severity while preserving the GVL effect.⁸ Similarly, Wu et al. reported that miR-31, a promoter of HIF1α activity, was essential for T cell pathogenicity in cGVHD development.⁹ In both studies, HIF1α-inhibition or -downregulation resulted in reduction of T cell expansion and an increase in donor Treg development while diminishing Th17 or Th1 responses.

In the current study published in Cell Reports Medicine, Bailey et al. evaluated HIF1α as a therapeutic target alone or in combination with ICI in the treatment of hematological malignancies after allo-HCT.¹ The clinical relevance of HIF1α in aGVHD pathogenesis was shown by elevated expression of HIF1α and its target genes in peripheral blood mononuclear cells (PBMCs) from the patients with aGVHD. By using genetic and pharmacologic approaches in clinically relevant murine and humanized GVH/GVL models, they convincingly demonstrated that targeted deletion of Hif1a in donor T cells or systemic inhibition of HIF1α with echinomycin significantly ameliorated GVHD and reduced mortality. The primary mechanism was identified through the IFNγ/PD-L1 axis, in that T cells produced higher levels of IFNγ following HIF1α inhibition, leading to PD-L1 expression in host tissues. Furthermore, neutralizing IFNγ or blocking PD-L1/PD-1 interaction abrogated the beneficial effect of HIF1α inhibition in GVHD prevention (Figure 1). Importantly, targeting HIF1α prevented GVHD without causing a negative impact on the GVL effect. Three potential mechanisms for the GVL maintenance upon targeting HIF1α were presented: (1) relatively preserved expansion of CD8 T cells and an increased CD8/CD4 T cell ratio; (2) enhanced PD-L1 expression on CD8 T cells, resulting in CD80 interaction and expansion of T cells; and (3) decreased PD-L1 expression on leukemia cells, causing attenuated immunosuppression. The other important finding is that targeting HIF1α allowed for anti-CTLA4 treatment to augment GVL effects without exacerbating GVHD.¹ While blocking CTLA4 exacerbated GVHD severity by increasing T cell infiltration in GVHD target organs, this adverse effect was largely abrogated when Hif1a was deficient in donor T cells or HIF1α was inhibited with echinomycin, which was likely attributed to elevated PD-L1 in these target organs, causing inhibition of the infiltrated donor T cells. On the other hand, blocking CTLA4 enhanced the expansion and activation of donor T cells presumably in hematologic and lymphatic compartments.

Figure 1.

Inhibition of HIF1α diminishes GVHD while enhancing GVL

While the study is exciting and of clinical significance, additional questions remain to be addressed. Given that HIF1α is a common TF and plays key roles in multiple immune cells including T cells, dendritic cells, macrophages, and neutrophils that contribute to GVHD pathogenesis, how inhibition of HIF1α pharmacologically impacts GVHD development through other immune cells should be considered and evaluated. Moreover, Seike et al. recently reported that HIF1α protein levels were reduced in the intestinal epithelial cells (IECs) from allogeneic when compared with syngeneic recipients after allo-HCT, and furthermore Hif1a-deficiency, specifically in IECs, resulted in greater GVHD severity and mortality.¹⁰ Thus, the potential negative effect on gut epithelia by HIF1α inhibition should also be taken into concern. Interestingly, the authors observed that inhibition of HIF1α promoted PD-L1 expression of donor T cells and host tissues but reduced PD-L1 expression on leukemia cells, which resulted in a favorable balance of GVH/GVL response. However, the underlying mechanisms by which HIF1α distinctly regulates PD-L1 expression on different cell types are not delineated. In addition, HIF1α has been shown to promote Th17 differentiation while limiting Treg development,⁶ which was not validated in the current study. Finally, while previous studies indicated that inhibition or downregulation of HIF1α reduced IFNγ production by donor T cells,^8,9 the current study showed that targeting HIF1α genetically or pharmacologically enhanced IFNγ production by donor T cells. Such a discrepancy deserves further investigation and reconciliation, given that IFNγ/PD-L1 was demonstrated as a key pathway in regulating GVHD.¹

In summary, the study by Bailey et al. demonstrates that targeting HIF1α confers GVHD protection without affecting GVL effect and that HIF1α inhibition allows ICI in combination with allo-HCT to improve therapeutic efficacy against hematologic malignances while limiting GVHD development.¹