Targeting neovascularization in GVHD

In this issue of Blood, Leonhardt et al¹  report that neovascularization during graft-versus-host disease (GVHD) is regulated by α_v integrins and the micro RNA miR-100.

Judah Folkman, the visionary physician-scientist who ultimately convinced doubting colleagues that neovascularization was critical to the development and metastasis of human cancers, often repeated the mantra, “Science goes where you imagine it.”²  Despite the often imaginative research of scientists who have devoted their lives to the study of mechanisms of GVHD following allogeneic hematopoietic stem cell transplantation (HSCT), meaningful therapeutic advances have so far been sparse, and the primary strategies used to prevent and treat GVHD in most centers have evolved little from the approaches advanced by Thomas, Storb, and colleagues more than a quarter century ago.³  The current study suggests that targeting of neovascularization might yield a new class of therapies aimed not only at effector T cells, the primary targets of therapies advanced to date.

As elegantly reviewed in Blood,⁴  studies of the role of vascular proliferation in GVHD were conducted surprisingly early in the course of the development of experimental and clinical HSCT. Indeed, pioneering transplant immunologists Brent and Medawar⁵  described in 1966 the “bright and scarlet reaction” of inflammation in GVHD that was clearly dependent on local microvascular changes. By the mid-1970s, Sidky and Auerbach⁶  quite clearly described a phenomenon that they termed “lymphocyte-induced angiogenesis” that occurred in the context of graft-versus-host reactions, demonstrating a direct correlation between the numbers of alloreactive (but not syngeneic) lymphocytes transferred and the extent of neovascularization observed. Less than 10 years later, studies including that by Cohen et al⁷  had identified links between lymphocytes, soluble cytokines, and the migration of endothelial cells, drawing together the processes of alloreactivity and vascularization, although cytokines were shown to be capable of negative, as well as positive, regulation of vessel formation. Despite these promising early demonstrations that neovascularization was closely associated with alloreactivity in vivo, there has been a relatively slow progression of research in this field, until relatively recently.

Fortunately, transplant immunologists have rekindled the hypothesis that targeting of neovascularization might yield successful therapies for GVHD. Recent studies have yielded mechanistic insights as well as potentially novel therapeutic strategies. Earlier, Penack et al,⁸  working in a murine HSCT model of GVHD, demonstrated that neovascularization, mediated by vasculogenesis, contributed to both GVHD and tumor formation following transplantation and could be inhibited using antibody-based therapy targeted at a cadherin superfamily member. In the current study, Leonhardt et al¹  extend our mechanistic understanding of neovascularization in GVHD and provide proof of concept regarding therapeutic targeting of this process.

The authors focused their studies on alloreactivity in the intestinal tract, demonstrating that gut GVHD is associated with neovascularization, including increased expression of vascular growth factors and of α_v integrin expressed on endothelial cells.¹  They demonstrated that cilengitide, a pentapeptide drug targeting α_v integrin, reduced neovascularization and inhibited GVHD severity. Preliminary studies of human subjects demonstrated a positive correlation between the severity of intestinal GVHD and endothelial cell marker staining, as well as the level of α_v integrin expression.¹ 

The authors then examined the expression of micro RNAs associated with neovascularization and determined that the expression of miR-100, a previously reported negative regulator of neovascularization, was downregulated in untreated mice, relative to levels in animals developing GVHD. Treatment with an miR-100–blocking antagomir reduced survival and increased neovascularization in the gut. As predicted, combined treatment with cilengitide partially ameliorated the effects of miR-100 blockade, resulting in significantly improved survival.¹ 

Overall, this study successfully extends our understanding of the interconnections between inflammation and neovascularization and suggests that targeting mediators of vessel formation may yield novel therapies for GVHD. Although this study will likely inspire further fruitful research, significant questions remain unanswered. Our understanding of the initial triggers of neovascularization, including the relationship between tissue injury, effector cytokines, and vasculogenesis, remains incomplete. Although the authors present convincing results related to intestinal GVHD, less is known regarding the importance of these mechanisms in other target organs (eg, the liver, skin, and hematolymphoid tissues, each a unique ecosystem). Although cilengitide demonstrated significant therapeutic promise in this model, the broad expression of α_v integrin (on T and B cells, natural killer cells, and dendritic cells) suggests that targeting might result in significant and potentially undesirable off-target effects on desirable T-cell responses to pathogens and cancers, especially during post-HSCT immune reconstitution. We previously demonstrated that engagement of α₄β₁ integrin, a surface receptor similarly implicated in neovascularization,⁹  may also play an important role in T-cell costimulation.¹⁰  The early studies of Cohen and others should remind us of the potentially complex interactions between cellular mediators of inflammation and potentially diverse effects on neovascularization.⁷ 

Despite these caveats, we can hope that targeting of neovascularization might facilitate the development of combination therapies targeting both cellular effectors (eg, T cells) and vessel formation in target tissues. There is also legitimate cause for optimism that such approaches might also limit disease recurrence,⁸  given the now uncontroversial importance of neovascularization as a mediator of relapse of many cancers.

Ultimately, to develop meaningful therapeutic advances in prevention and treatment, we will need to continue to follow the late Dr Folkman’s maxim and pursue imaginative science that uncovers the synergistic mechanisms leading to GVHD. Imagination, as demonstrated here, will first yield possibilities and, eventually, therapeutic success.