Sildenafil treatment to reduce endothelial damage in aGvHD and SR-aGvHD

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is the only curative treatment option for many patients with hematological disorders. However, acute graft-versus-host disease (aGvHD) is a common and severe complication associated with allo-HSCT. Although steroid treatment is mostly successful in these patients, approximately 20% to 25% of patients fail the initial treatment with steroids and are referred to as having steroid-refractory aGvHD (SR-aGvHD). These patients have a high mortality rate and currently there is no standard treatment available. Endothelial dysfunction and damage are thought to be associated with the development and severity of aGvHD and SR-aGvHD, however, the biology of SR-aGvHD is still poorly understood and the development of novel therapies has been stagnant. Therefore, Steffen Cordes et al published in Haematologica a study that analyzed the association of endothelial dysfunction and aGvHD using tissue biopsies. Furthermore, they tested the protective effect of the PDE5 inhibitor sildenafil, also known as Viagra, on endothelial cells and its impact on target organ damage and survival in a mouse model.¹

Methods

Human samples

The following human samples were collected from patients treated at Charité Berlin and Hannover Medical School:

• Samples from Charité Berlin included 12 duodenal and 11 colonic biopsies from patients with intestinal aGvHD Grade III–IV after allo-HSCT
• Samples from the Hannover Medical School contained colon biopsies from 11 patients with aGvHD
• Biopsies were taken at two time points from each patient, once at the diagnosis of aGvHD and later at the diagnosis of SR-aGvHD
• Nineteen duodenal and 10 colonic biopsies from recipients of allo-HSCT without histological evidence of aGvHD were used as a control

Mouse models

Two types of murine models were used to mimic SR-aGvHD:

• The chemotherapy-based murine models 129→B6 MHC-matched and B6→B6D2F1 (haploidentical)
• The radiation-based murine model BALB/c→B6 major MHC-mismatch
• All recipient mice were treated with the same conditioning and dosage: intraperitoneal administration of 0.5 mg/kg/day dexamethasone beginning at Day 4 after allo-HSCT
• Control mice (i.e., no aGvHD group) were transplanted with T cells and BM cells from syngeneic donors

Testing

• Caspase 3 (Casp3⁺) staining as an apoptotic cell marker to quantify endothelial apoptosis
• Transmission electron microscopy to investigate microstructural changes of the endothelium
• Fluorescence microscopy and immunostaining for ZO-1 (tight junction protein) and VE-cadherin (intercellular junction protein) to assess alterations in the endothelial barrier function
• Light sheet fluorescence microscopy of whole colons labelled with anti-VE-cadherin antibodies to observe the three-dimensional changes of the vasculature
• Myographic measurements to examine vessel contraction as a readout of macrovascular function
• Gene array expression analysis by flow cytometry to observe the molecular pathways involved in endothelial dysfunction during aGvHD
• Characterization of inflammatory infiltrates and endothelial damage during SR-aGvHD
• Administration of sildenafil both in vitro and in vivo to assess the effect of pharmacologic protection of the endothelium on SR-aGvHD vs treatment of aGvHD

Results

Intestinal biopsies of patients diagnosed with intestinal Grade III–IV aGvHD had a high frequency of endothelial apoptosis, as quantified by Casp3⁺ endothelial cells. There were similar levels of endothelial apoptosis seen in SR-aGvHD biopsies. In contrast, low levels of endothelial apoptosis were observed in biopsies of patients who had allo-HSCT without aGvHD. This data suggests that endothelial apoptosis occurs during severe intestinal aGvHD and SR-aGvHD.  

At the microstructural level, electron microscopy analysis revealed severe damage and blistering in the hepatic sinusoidal endothelium in biopsies from mice with aGvHD. In contrast, the hepatic sinusoidal endothelium was not affected in control mice without aGvHD. This data suggests that extensive endothelial damage occurs also in other organs during aGvHD. Loss of tight junctions and endothelial leakage was significantly higher in the liver and colon in mice with aGvHD vs mice without aGvHD.

aGvHD was associated with structural changes in colonic vasculature as demonstrated by a significantly increased vessel branching from vessel segments in mice that developed aGvHD vs control mice. Intestinal aGvHD was also associated with a significantly increased vessel diameter.

Analysis in mouse models showed an increased contraction and reduced relaxation potential of mesenteric arteries during aGvHD compared with mice without aGvHD, suggesting an association of systemic hypertension and aGvHD.

There were significant endothelial gene expression changes in many clinically relevant pathways from biopsies of mice with aGvHD, including apoptosis, complement activation, oxidative damage, IL-1 signalling, and cell cycle pathways. A mouse model of SR-aGvHD and human biopsies from patients with SR-aGvHD demonstrated reduced inflammatory T-cell infiltrates after steroid treatment, compared with untreated aGvHD.

Sildenafil treatment

• In cell culture models, sildenafil protected endothelial cells from etoposide-induced apoptosis and etoposide-induced reduction of endothelial metabolic activity and proliferation in vitro
• Significantly lower clinical and histopathological aGvHD scores were observed in sildenafil-treated mice with aGvHD vs untreated mice with aGvHD. However, sildenafil treatment showed only a non-significant trend for improved survival in mice with aGvHD
• In a SR-aGvHD mouse model, sildenafil treatment significantly improved survival, with significantly improved clinical aGvHD scores in the early phase of the disease. The impact of sildenafil on the late phase of the disease was difficult to assess due to the high mortality
• A non-significant trend of increased vascular density and lymphatic vascular density was observed in sildenafil-treated mice with SR-GvHD compared with the control group
• In addition, the sildenafil treated mice with SR-aGvHD showed less endothelial damage in the liver and reduced fibrinogen deposits in colon biopsies compared with the control group
• Sildenafil treatment had no significant effects on T-cell proliferation in SR-aGvHD mice, suggesting that the positive effect of sildenafil on SR-aGvHD is mostly mediated by its effect on endothelial cells rather than on immune cells

Conclusion

In conclusion, the results show the extensive vasculature changes and endothelial damage that occurs during intestinal aGvHD in humans and mice, and the beneficial effect of sildenafil in mice with aGvHD and SR-aGvHD. These findings suggest that endothelium-protecting agents, such as sildenafil, may constitute a promising treatment approach for SR-aGvHD. Further studies are needed to test whether endothelium-targeting approaches can complement the established treatment options in SR-aGvHD, such as JAK-2 inhibitors.