Hepatic SOS, formerly referred to as veno-occlusive disease, develops in up to 10% of patients undergoing stem cell transplantation, a substantial percentage of whom succumb to this disorder.1 A number of therapeutic approaches have failed to significantly alter the relentless course of SOS, though recent evidence suggests that defibrotide ameliorates SOS and may improve survival. In this issue of Blood, Benimetskaya and colleagues characterize the interactions of defibrotide with endothelial cells, providing new insight into potential mechanisms underlying its efficacy in SOS.
Sinusoidal obstruction syndrome (SOS) is initiated by exposure to naturally toxic pyrrolizidine alkaloids, liver irradiation, or conventional chemotherapy. More commonly, however, SOS occurs after high-dose chemotherapy and hematopoietic stem cell infusion,1-3 especially after prior exposure to the immunoconjugate gemtuzumab ozogamicin (Mylotarg). Patients with SOS generally present with tender hepatomegaly, jaundice and ascites, or unexplained weight gain, most often within the first 3 weeks after a hematopoietic stem cell transplant.4 Attempts to treat SOS with vigorous supportive care, systemic anticoagulation, thrombolytic therapy, and/or surgical shunting have not proven effective.5 Recent reports, however, suggest that defibrotide, a mixture of porcine-derived phosphodiester oligonucleotides, has significant efficacy in the treatment of SOS. This investigational agent is now being used with increased frequency in the active treatment setting, as well as in prophylaxis in high-risk stem cell transplant situations.1 However, there is little information available concerning the cellular mechanisms that account for the activity of defibrotide in SOS.
The pathogenesis of SOS appears to reflect direct insult to hepatic sinusoidal endothelial cells.1 In an animal model of SOS prepared by treating Sprague-Dawley rats with monocrotaline, the earliest morphologic changes included loss of fenestration of sinusoidal endothelial cells and gaps in the sinusoidal endothelial cell barrier.6 Subsequently, endothelial cells rounded up, red blood cells penetrated into the space of Disse beneath the damaged endothelium, and the sinusoidal lining cells (endothelium, Kuppfer cells, and stellate cells) were sloughed and embolized distally, resulting in obstruction of sinusoidal flow. In the rat model, SOS is ameliorated by concomitant administration of glutathione, which prevents endothelial cell rounding and sloughing of the sinusoidal lining, possibly by inhibiting matrix metalloproteases released by endothelial cells following monocrotaline-induced depolymerization of endothelial actin.6
The article by Benimetskaya et al provides new information concerning the interactions of defibrotide with endothelial cells. These investigators demonstrate that defibrotide, as well as a series of well-defined phosphodiester oligonucleotides, bind to heparin-binding proteins, in particular bFGF, but not VEGF-165. Once bound by defibrotide, bFGF retains its ability to bind FGFR1c with high affinity and stimulate endothelial cell mitogenesis. Defibrotide also mobilizes bFGF from storage sites in the endothelial matrix and protects bFGF from degradation by trypsin and chymotrypsin as well as air oxidation. Finally, defibrotide binds collagen I with nanomolar affinity and promotes endothelial tubular morphogenesis in 3-dimensional collagen I gels, perhaps through enhancing either α2 or β1 integrin interactions with collagen I. Taken together, these effects would clearly favor angiogenesis, and the authors hypothesize that the efficacy of defibrotide in SOS may be related to its ability to promote revascularization of an injured, hypoxic hepatic parenchyma.
By defining in vitro interactions of defibrotide with endothelial cells, this report provides clues to the pathophysiology of SOS as well as to the potential therapeutic mechanisms of defibrotide. However, additional work is required to validate these mechanisms in the in vivo setting, which is far more complex than a cell culture system. For example, though the authors hypothesize that the activity of defibrotide results from its proangiogenic activity and revascularization of the hepatic parenchyma after angiotoxic injury, it seems plausible that defibrotide may also directly protect sinusoidal endothelium from toxin-induced apoptosis or necrosis through activation of direct or indirect (induction of local VEGF release) prosurvival pathways.7 In the rat model of SOS, it is hypothesized that the protective activity of glutathione results from inhibition of matrix metalloproteases,6 although one might wonder whether the angiotoxic or proaptotic effect of monocrotaline involves oxidant stress pathways, which may be counteracted by glutathione or perhaps even defibrotide. Could defibrotide also protect bFGF or other endothelial growth and survival factors from degradation by metalloproteases, which may have greater pathophysiologic importance in SOS than trypsin or chymotrypsin?
Benimetskaya et al have raised additional questions and provided important insight into the pathogenesis of SOS by describing novel interactions of defibrotide with endothelial cells. It is hoped that these studies will stimulate additional exploration in attempts to validate the authors' observations in vitro.
Conflict-of-interest disclosure: The authors declare no competing financial interests. ■
