Mini-Review Avoiding Hepatic Veno-Occlusive Disease: What Do We Know and Where Are We Going?

Home , Defibrotide

Bone Marrow Transplantation (2001) 27, 1113–1120  2001 Nature Publishing Group All rights reserved 0268–3369/01 $15.00 www.nature.com/bmt Mini-review Avoiding hepatic veno-occlusive disease: what do we know and where are we going?

SI Bearman

Bone Marrow Transplant Program, University of Colorado Health Sciences Center, Denver, CO, USA

Summary: example, if on day 1 a patient’s bilirubin is 5 mg/dl and his weight has increased 5% from baseline, z would equal Hepatic venocclusive disease (V0D) is a common toxicity Ϫ2.36 (␤ on days Ϫ1toϩ1) ϩ [0.79 associated with myeloablative chemotherapy or chemor- 0 ␤ Ϫ ϩ ϫ adiotherapy used to prepare patients for stem cell trans- ( 1 on days 1to 1) ln5] plantation. A sizable proportion of patients who develop ϩ [0.14 (␤ on days Ϫ1toϩ1) ϫ 5] VOD die. It is clear that injury to endothelial cells and 2 hepatocytes in zone 3 of the liver acinus is the initial or Ϫ0.39. The probability of developing severe VOD event in the pathogenesis of VOD. What are less clear would therefore be 1/[1 ϩ eϪ(Ϫ0.39)] or 0.40. This model are the mechanisms of injury and whether this knowl- has been shown to be highly specific but only moderately edge can be exploited. This manuscript will briefly sensitive. Therefore, it is suitable for use to screen patients review some recent data and discuss how that infor- for therapeutic trials of investigational or potentially toxic mation has influenced clinical practice. Bone Marrow agents. Richardson and colleagues6 used the model to deter- Transplantation (2001) 27, 1113–1120. mine eligibility for treatment with defibrotide. Patients Keywords: veno-occlusive disease; stem cell transplan- whose probability of developing fatal VOD was 40% or tation; toxicity greater were eligible for that study. Pathologic changes of VOD are centered in zone 3 of the liver acinus. Hepatic venular occlusion is the pathogno- monic change described in the literature, although patients VOD is characterized by a syndrome of jaundice, fluid without this finding can develop the clinical syndrome. retention and painful hepatomegaly and occurs in up to half Other pathologic changes include eccentric venular luminal 1 of patients who receive ablative preparative therapy. As narrowing, phlebosclerosis, sinusoidal fibrosis, and necrosis many as half of patients with VOD die, usually of multior- of hepatocytes.7 The severity of the clinical illness is gan failure. Signs of VOD usually present around the day strongly related to the number of pathologic changes in 2 of transplant, although well described ‘late onset VOD’ zone 3. Elevated hepatic venous pressure gradients also cor- 3,4 has been reported. Fluid retention and hepatomegaly are relate with histologic features of VOD.7 usually the first manifestations and typically present on or Increased platelet transfusion requirement is usually around the day of transplant. Jaundice develops later, associated with VOD. Gordon et al8 retrospectively 2 around day 6. The rate of rise in bilirubin and weight is reviewed platelet requirements in 199 patients who also faster for patients whose illness is likely to be severe underwent transplant at the University of Nebraska. They vs those who are likely to have self-limited disease. Further- found that patients with dysfunction of a single organ more, fewer than 20% of patients with mild VOD and 48% (VOD, pulmonary or CNS) received more platelet trans- 2 with severe VOD develop ascites. These observations led fusions per week compared to patients without organ dys- to the development of a model function. Furthermore, an increased number of platelet p ϭ 1/(1 ϩ eϪz) transfusions during any 1 week was predictive for development of organ toxicity, including VOD, in the subsequent which would predict which patients with VOD were likely week. Thrombocytopenia in VOD patients is not the result 5 ϭ ␤ ϩ ␤ to die. In this model, z 0 1 (in total serum bilirubin) of a deficiency in thrombopoietin.9 ϩ ␤ ␤ ␤ ␤ 2 (percent weight gain). The intercepts 0, 1 and 2 change as day relative to transplant increases.5 For Pathogenesis, prevention and treatment

Research in VOD pathogenesis has concentrated in several Correspondence: Dr SI Bearman, Bone Marrow Transplant Program, Uni- versity of Colorado Health Sciences Center, 4200 East Ninth Avenue, Box areas: glutathione depletion, inﬂammatory mediators, B-190, Denver, CO 80262, USA coagulation, and ﬁbrogenesis. These areas of investigation Received 9 February 2001; accepted 8 February 2001 frequently overlap. Not surprisingly, prevention (Table 1) Avoiding hepatic veno-occlusive disease SI Bearman 1114 Table 1 Approaches to prevention of VOD

Strategy Mechanism Comments

Heparin Prevents deposition of clotting material Only one randomized trial performed. Patients who received heparin had in terminal hepatic venules less VOD than those who did not receive heparin.34 However, severe VOD was not different in the two groups. In addition, few patients were at high risk prior to transplant. In a single arm study of high risk patients, heparin was ineffective.32 A randomized trial with low molecular weight heparin showed that the duration of signs of VOD (hyperbilirubinemia, hepatomegaly) was less in LMWH-treated patients.36 Outcome between the two groups was not reported. Ursodiol ?downregulates inﬂammatory mediators Two randomized trials have reported a lower incidence of VOD in ursodiol-treated patients. Survival was not different between the groups.72,73 N-acetylglucosamine precursor of glutathione Very few patients treated. l-glutamine precursor of glutathione Very few patients treated.

and treatment (Table 2) studies have been based on these serine and methionine. However, when propargylglycine, observations. an inhibitor of glutathione synthesis was present, this pro- tective effect was abolished. These studies suggest that Glutathione increased exposure to the toxic metabolites of cyclophosphamide contribute to development of VOD and that sup- Zone 3 of the liver acinus contains the greatest concen- porting hepatocyte glutathione levels might prevent devel- tration of cytochrome P450 and glutathione-S-transferase opment of or reverse VOD. Recently, DeLeve and enzymes.10,11 P450 enzymes metabolize many chemothera- colleagues demonstrated that infusion of reduced gluta- peutic agents while the glutathione-S-transferase enzymes thione could protect rats from VOD caused by monocrota- catalyze the reaction of glutathione with electrophilic com- line, a pyrrolizidine alkaloid.14 pounds. There is strong evidence that depletion of gluta- Three patients with VOD have been treated with N-ace- thione results in hepatocyte necrosis.12 DeLeve13 studied tylcysteine, a thiol antioxidant which is a precursor of intra- the effect of cyclophosphamide on sinusoidal endothelial cellular reduced glutathione synthesis.15 The diagnosis was cells (SEC) and hepatocytes in culture. She reported that made on days 10, 14, and 18 post transplant. All had at direct exposure of SEC to cyclophosphamide did not result least a 10% weight gain and moderate hyperbilirubinemia. in toxicity while exposure to the cyclophosphamide metab- One had signiﬁcant ascites, pleural effusions, and encepha- olites acrolein or 4-hydroxycyclophosphamide resulted in a lopathy. Tissue plasminogen activator was given to all three dramatic dose-dependent toxicity. When SEC and hepato- patients for one to four doses. NAC was also given for 12– cytes were cultured together in the presence of cyclophos- 31 days. All patients recovered normal hepatic function and phamide, SEC experienced signiﬁcantly more toxicity than were alive 4, 8, and 9 months after transplant. Since all did hepatocytes. DeLeve postulated that toxicity to SEC patients had a low probability of fatal VOD based on the was due to acrolein generated by the metabolic activation model discussed above,5 the relative contributions of of cyclophosphamide by hepatocytes. This effect could be tPA and NAC to the resolution of VOD cannot be fully reversed by sustaining levels of hepatocyte glutathione with ascertained.

Table 2 Approaches to treatment of VOD

Therapy Efﬁcacy Comments

Tissue plasminogen activator 23–27% Life-threatening hemorrhage in 20–30% of patients. Ineffective in patients with multiorgan dysfunction. We no longer use tPA because efficacy is not greater than risk. Antithrombin III ෂ45% Only 13 patients treated. All patients had some improvement. Defibrotide 35% No significant hemorrhagic complications. Has been effective in patients with multiorgan failure.6,41 High-dose corticosteroids 61% Experience using steroids is very limited experience. The patients treated by Khoury et al54 were very early in the course of disease, with low probabilities of dying of VOD. TIPS ෂ55% Experience is very limited. Although successful in the majority of patients, most ultimately die of other problems. Should probably be reserved for patients with intractable portal hypertension who have large ascites and are unable to mobilize fluid. Persistent encephalopathy may be a sequelae.

Bone Marrow Transplantation Avoiding hepatic veno-occlusive disease SI Bearman 1115 Like NAC, glutamine is utilized by hepatocytes to those who did not. A multivariate analysis, which also replenish intracellular glutathione.16 Two patients with included male patients, showed norethisterone to be an VOD were treated using l-glutamine (in dipeptide form) independent predictor of VOD. plus vitamin E.17 Although VOD resolved in both patients, The studies above led to trials of various agents with the role of l-glutamine is difficult to determine. One of the anticoagulant activity in patients with VOD or to prevent patients had late-onset VOD, which is not considered in the VOD. The earliest studies, with perhaps the most disagreed above model.5 The second patient had VOD earlier in the upon results, used heparin. My colleagues and I adminis- post-transplant course with a more fulminant presentation tered heparin by continuous infusion to 28 patients at high and may, indeed, have benefited from therapy. risk for developing VOD.32 Heparin was dosed to result The same group also studied l-glutamine as prophylaxis in varying degrees of prolongation of the activated partial of liver toxicity after stem cell transplantation.18 Eighteen thromboplastin time. VOD occurred in 70% of patients, patients were randomized to receive l-glutamine while 16 four of whom developed severe VOD. The incidence and received non-essential amino acids. VOD did not occur in severity of VOD was not influenced by the degree or dur- any patients. Thrombin-antithrombin (TAT), plasmin-␣2- ation of anticoagulation. Two patients at the highest dose antiplasmin (PAP), prothrombin fragment F1ϩ2, protein C level of heparin developed significant, but non-fatal and albumin were measured in all patients. The authors hemorrhagic complications. Lack of benefit from heparin reported that mean protein C and albumin levels were sig- was also reported by Marsa-Vila et al.33 nificantly higher in the glutamine group on day 4 post trans- On the other hand, there have been several studies sup- plant and suggest that glutamine may work by preserving porting the use of heparin to prevent VOD. Attal and col- protein C levels. However, since none of the patients not leagues34 conducted the only prospective randomized trial receiving glutamine developed VOD, despite lower protein of heparin for VOD prophylaxis. They reported a 13.7% C levels, the benefit of glutamine in VOD prophylaxis incidence of VOD in the non-treated group compared to remains questionable. 2.5% in the heparin-treated group (13.7% vs 2.5%). Severe VOD was not different between the two groups. Rosenthal 35 Coagulation parameters et al conducted a retrospective analysis of heparin prophylaxis in children. They found that heparin given from In 1987, Shulman et al19 described the deposition of the start of preparative therapy until day ϩ30 (or discharge) fibrinogen and factor VIII in vessel walls at the interface resulted in significantly less VOD than historical controls of hepatic sinusoids and terminal hepatic venules. This (10% vs 28%). Moderate to severe VOD occurred in 4% observation provided the rationale for studying the role of of study patients and 9% of control patients. the clotting system in the pathogenesis of VOD. A number More recently, Or et al36 conducted a prospective ran- of investigators have found evidence consistent with hyper- domized trial of low molecular weight (LMW) heparin vs coagulable state at the start of the preparative regimen or placebo to prevent VOD. They reported a reduction in the in the early post-transplant period. These include duration of hyperbilirubinemia and the incidence of hepato- deficiencies of protein C, antithrombin III, protein S and megaly in the LMW heparin-treated patients. They also elevated levels of D-dimer and prothrombin fragment reported that hemorrhagic complications were fewer, time F1ϩ2.20–27 Why this is so is unclear. to platelet recovery was shorter, and platelet transfusion Deficiencies in protein C and antithrombin III (ATIII) in requirements were less in the LMW heparin-treated patients with VOD (or organ toxicity, in general) have been patients. The degree of hyperbilirubinemia, fluid retention reported by numerous investigators.20–24 Faioni et al20 and outcome was not reported. reported that baseline levels of protein C were lower in The efficacy of heparin remains in dispute. While all patients who developed VOD and continued to decrease agree that low doses of heparin or LMW heparin can be post transplant. Tanikawa et al21 and Gordon et al22 safely administered in the transplant setting, most patients reported that protein C levels decreased post transplant in reported in the literature have a low likelihood of patients who developed VOD. The University of Nebraska developing VOD based on pretransplant risk factors. The group also reported that ATIII levels fell after high-dose only study where most patients were at high risk showed chemotherapy.22 They also showed that patients who no benefit.32 I do not believe heparin is efficacious for developed organ dysfunction had lower levels of ATIII than prevention of VOD. those without organ dysfunction. Low levels of protein C Several agents with antithrombotic activity have been and ATIII post transplant both correlated with death.23 studied. These include recombinant human tissue plasmin- There are several reports which suggested that VOD was ogen activator, antithrombin III, and defibrotide. While more common in females.28,29 One possible explanation for recombinant tissue plasminogen activator (rh-tPA) has been this is the use of progestins to minimize menstrual bleeding used in many patients, only two series have included sub- during myelosuppression. Progestins have been associated stantial numbers of patients.37,38 We reported that 29% of with an increased risk of thromboembolic events, in part patients responded and that 24% developed life-threatening due to acquired protein C resistance.30 Ha¨gglund and col- hemorrhage. In Scriber’s series,38 most patients began treat- leagues31 recently evaluated whether norethisterone admin- ment when they were suspected to be developing VOD istration might be associated with VOD. Fifty-five women rather than after established criteria were met. Of 24 transplanted between January 1990 and June 1995 received patients who were treated without having met criteria for norethistherone and 38 did not. The incidence of VOD was VOD, 67% responded. The remaining 13 patients, who 27% in women who received norethisterone and 3% in began treatment with established VOD, had a response rate

Bone Marrow Transplantation Avoiding hepatic veno-occlusive disease SI Bearman 1116 of 23%. They, too, reported a very high incidence (35%) phase II studies. Of 103 patients transplanted, six developed of clinically significant hemorrhage. We have ceased using VOD, which was fatal in two. The first dose mean busulfan rh-tPA because the risk of severe hemorrhage is about the AUC was 1156 ␮mol/min for all patients and 1302 same as the response rate. Treating patients with existing ␮mol/min for the six VOD patients. The AUC observed in multiorgan dysfunction has been shown to be ineffective. this trial are sufficient to ensure engraftment (known to be Antithrombin III has been administered to a small num- poor when drug exposure is low)45 and to minimize the risk ber of patients with VOD.39,40 Compared to historical con- of VOD. trols not treated with ATIII, death due to VOD or multiorgan failure was significantly decreased. Platelet Inflammatory mediators consumption, painful hepatomegaly and fluid retention resolved in the majority of patients. Alterations in plasma levels of several inflammatory Defibrotide (DF) is a single-stranded polydeoxyribonu- mediators have been reported in patients with VOD. Holler cleotide with anti-ischemic, antithrombotic and throm- et al50 first reported in 1990 that patients with major com- bolytic activity, but no significant anticoagulant effects. plications of transplantation, ie acute graft-versus-host dis- Richardson and coworkers6,41 have studied DF in 75 ease, idiopathic pneumonia, endothelial leak syndrome and patients with severe VOD. Eighty-four percent of patients VOD, had significant elevations in TNF-␣ levels after had ascites and 92% had evidence of multiorgan dysfunc- transplant which often preceded the clinical complication. tion. Of 67 patients treated for 4 or more days, the complete Holler’s initial observation was the impetus for studying response rate was 35%. There was no significant toxicity. pentoxifylline in regimen-related toxicities, which was ulti- Patients younger than 18 years of age and those with dys- mately found to be ineffective.51 Other data for cytokine function in no more than one additional organ system did activation in VOD include increased expression of genes better. Since responses were greater in patients prepared for interleukin-1␤ and TNF-␣ in peripheral blood mono- with cyclophosphamide-based regimens, the authors postu- nuclear cells of patients who developed VOD,52 as well as lated that DF worked by repairing the endothelial damage increased levels of circulating IL-2 receptor.53 Khoury et known to occur as a result of high-dose cyclophosphamide. al54 studied whether high doses of corticosteroids given at the earliest evidence of VOD or liver dysfunction of Busulfan pharmacokinetics unknown etiology alters the course of hepatic regimen- related toxicity. Twenty-eight consecutive patients who met The importance of busulfan in VOD pathogenesis has been Seattle criteria for VOD or liver dysfunction of unknown the subject of intense investigation. Several investigators etiology were treated with 500 mg/m2 of methylpredniso- have reported that a higher busulfan area-under-the-curve lone every 6 h for six doses. Seventeen patients (61%) of concentration vs time (AUC) or steady-state concen- responded as defined by a decrease in total bilirubin within tration is associated with an increased risk of developing 10 days of treatment. The only distinguishing factor VOD42–45 and that adjusting the dosage based on first dose between responders and nonresponders was that non- pharmacokinetics can result in a decrease in VOD.45,46 responders had a lower pretransplant DLCO. Virtually all patients prepared for transplant using busulfan receive a combination of busulfan and cyclophospham- Generalized endothelial injury ide. Slattery et al47 found that the average plasma steady- state concentrations of busulfan correlated with exposure Endothelial injury is clearly an early event in the pathogen- to cyclophosphamide. In their study comparing the pharma- esis of VOD. Circulating levels of the endothelial marker cokinetics in seven patients who received busulfan plus thrombomodulin55–57 increase after high-dose cytoreductive cyclophosphamide (BU/CY) with seven patients who therapy. Plasminogen activator inhibitor-1 (PAI-1), another received cyclophosphamide plus TBI (CY/TBI), they found endothelium-derived molecule, also increases after high- that all measures of exposure to hydroxycyclophosphamide dose cytoreductive therapy in patients with VOD.58 Salat were greater in BU/CY patients than in CY/TBI.47 This is and colleagues59 have shown that elevations in PAI-1 are consistent with the results of DeLeve (see Glutathione, specific for VOD and were not significantly elevated from above). Busulfan depletes glutathione, enhancing hydroxy- other causes of liver injury. Successful treatment of VOD cyclophosphamide exposure and increasing toxicity. with DF lowers PAI-1 levels.57 Part of the difficulty in using busulfan is that the absorp- There is also evidence that the vascular endothelium may tion from the gastrointestinal tract is unpredictable and the be injured and activated before the start of conditioning plasma levels, therefore, variable. Andersson et al48 perfor- therapy, presumably due to previous chemotherapy. Rich- med a phase I trial of a single dose of intravenous busulfan ard and colleagues60 measured the endothelial cell markers in patients being prepared for transplant with oral busulfan von Willebrand factor (vWF), soluble thrombomodulin, plus cyclophosphamide (15 additional doses). When com- soluble intercellular adhesion molecule-1 (ICAM-1), and pared with the bioavailability of the single dose of i.v. bus- their inducers TNF-␣ and elastase in the plasma of patients ulfan, the bioavailability of oral busulfan ranged from 10% prior to and 1 and 3 weeks after bone marrow transplant. to 100% (average 69%). The time to maximum plasma con- They observed increases in ICAM-1 and vWF after trans- centration and the clearance of the i.v. preparation was plant but no increases in thrombomodulin or TNF-␣. How- similar to that of oral busulfan. The optimal dose of i.v. ever, levels of thrombomodulin and vWF were increased busulfan was determined to be 0.8 mg/kg. Vaughan and at baseline, suggesting that endothelial injury already colleagues49 reported on the use of i.v. busulfan in two existed and that changes occurring after transplant were

Bone Marrow Transplantation Avoiding hepatic veno-occlusive disease SI Bearman 1117 more consistent with endothelial activation rather than not yet known is whether the TNFd3 allele correlates with further damage. The Seattle group reported that the risk organ dysfunction independently of ATIII and protein C of dying within 6 months of stem cell transplantation was levels. correlated with reduced diffusing capacity of carbon mon- oxide corrected for hemoglobin (DLCO) and increased Other pharmacologic therapy: ursodiol alveolar-arterial oxygen gradient.61 In a subsequent report, they showed that DLCO less than 70% of predicted was Several years ago, Essell and colleagues72 published their independently associated with an increased risk of severe results of a randomized trial of ursodiol for prevention of VOD (odds ratio 2.4).62 They, too, hypothesized that the VOD. Patients were prepared with busulfan and cyclophos- decreased DLCO reflected systemic endothelial damage phamide and received methotrexate plus cyclosporine as from prior therapy. GVHD prophylaxis. VOD occurred in 40% of placebo recipients and 15% of ursodiol recipients. While develop- Fibrogenesis/fibrolysis ment of VOD was the strongest predictor of survival to day 100 post transplant, day 100 survival was not statistically Fibrosis of centrilobular sinusoids is one of the classic his- different between the two groups. More recently, a second tologic features of VOD and is the result of a balance randomized trial of ursodiol was conducted in Japan. VOD between fibrogenesis and the dissolution of fibrosis occurred significantly less often in the ursodiol-treated (fibrolysis). How this comes about in VOD and why, has patients (3%) compared with the control group (18%). been the subject of studies by several groups of investi- Again, no statistically significant difference was observed gators. in survival between the two groups.73 Hepatic stellate cells are the most important mediators There are data which show that bile steroids may prevent of hepatic fibrosis.63 Activation of stellate cells results in endothelial cell74 and hepatocyte injury,75 possibly by induction of extracellular matrix genes, expression of downregulating the expression of inflammatory mediators. smooth muscle-specific ␣ actin, and increases in types III How it may actually work to prevent VOD is unclear. and IV collagen.64 Sato and colleagues65 reported that significantly more stellate cells in zone 3 were activated in Non-pharmacologic therapy: transjugular intrahepatic patients who developed VOD compared to those who did portosystemic shunting (TIPS) not. This results in production of the N-terminal propeptide of type III procollagen (PIIINP), which has been reported to Several groups have reported using TIPS for the treatment be elevated in patients with VOD by several groups,21,66–68 as of VOD. In this procedure, a channel is created between early as day 0.68 Tanikawa et al21 reported that PIIINP lev- the hepatic and portal vein using a percutaneously inserted els were elevated prior to the start of conditioning therapy, catheter and kept patent with a metal stent. A total of 19 again suggesting that hepatic injury and repair processes patients have been reported in the literature.76–79 Significant had occurred or were ongoing at the time high-dose cytore- reduction in hepatic venous pressure gradient was reported ductive therapy began. In this study, however, pretransplant after the procedure. Eight patients had no improvement in ALT and AST were not different between patients who did VOD and died within 2 weeks. Of the remaining 11 or did not develop VOD. It would be interesting to know patients, improvement in ascites, urinary output, and coagu- whether patients whose transaminases are elevated prior to lation parameters was noted. Unfortunately, all but three transplant also have elevated PIIINP levels. died of other complications. There is some suggestion from Other molecules which promote fibrosis have also been the report of Azoulay and colleagues,79 that responding studied in VOD patients. These include tenascin, a marker patients have less severe disease, as demonstrated by lower of fibrogenesis, and tissue inhibitor of metalloproteinase bilirubin and ALT prior to the TIPS and a greater reduction type 1 (TIMP-1), an inhibitor of fibrolysis, which are also in the hepatic venous pressure gradient after the procedure. elevated within 1 week of onset of VOD.69 Thus, hepatic In addition, their patients who died shortly after TIPS had injury both increases fibrogenesis and inhibits fibrolysis. a shorter interval from transplant to TIPS than patients who Unfortunately, how to exploit this information with regard survived more than 10 days, suggesting that the pace of to prevention and treatment is unclear. their disease was faster. These data indicate that TIPS may benefit more than half of patients with VOD. However, they A unifying hypothesis? also indicate that these patients have other life-threatening problems, as well. There are some data which suggest that particular TNF gene polymorphisms may be associated with organ injury. Cavet et al70 reported that patients homozygous for TNFd Conclusion microsatellite allele 3 (TNFd3) had a much greater likelihood of dying by day 30 than patients not homozygous for We have known for a long time that patients with active TNFd3 (24% vs 7%). Haire et al71 followed 42 consecutive liver injury, as indicated by elevated transaminases prior to patients for development of VOD, pulmonary dysfunction the start of preparative therapy, were more likely to develop and CNS dysfunction. The TNFd genotype of these patients severe VOD. Recent data indicate that the endothelial dam- was determined by polymerase chain reaction. Organ tox- age and repair processes are ongoing at the start of con- icity occurred in 57% of patients with at least one TNFd3 ditioning in many patients, as evidenced by elevated base- allele and 17% of patients with no TNFd3 alleles. What is line levels of PIIINP,21 vWF and thrombomodulin60 in

Bone Marrow Transplantation Avoiding hepatic veno-occlusive disease SI Bearman 1118 patients who develop VOD. Prophylaxis might be more Liver: Biology and Pathobiology, 2nd edn. Raven: New York, effective if it were directed towards endothelial protection 1988, pp 931–947. and reduction of fibrosis. Further study of agents like NAC 11 el Mouelhi M, Kauffman FC. Sublobular distribution of trans- or l-glutamine, which would target endothelial protection ferases and hydrolases associated with glucuronide, sulfate by supplementing sources of glutathione, are warranted. and glutathione conjugation in human liver. Hepatol 6: 1986; 450–456. How to stop ongoing fibrosis is unclear. If such prophylac- 12 DeLeve LD. Dacarbazine toxicity in murine liver cells: a tic strategies are ineffective, it is doubtful that treatment model of hepatic endothial injury and glutathione defense. J of VOD will ever be better than the 30–35% reported for Pharm Exp Ther 1994; 286: 1261–1270. defibrotide. Most patients, even if successfully treated, die 13 DeLeve LD. Cellular target of cyclophosphamide toxicity in of other complications. murine liver: role of glutathione and site of metabolic acti- This inability may be less problematic than it appears, vation. Hepatol 1996; 24: 830–837. however. New nonmyeloablative immunosuppressive stra- 14 Wang X, Kanel GC, DeLeve LD. Support of sinusoidal endo- tegies are being actively studied and preliminary reports thelial cell glutathione prevents hepatic veno-occlusive disease using these regimens suggest that VOD occurs less fre- in the rat. Hepatol 2000; 31: 428–434. quently. Conceivably, we may be closer to transplantation 15 Ringden O, Remberger M, Lehmann S et al. N-acetylcysteine for hepatic veno-occlusive disease after allogeneic stem cell with little regimen-related toxicity. Whether we get there transplantation. Bone Marrow Transplant 2000; 25: 993–996. by figuring out how to interfere with VOD pathogenesis or 16 Hong RW, Rounds JD, Helton WS et al. Glutamine preserves by circumventing the problem really does not matter. liver glutathione after lethal hepatic injury. Ann Surg 1992; 215: 114–119. 17 Goringe AP, Brown S, O’Callaghan U et al. Glutamine and Acknowledgements vitamin E in the treatment of hepatic veno-occlusive disease following high-dose chemotherapy. Bone Marrow Transplant 21 Dr Bearman is the recipient of a Midcareer Investigator Award 1998; : 829–832. et al in Patient-Oriented Research (1K24 CA81408) from the National 18 Brown SA, Goringe A, Fegan C . Parenteral glutamine Cancer Institute. protects hepatic function during bone marrow transplantation. Bone Marrow Transplantation 1998; 22: 281–284. 19 Shulman HM, Gown AM, Nugent DJ. Hepatic veno-occlusive disease after bone marrow transplantation. Immunohistochem- References ical identification of the material within occluded central venules. Am J Pathol 1987; 127:549–558. 1 Bearman SI. The syndrome of hepatic veno-occlusive disease 20 Faioni EM, Krachmalnicoff A, Bearman SI et al. Naturally after marrow transplantation. Blood 85: 1995; 3005–3020. occuring anticoagulants and bone marrow transplantation: 2 McDonald GB, Hinds MS, Fisher LD et al. Veno-occlusive Plasma protein C predicts development of venocclusive disease of the liver and multiorgan failure after bone marrow disease of the liver. Blood 1993; 81: 3458–3462. transplantation: a cohort study of 355 patients. Ann Int Med 21 Tanikawa S, Mori S, Ohhashi K et al. Predictive markers for 1993; 118: 255–267. hepatic veno-occlusive disease after hematopoietic stem cell 3 Toh HC, McAfee SL, Sackstein R et al. Late onset veno- transplantation in adults: a prospective single center study. occlusive disease following high-dose chemotherapy and stem Bone Marrow Transplant 2000; 26: 881–886. cell transplantation. Bone Marrow Transplant 1999; 24: 22 Gordon BG, Haire W, Kessinger A et al. High frequency of 891–895. antithrombin 3 and protein C deficiency following autologous 4 Lee JL, Gooley T, Bensinger W et al. Veno-occlusive disease bone marrow transplantation for lymphoma. Bone Marrow of the liver after busulfan, melphalan, and thiotepa condition- Transplant 1991; 8: 497–502. ing therapy: incidence, risk factors, and outcome. Biol Blood 23 Haire WD, Ruby EI, Gordon BG et al. Multiple organ dys- Marrow Transplant 1999; 5: 306–315. function syndrome in bone marrow transplantation. J Am Med 5 Bearman SI, Anderson G, Mori M et al. Venocclusive disease Assoc 1995; 274: 1289–1295. of the liver. Development of a model for predicting fatal out- 24 Lee J-H, Lee K-H, Kim S et al. Relevance of proteins C and come after marrow transplantation. J Clin Oncol 1993; 11: S, antithrombin III, von Willebrand factor, and factor VIII for 1729–1736. the development of hepatic veno-occlusive disease in patients 6 Richardson PG, Elias AD, Krishnan A et al. Treatment of sev- undergoing allogeneic bone marrow transplantation: a pro- ere veno-occlusive disease with defibrotide: compassionate spective study. Bone Marrow Transplant 1998; 22: 883–888. use results in efficacy without significant toxicity in a high 25 Catani L, Gugliotta L, Mattioli Belmonte M et al. Hypercoag- risk population. Blood 1998; 92: 737–744. ulability in patients undergoing autologous or allogeneic BMT 7 Shulman HM, Fisher LB, Schoch HG et al. Venocclusive dis- for hematological malignancies. Bone Marrow Transplant ease of the liver after marrow transplantation: histological cor- 1993; 12: 253–259. relates of clinical signs and symptoms. Hepatol 1994; 19: 26 Harper PL, Jarvis J, Jennings I et al. Changes in the natural 1171–1181. anticoagulants following bone marrow transplantation. Bone 8 Gordon B, Tarantolo S, Ruby E et al. Increased platelet trans- Marrow Transplantation 1990; 5: 39–42. fusion requirement is associated with multiple organ dysfunc- 27 Sudhoff T, Heins M, Sohngen D et al. Plasma levels of D- tions in patients undergoing hematopoietic stem cell transplan- dimer and circulating endothelial adhesion molecules in veno- tation. Bone Marrow Transplant 1998; 22: 999–1003. occlusive disease of the liver following allogeneic bone mar- 9 Oh H, Tahara T, Bouvier M et al. Plasma thrombopoietin lev- row transplantation. Eur J Haematol 1998; 60: 106–111. els in marrow transplant patients with veno-occlusive disease. 28 Ganem G, Saint-Marc Girardin M-F, Kuentz M et al. Ven- Bone Marrow Transplant 1998; 22: 675–679. occlusive disease of the liver after allogeneic bone marrow 10 Gumucio JJ, Chianale J. Liver cell heterogeneity and liver transplantation in man. Int J Radiat Oncol Biol Phys 1998; function. In: Arias IM, Jakoby WP, Popper H et al. (eds). The 14: 879–884.

Bone Marrow Transplantation Avoiding hepatic veno-occlusive disease SI Bearman 1119 29 Nevill TJ, Barnett MJ, Klingemann H-G et al. Regimen- and hepatic veno-occlusive disease in children undergoing related toxicity of a bulsulfan-cyclophosphamide conditioning bone marrow transplantation. Cancer Chemother Pharmacol regimen in 70 patients undergoing allogeneic bone marrow 1996; 37: 247–253. transplantation. J Clin Oncol 1991; 9: 1224–1232. 47 Slattery JT, Kalhorn TF, McDonald GB et al. Conditioning 30 Rosing J, Tans G. Effects of oral contraceptives on hemostasis regimen-dependent disposition of cyclophosphamide and hyd- and thrombosis. Am J Obstet Gynecol 180: S375–S382. roxycyclophosphamide in human marrow transplantation 31 Ha¨gglund H, Remberger M, Klaesson S et al. Norethisterone patients. J Clin Oncol 1996; 14: 1484–1494. treatment, a major risk-factor for veno-occlusive disease in the 48 Andersson BS, Madden T, Tran HT et al. Acute safety and liver after allogeneic bone marrow transplantation. Blood pharmacokinetics of intravenous busulfan when used with oral 1998; 92: 4568–4572. busulfan and cyclophosphamide as pretransplantation con- 32 Bearman SI, Hinds MS, Wolford JL et al. A pilot study of ditioning therapy: a phase I study. Biol Blood Marrow Trans- continuous infusion heparin for the prevention of hepatic ven- plant 2000; 6: 548–554. occlusive disease after bone marrow transplantation. Bone 49 Vaughan WP, Cagnoni P, Fernandez H et al. Decreased inci- Marrow Transplant 1990; 5: 407–411. dence of and risk factors for hepatic veno-occlusive disease 33 Marsa-Vila L, Gorin NC, Laport JP et al. Prophylactic heparin with an intravenous busulfan containing preparative regimen does not prevent liver veno-occlusive disease following auto- for hematopoietic stem cell transplantation. Blood 1998; 92 logous bone marrow transplantation. Eur J Haematol 1991; (Suppl. 1): 516a. 47: 346–354. 50 Holler E, Kolb H-J, Moller A et al. Increased serum levels of 34 Attal M, Huguet F, Rubie H et al. Prevention of hepatic veno- tumor necrosis factor alpha precede major complications of occlusive disease after bone marrow transplantation by con- bone marrow transplantation. Blood 1990; 75: 1011–1016. tinuous infusion of low-dose heparin: a prospective, ran- 51 Ferra C, de Sanjose S, Lastra CF et al. Pentoxifylline, cipro- domized trial. 1992; Blood 79: 2834–2840. floxacin and prednisone failed to prevent transplant-related 35 Rosenthal J, Sender L, Secola R et al. Phase II trial of heparin toxicities in bone marrow transplant recipients and were asso- prophylaxis for veno-occlusive disease of the liver in children ciated with an increased incidence of infectious complications. undergoing bone marrow transplantation. Bone Marrow Bone Marrow Transplant 1997; 20: 1075–1080. Transplant 1996; 18:185–191. 52 Tanaka J, Imamura M, Kasai M et al. Cytokine gene 36 Or R, Nagler A, Shpilberg O et al. Low molecular weight expression in peripheral blood mononuclear cells during graft- heparin for the prevention of veno-occlusive disease of the versus-host disease after allogeneic bone marrow transplan- liver in bone marrow transplantation patients. Transplantation tation. Br J Haematol 1993; 85: 558–565. 1996; 61:1067–1071. 53 Remberger M, Ringden O. Increased levels of soluble interleu- 37 Bearman SI, Lee JL, Baron AE, McDonald GB. Treatment of kin-2 receptor in veno-occlusive disease of the liver after hepatic venocclusive disease with recombinant human tissue allogenic bone marrow transplantation. Transplantation 1995; plasminogen activator and heparin in 42 marrow transplant 60: 1293–1299. patients. Blood 1997; 89: 1501–1506. 54 Khoury H, Adkins D, Brown R et al. Does early treatment 38 Schriber J, Milk B, Shaw D et al. Tissue plasminogen acti- with high-dose methylprednisolone alter the course of hepatic vator (tPA) as therapy for hepatotoxicity following bone mar- regimen-related toxicity? Bone Marrow Transplant 2000; 25: row transplantation. Bone Marrow Transplant 1999; 24: 737–743. 1311–1314. 55 Testa S, Manna A, Porcellini A et al. Increased plasma level 39 Morris JD, Harris RE, Hashmi R et al. Antithrombin-III for of vascular endothelial glycoprotein thrombomodulin as an the treatment of chemotherapy-induced organ dysfunction fol- early indicator of endothelial damage in bone marrow trans- lowing bone marrow transplantation. Bone Marrow Trans- plantation. Bone Marrow Transplant 1996; 18: 383–388. plant 1997; 20: 871–878. 56 Catani L, Gugliotta L, Vianelli N et al. Endothelium and bone 40 Mertens R, Brost H, Granzen B, Nowak-Gottl U. Antithrom- marrow transplantation. Bone Marrow Transplant 1996; 17: bin treatment of severe hepatic veno-occlusive disease in chil- 277–280. dren with cancer. Eur J Pediatr 1999; 158 (Suppl. 3): 57 Richardson P, Hoppensteadt D, Elias A et al. Elevation of S154–S158. tissue factor pathway inhibitor (TFPI), thrombomodulin (TM) 41 Richardson PG, Murakami C, Warren DL et al. Multi-insti- and plasminogen activator inhibitor-1 (PAI-1) levels in stem tutional emergency use of defibrotide in 75 patients post SCT cell transplant (SCT)-associated veno-occlusive disease. Blood with severe VOD and multisystem organ failure: Response 1997; 90 (Suppl. 1): 219a. without significant toxicity in a high risk population. Blood 58 Salat C, Holler E, Reinhardt B et al. Parameters of the 2000; 96 (Suppl. 1): 2510. fibrinolytic system in patients undergoing BMT: elevation of 42 Grochow LB, Jones RJ, Brundrett RB et al. Pharmacokinetics PAI-1 in veno-occlusive disease. Bone Marrow Transplant of busulfan: correlation of veno-occlusive disease in patients 1994; 14: 747–750. undergoing bone marrow transplantation. Cancer Chemother 59 Salat C, Holler E, Kolb H-J et al. Plasminogen activator Pharmacol 1989; 25: 55–61. inhibitor-1 confirms the diagnosis of hepatic veno-occlusive 43 Dix SP, Wingard JR, Mullins RE et al. Association of busul- disease in patients with hyperbilirubinemia after bone marrow fan area under the curve with veno-occlusive disease follow- transplantation. Blood 1997; 89: 2184–2188. ing BMT. Bone Marrow Transplant 1996; 17: 225–230. 60 Richard S, Seigneur M, Blann A et al. Vascular endothelial 44 Ljungman P, Hassan M, Bekassy AN et al. High busulfan lesion in patients undergoing bone marrow transplantation. concentrations are associated with increased transplant-related Bone Marrow Transplant 1996; 18: 955–959. mortality in allogeneic bone marrow transplant patients. Bone 61 Crawford SW, Fisher L. Predictive value of pulmonary func- Marrow Transplant 1997; 20: 909–913. tion tests before marrow transplantation. Chest 1992; 101: 45 Slattery JT, Sanders JE, Buckner CD et al. Graft rejection and 1257–1264. toxicity following bone marrow transplantation in relation to 62 Manute-Bello G, McDonald GB, Hinds MS et al. Association busulfan pharmacokinetics. Bone Marrow Transplant 1995; of pulmonary function testing abnormalities and severe veno- 16: 31–42. occlusive disease of the liver after marrow transplantation. 46 Vassal G, Koscielny S, Challine D et al. Busulfan disposition Bone Marrow Transplant 1998; 21: 1125–1130.

Bone Marrow Transplantation Avoiding hepatic veno-occlusive disease SI Bearman 1120 63 Friedman SL. The cellular basis of hepatic ﬁbrosis. Mech- 72 Essell JH, Schroeder MT, Harman GS et al. Ursodiol prophy- anisms and treatment strategies. New Engl J Med 1993; 328: laxis against hepatic complications of allogeneic bone marrow 1828–1835, 1993. transplantation. A randomized, double-blind, placebo-con- 64 Scanga AE, Schrum LW, Iozzo RV et al. Transcriptional regu- trolled trial. Ann Intern Med 1998; 128: 975–981. lation of the biglycan gene. Hepatol 1997; 26 (Suppl.): 597A. 73 Ohashi K, Tanabe J, Watanabe R et al. The Japanese multi- 65 Sato Y, Asada Y, Hara S et al. Hepatic stellate cells (Ito cells) center open randomized trial of ursodeoxycholic acid prophy- in veno-occlusive disease of the liver after allogeneic bone laxis for hepatic veno-occlusive disease after stem cell trans- marrow transplantation. Histopathology 1999; 34: 66–70. plantation. Am J Hematol 2000; 64: 32–38. 66 Eltumi M, Trivedi P, Hobbs JR et al. Monitoring of veno- 74 Ishida H, Nakayasu H, Nukaya H, Tsuji K. Study of the phar- occlusive disease after bone marrow transplantation by serum macological effect of the bile salt, sodium scymnol sulfate, amino-propeptide of type III procollagen. Lancet 1993; 342: from Rhizopriondon acutus. II. Prophylactic effect of scymnol 518–521. on lesion development in a rat peripheral arterial occlusion 67 Heikinheimo M, Halili R, Fasth A. Serum procollagen type model. Biol Pharm Bull 1998; 21: 240–244. III is an early and sensitive marker for veno-occlusive disease 75 Neuman MG, Shear NH, Bellentani S, Tiribelli C. Role of of the liver in children undergoing bone marrow transplan- cytokines in ethanol-induced cytotoxicity in vitro in Hep G2 tation. Blood 1994; 83: 3036–3040. cells. Gastroenterol 1998; 115: 157–166. 68 Rio B, Bauduer F, Arrago JP, Zittoun R. N-terminal peptide 76 Fried MW, Connaghan DG, Sharma S et al. Transjugular of type III procollagen: a marker for the development of hep- intrahepatic portosystemic shunt for the management of severe atic veno-occlusive disease after BMT and a basis for venocclusive disease following bone marrow transplantation. determining the timing of prophylactic heparin. Bone Marrow Hepatol 1996; 24: 588–591. Transplant 1993; 11: 471–472. 77 de la Rubia J, Carral A, Montes H et al. Successful treatment 69 Schuppan D, Farrand A, Oesterling C et al. Circulating mark- of hepatic veno-occlusive disease in a peripheral blood pro- ers of hepatic ﬁbrosis predict evolution of venocclusive dis- genitor cell transplant patient with a transjugular intrahepatic ease after marrow transplantation. Hepatol 1997; 26 portosystemic stent-shunt (TIPS). Haematologica 1996; 81: (Suppl.): 452A. 536–539. 70 Cavet J, Middleton PG, Segall M et al. Recipient tumor 78 Smith FO, Johnson MS, Scherer LR et al. Transjugular necrosis factor-alpha and interleukin-10 gene polymorphisms intrahepatic portosystemic shunting (TIPS) for treatment of associate with early mortality and acute graft-versus-host dis- severe hepatic veno-occlusive disease. Bone Marrow Trans- ease severity in HLA-matched sibling bone marrow transplant 1996; 18: 643–646. plants. Blood 1999; 94: 3941–3446. 79 Azoulay D, Castaing D, Lemoine A et al. Transjugular 71 Haire WD, Cavet J, Pavletic SZ et al. Tumor necrosis factor intrahepatic portosystemic shunt (TIPS) for severe veno- d3 allele predicts for organ dysfunction after allogeneic blood occlusive disease of the liver following bone marrow trans- stem cell transplantation. Blood 2000; 96 (Suppl. 1): 584a. plantation. Bone Marrow Transplant 2000; 25: 987–992.

Bone Marrow Transplantation