Hepatic sinusoidal obstruction syndrome (veno-occlusive disease) in children

INTRODUCTION — Hepatic sinusoidal obstruction syndrome (SOS), also called veno-occlusive disease (VOD), is a toxic injury to liver sinusoidal endothelium that occurs in up to one-third of children who undergo hematopoietic cell transplantation (HCT). Hepatic SOS typically presents with transfusion-refractory thrombocytopenia, right upper quadrant pain, hepatomegaly, ascites, and jaundice within 30 days of HCT, but the clinical presentation can vary. Infrequently, hepatic SOS occurs after liver injury from chemotherapy, immunotherapy, or radiation therapy.

Hepatic SOS can rapidly progress to multiorgan dysfunction and death. A high index of suspicion is needed to evaluate and diagnose SOS, and prompt intervention is critical for reducing the associated morbidity and mortality. The incidence, risk factors, clinical presentation, evaluation, diagnosis, and management of hepatic SOS differ between children and adults.

This topic discusses evaluation, diagnosis, and management of hepatic SOS in children.

Hepatic SOS in adults is discussed separately. (See "Hepatic sinusoidal obstruction syndrome (veno-occlusive disease) in adults".)

EPIDEMIOLOGY — Hepatic SOS develops in up to one-third of children after undergoing hematopoietic cell transplantation (HCT) and much less often following liver injury from chemotherapy, immunotherapy, or radiation therapy (RT).

●Post-hematopoietic cell transplantation – Hepatic SOS is estimated to occur in 30 percent of children after transplantation, but the incidence varies with the patient population, underlying disease, and conditioning regimen [1]. Previous estimates of incidence have ranged widely, in part, because of better recognition and evolving diagnostic criteria [2].

In most studies of children, the incidence of SOS after HCT is 20 to 30 percent, but it can reach as high as 60 percent with certain diseases (eg, hemophagocytic lymphohistiocytosis, osteopetrosis, thalassemia) [1,3,4]. The incidence was estimated to be 15 percent in a meta-analysis of 21 studies that included 27,679 children and adults [5]. The rate of overall survival is two- to threefold higher in children than in adults due to the underlying disease, young age, and other risk factors.

The incidence of SOS varies according to aspects of transplantation. In a review of SOS in children and adults, allogeneic HCT using myeloablative conditioning was associated with 10 and 60 percent incidence, but it is lower with reduced-intensity conditioning; the incidence was 5 and 30 percent after autologous HCT [6].

●Other causes – Particularly in children, hepatic SOS can occur outside of the setting of HCT as a complication of immunotherapy, chemotherapy, or RT [6,7].

This occurs most often after treatment with antibody-drug conjugates (eg, gemtuzumab ozogamicin, inotuzumab ozogamicin), actinomycin D, or RT. It can arise after chronic treatment with antimetabolites (eg, thioguanine, azathioprine, mercaptopurine), where the presentation is typically subacute or chronic.

RISK FACTORS — Risk factors for hepatic SOS after hematopoietic cell transplantation (HCT) include patient characteristics (eg, age, underlying condition, pre-existent liver disease) and transplantation-related factors (eg, conditioning regimen, graft source, regimen for graft-versus-host disease [GVHD] prophylaxis) (table 1). The strength of these associations varies among studies [8-10].

Patient characteristics — Pretransplantation characteristics associated with increased risk of hepatic SOS include:

●Age – Children <2 years [4,10,11]; infancy is the greatest risk factor for SOS in children undergoing HCT.

●Underlying disease – Osteopetrosis, hemophagocytic lymphohistiocytosis, juvenile myelomonocytic leukemia, neuroblastoma, and diseases associated with significant liver iron overload and fibrosis (eg, advanced thalassemia) [3,12,13].

●Liver disease – Pre-existent liver disease is associated with increased risk for SOS [6,8,9,14-19].

Patients with liver cirrhosis are at great risk for developing SOS and are generally considered ineligible for myeloablative HCT. (See "Determining eligibility for autologous hematopoietic cell transplantation" and "Determining eligibility for allogeneic hematopoietic cell transplantation".)

Elevated serum aspartate aminotransferase (AST) reflects acute inflammation in the liver; in one series, the risk of developing SOS was increased three- to 10-fold in patients with elevated AST [8].

Aspects of transplantation — Transplantation-related factors associated with increased risk for hepatic SOS include:

●Preparative regimen – Increased risk for SOS is associated with preparative regimens that use certain alkylating agents (eg, busulfan, cytarabine, cyclophosphamide) [4,8,15,20-32] and/or high dose-fraction radiation therapy (RT; ≥12 gray) [16,33-35]. (See "Chemotherapy hepatotoxicity and dose modification in patients with liver disease: Conventional cytotoxic agents", section on 'Hepatic vascular injury'.)

●Graft source – The risk of SOS is generally higher with allogeneic grafts than autologous grafts. This may be related to the degree of alloreactivity; there is a higher risk with an unrelated donor, human leukocyte antigen (HLA)-mismatched donor, or non-T cell depleted graft [2,9,36,37]. The risk associated with HLA-haploidentical donors is uncertain.

●Graft-versus-host disease prophylaxis – Increased risk has been reported in association with certain GVHD prophylaxis regimens, including use of sirolimus in patients receiving cyclophosphamide/total body irradiation and methotrexate used with busulfan- or high-dose etoposide-containing conditioning regimens [11,38-40].

Other causes — In children, hepatic SOS can also occur infrequently as a complication of chemotherapy, immunotherapy, or RT outside of the setting of HCT [6]:

●Calicheamicin-linked immunoconjugates – Treatment with monoclonal antibodies conjugated with calicheamicin (eg, gemtuzumab ozogamicin, inotuzumab ozogamicin) can cause hepatic SOS and are also associated with increased risk in patients who subsequently undergo HCT [41-44].

●Actinomycin D – Treatment with actinomycin D for nephroblastoma or sarcomas is associated with a higher risk for SOS [45,46].

●Other chemotherapy or radiation – Conventional RT and high-dose chemotherapy, exposure to pyrrolizidine alkaloids, and solid-organ transplantation have been associated with SOS [1,47,48].

PATHOGENESIS — Liver necrosis in hepatic SOS results from multifactorial injury to sinusoidal endothelial cells that is amplified by a local inflammatory response and activation of coagulation and fibrinolytic pathways.

The cellular injury is thought to be initiated by toxic metabolites generated by alkylating chemotherapy conditioning regimens (eg, busulfan, cyclophosphamide, melphalan), ionizing radiation, or hepatotoxins [1,49]. These products damage sinusoidal endothelial cells and hepatocytes in the hepatic acinus, which creates gaps in the sinusoidal barrier through which cells and cellular debris pass into the space of Disse beneath the endothelial cells. The narrowed venous lumen reduces sinusoidal venous outflow, causes post-sinusoidal portal hypertension, and leads to widespread zonal liver disruption and centrilobular hemorrhagic necrosis. This is compounded by cell damage from locally released cytokines and activation of the coagulation and fibrinolytic pathways. The process may be exacerbated by impaired drug metabolism and abnormal expression of adhesion molecules and procoagulant factors in patients with pre-existent liver disease.

The pathophysiology of hepatic SOS shares features with other transplant-related systemic endothelial diseases (eg, engraftment syndrome, acute graft-versus-host disease, transplant-associated microangiopathy) and drug-associated endothelial cell injury. (See "Pathogenesis of graft-versus-host disease (GVHD)", section on 'Acute GVHD' and "Cancer-associated hypercoagulable state: Causes and mechanisms", section on 'Therapy-related factors'.)

CLINICAL PRESENTATION — The most sensitive clinical indicator for hepatic SOS in children is a sudden increase in the need for platelet transfusions; this is generally followed by fluid overload/weight gain, abdominal complaints, and jaundice. The peak incidence of hepatic SOS is 12 days after hematopoietic cell transplantation (HCT).

●Typical presentation – Most children present within 21 days of HCT with thrombocytopenia refractory to platelet transfusions, unexplained weight gain, ascites, hepatomegaly, and jaundice; abdominal pain/tenderness can be challenging to recognize in infants and toddlers [50]. Refractory or consumptive thrombocytopenia is often the earliest manifestation of SOS. Dyspnea, tachypnea, or other evidence of fluid overload may accompany renal, cardiac, or pulmonary dysfunction in children who progress to multiorgan dysfunction/failure.

●Late presentation – One-fifth of children present with hepatic SOS >21 days after HCT [1,50].

●Anicteric presentation – Hyperbilirubinemia is a characteristic hallmark of hepatic SOS, but it was seen in only 71 percent of children at diagnosis [51].

The clinical presentation of hepatic SOS that follows chemotherapy, immunotherapy, or radiation therapy outside of the setting of HCT is more likely to be subacute or chronic, but it can also arise after fulminant liver failure in the company of acute hepatic necrosis [7].

EVALUATION — A high index of suspicion is required to diagnose hepatic SOS. Suspected cases are evaluated clinically with laboratory studies and abdominal imaging.

Clinical evaluation — Every child who undergoes hematopoietic cell transplantation (HCT) requires daily clinical evaluation with an eye toward features that might suggest development of hepatic SOS. Daily evaluation should include:

●Fluid balance – Fluid intake/output and weight.

●Interval history – Abdominal swelling or pain (especially right upper quadrant or epigastrium), nausea/vomiting, dyspnea, peripheral edema, and headache.

●Physical examination – Daily abdominal girth and other evidence of fluid accumulation, hepatomegaly, and abdominal tenderness.

Laboratory studies

●Hematology – Daily complete blood count and differential count.

●Chemistries – Daily serum chemistries, including electrolytes, renal function tests, and frequent liver function tests (eg, aspartate aminotransferase [AST], alanine aminotransferase [ALT], alkaline phosphatase, gamma-glutamyl transpeptidase [GGT], total and direct bilirubin, albumin, lactate dehydrogenase).

Amylase and lipase should be measured, if clinically warranted.

●Coagulation – Prothrombin time (PT)/international normalized ratio (INR) and partial thromboplastin time (PTT); fibrinogen and other tests of coagulation, as clinically indicated.

Abdominal ultrasound — An abdominal ultrasound (US) should be performed prior to HCT in children; this is especially important in children who have hepatomegaly and/or ascites prior to transplantation.

All children should have baseline abdominal imaging according to pediatric EBMT guidelines [1]. Practices vary, but most institutions obtain a baseline abdominal US. Computed tomography (CT) or magnetic resonance imaging (MRI) can substitute for or complement US, but they require sedation for younger patients, may expose the child to radiation, and are less amenable to serial studies.

●Serial ultrasound – Comparison of US images to the baseline study is important for recognizing hepatic SOS, grading disease severity, and assessing treatment response.

Serial US can detect changes in liver size and/or ascites when compared with the baseline study. Many pediatric patients come to transplant with pre-existing hepatomegaly and/or ascites, so it can be difficult to distinguish SOS from abnormalities related to the underlying disease or other transplant-associated complications.

●Findings – Findings consistent with SOS include ascites, hepatomegaly, gallbladder changes, reversal of portal vein blood flow, decreased size of hepatic veins, and/or elevated hepatic arterial resistive indices [50]. Although such findings are important indicators of hepatic SOS, they are not required for diagnosis, as discussed below. (See 'Diagnosis' below.)

Imaging is especially important for children who may have SOS in association with nontransplant-related causes because they may not be available for daily clinical evaluation.

Liver biopsy — Liver biopsies should generally not be performed in children.

Importantly, by the time a liver biopsy reveals the diagnosis of SOS, the child is likely to have severe disease. Prompt recognition of SOS is important for successful treatment, as described below.

Liver biopsy carries a high risk of complications in this setting; a study of 16 children who underwent liver biopsies at one institution reported hemorrhage or other complications in nearly one-third [52].

DIAGNOSIS AND SEVERITY — Our approach to diagnosis, grading, and management of hepatic SOS in children is consistent with recommendations of EBMT (European Society for Blood and Marrow Transplantation) and GITMO (Gruppo Italiano Trapianto Midollo Osseo e Terapia Cellulare) [1,53].

Diagnosis — A high index of suspicion is needed to diagnose hepatic SOS, as the clinical presentation varies widely, and there is no pathognomonic diagnostic biomarker, imaging characteristic, or biopsy feature.

●When to suspect hepatic SOS – The diagnosis of hepatic SOS should be considered in any child who has undergone hematopoietic cell transplantation (HCT) and develops one or more of the following: refractory thrombocytopenia, hepatomegaly, abdominal pain, ascites, fluid overload, and/or weight gain.

Hepatic SOS is increasingly recognized in children who received chemotherapy, immunotherapy, or radiation therapy without transplantation; the diagnosis must be considered in such children with any of the above findings.

●Diagnostic criteria – Diagnosis of hepatic SOS in children should use EBMT revised diagnostic criteria [1].

Diagnosis of hepatic SOS in children requires ≥2 of the following [1]:

•Thrombocytopenia – Unexplained transfusion-refractory or consumptive thrombocytopenia (eg, unexpected need for platelet transfusions at least once daily).

•Weight gain – Otherwise unexplained weight gain on three consecutive days despite the use of diuretics, or a weight gain >5 percent above baseline value.

•Hepatomegaly – Liver enlargement above baseline, best judged using ultrasound (US) imaging.

•Ascites – Ascites above baseline, best judged with US.

•Bilirubin – Bilirubin rising above the baseline value on three consecutive days or bilirubin ≥2 mg/dL within 72 hours.

We advise against performing a liver biopsy to diagnose hepatic SOS in children. Liver biopsy is associated with substantial risks in this setting and may delay initiation of treatment, which should begin with the first evidence of hepatic SOS (when liver biopsy may be unrevealing).

The Seattle and Baltimore diagnostic models should not be used in children because they do not adequately account for late and anicteric presentations of SOS, do not include refractory thrombocytopenia as a diagnostic criterion, and do not account for the prevalence of pre-existent hepatomegaly/hyperbilirubinemia in many children undergoing HCT.

The EBMT model provides greater sensitivity and enables earlier diagnosis in children than the Seattle and Baltimore models. A retrospective study of pediatric and adolescent/young adult patients who underwent HCT reported earlier diagnosis (by approximately three days) and a higher incidence of SOS (16 percent) using EBMT criteria, compared with the modified Seattle (12 percent) and Baltimore (7 percent) criteria [54]. Three-quarters of children had refractory thrombocytopenia at the time of diagnosis, which is not a diagnostic criterion in the Seattle and Baltimore models; nearly two-thirds were anicteric at diagnosis, although nearly all later developed hyperbilirubinemia (over a median of four days). Another retrospective study reported that nearly one-third of children had an anicteric presentation, which would not have been captured using the Seattle and Baltimore models [55].

The Seattle and Baltimore models for diagnosing hepatic SOS in adults are described separately. (See "Hepatic sinusoidal obstruction syndrome (veno-occlusive disease) in adults", section on 'Other diagnostic models'.)

Severity grading — The EBMT grading system distinguishes mild/moderate hepatic SOS from severe/very severe disease in children.

Severity grading in children (<18 years) is based on clinical status, laboratory findings, and imaging [1]. EBMT criteria include clinical features (eg, ascites, oxygen requirement, cognitive impairment) and the severity and the kinetics of laboratory abnormalities (eg, persistent thrombocytopenia, transaminases, bilirubin, kidney function, coagulation).

US imaging is an important adjunctive technique, but it is not required to diagnose SOS.

●Severe/very severe – Severe/very severe hepatic SOS is characterized by at least one of the following [1]:

•Liver function tests – Alanine transaminase (ALT), aspartate transaminase (AST), glutamate dehydrogenase (GLDH) >5 times upper limit of normal

•Refractory thrombocytopenia – Refractory for >7 days

•Bilirubin – Either of the following:

-≥2 mg/dL (≥34 micromol/L)

-Doubling ≤48 hours

•Ascites – Requiring paracentesis/external drainage

•Impaired coagulation

•Renal function – Glomerular filtration rate (GFR) <30 mL/min

•Oxygen needs – Disease severity that necessitates invasive pulmonary ventilation (including continuous positive airway pressure [CPAP])

•Mental status – New onset of cognitive impairment

●Mild/moderate – None of the criteria for severe/very severe disease (above) are met.

DIFFERENTIAL DIAGNOSIS — The differential diagnosis includes other causes of liver dysfunction, refractory thrombocytopenia, fluid overload, or multiorgan failure in patients who recently underwent hematopoietic cell transplantation (HCT).

For patients who underwent HCT, the differential diagnosis includes:

●Engraftment syndrome/capillary leak syndrome/peri-engraftment respiratory distress syndrome – These syndromes generally occur 9 to 16 days after HCT and are thought to be related to the release of proinflammatory cytokines during neutrophil recovery. Each can manifest as ascites, edema, and weight gain that resembles hepatic SOS; respiratory distress is a prominent feature of peri-engraftment respiratory distress syndrome (PERDS), which can resemble advanced SOS.

These syndromes differ from SOS in that they are generally associated with noninfectious fever and/or maculopapular rash, while abdominal pain, hepatomegaly, and liver dysfunction are less prominent features than in SOS. (See "Approach to the immunocompromised patient with fever and pulmonary infiltrates".)

●Acute graft-versus-host disease – Both SOS and acute graft-versus-host disease (aGVHD) can present with abdominal pain and a rising serum bilirubin, but patients with aGVHD usually have concurrent rash and involvement of the gastrointestinal tract. The timing of aGVHD is generally later and coincides with engraftment, but this can overlap the timing of SOS (especially late onset SOS). Skin biopsy can distinguish these processes, but liver biopsy should not be performed in children suspected to have SOS because of excessive bleeding risk. Diagnosis of aGVHD is discussed separately. (See "Clinical manifestations, diagnosis, and grading of acute graft-versus-host disease".)

●Abnormal liver function – Numerous conditions cause abnormal liver function tests in the post-transplant period (table 2 and table 3). SOS may also share features of other causes of fulminant hepatic failure, including ischemic, viral, malignant/infiltrative, and toxic causes. Some pertinent entities include:

•Hepatic infections – Abnormal liver function tests may be due to viral hepatitis (eg, hepatitis B, hepatitis C), other viruses (eg, cytomegalovirus [CMV], varicella zoster [VZV], Epstein-Barr [EBV], human herpesvirus 6 [HHV-6], adenovirus), and hepatosplenic candidiasis. The most likely causes vary with the timing after transplantation. A substantial rise of the transaminases is a hallmark of most such liver infections; in contrast, SOS is typically marked by early refractory thrombocytopenia, followed by a later rise in transaminases or bilirubin. (See "Overview of infections following hematopoietic cell transplantation".)

•Drug toxicity – Many drugs used in the setting of HCT, such as calcineurin inhibitors (cyclosporine, tacrolimus, sirolimus), methotrexate, azole antifungal agents, trimethoprim-sulfamethoxazole, ribavirin, and busulfan are associated with cholestasis, but they primarily cause hepatocytic damage with elevated transaminases and are not associated with refractory thrombocytopenia that is seen in SOS. (See "Hepatotoxicity associated with chronic low-dose methotrexate for nonmalignant disease".)

●Budd-Chiari syndrome – The acute form of Budd-Chiari syndrome (BCS) can occasionally resemble hepatic SOS. BCS is caused by the obstruction of hepatic veins and inferior vena cava, which can be established noninvasively by US with Doppler studies, CT scan, or magnetic resonance angiography. (See "Budd-Chiari syndrome: Epidemiology, clinical manifestations, and diagnosis".)

Outside of the context of HCT, other causes of acute or chronic liver failure, such as hemophagocytic lymphohistiocytosis, viral hepatitis, hepatotoxins, inherited metabolic diseases, hypoperfusion, and immune dysregulation, should be considered, as discussed separately. (See "Acute liver failure in children: Etiology and evaluation".)

PREVENTION — To reduce the incidence and severity of hepatic SOS and lessen associated morbidity/mortality, it is important to mitigate risk factors in all patients and to implement prophylaxis for selected patients.

Modifiable risk factors — Risk factors for hepatic SOS should be reviewed before hematopoietic cell transplantation (HCT). (See 'Risk factors' above.)

Certain risk factors for SOS can be modified, while others (eg, age, underlying disease) cannot. The presence of unmodifiable risk factors may influence selection of bridging therapy, conditioning regimen, graft source, and other aspects of transplantation. (See 'Prophylaxis' below.)

●Pre-existent liver disease/injury – Pre-existent liver disease/injury should be optimized prior to HCT.

•Medications/supplements – Medications and supplements associated with liver injury (eg, azole antifungal agents, acetaminophen) should be avoided in the peri-transplant period. (See "Drug-induced liver injury".)

•Iron chelation – Aggressive iron chelation ("downstaging") can reduce the risk for SOS in children with hepatic inflammation/fibrosis caused by iron overload in association with advanced thalassemia, chronic transfusions for aplastic anemia, or myelodysplastic syndromes/neoplasms.

The decision to pursue months of chelation therapy to reduce iron load prior to HCT must be individualized and should balance the risks of progression of the underlying condition versus the potential risk for hepatic SOS. (See "Iron chelators: Choice of agent, dosing, and adverse effects".)

●Pretransplant/bridging therapy – When possible, we avoid treatment with gemtuzumab ozogamicin or inotuzumab ozogamicin immediately prior to transplantation. We generally use an alternative bridging therapy or wait four to six weeks between exposure to the immunoconjugate and the start of HCT conditioning.

In patients treated with inotuzumab ozogamicin, there was no correlation between SOS incidence and time to transplant [56].

●Aspects of transplantation – A decision to modify aspects of transplantation that can reduce risk must be made in the context of the preferred techniques for the underlying disease, comorbid conditions, and institutional approach [48,57]:

•Conditioning regimen – The following options can reduce the risk for SOS:

-Nonmyeloablative conditioning/reduced-intensity conditioning, instead of myeloablative conditioning.

-Avoidance of busulfan-based and myeloablative total body irradiation-containing conditioning regimens.

-If a busulfan-containing regimen is chosen, administration of busulfan prior to cyclophosphamide (rather than the reverse order) and targeted dosing of busulfan is preferred [2].

●Graft-versus-host disease prophylaxis – When possible, we avoid higher-risk graft-versus-host disease (GVHD) prophylaxis regimens, such as:

•Sirolimus plus methotrexate plus tacrolimus

•Methotrexate plus cyclosporine

Some experts avoid cyclosporine-containing regimens or use a continuous infusion of calcineurin inhibitors (to avoid peak serum levels) to mitigate endothelial damage.

Prophylaxis

Selection of patients — We suggest hepatic SOS prophylaxis for children undergoing HCT who have at least one major risk factor for SOS.

We consider any of the following features to warrant SOS prophylaxis.

●Age – <2 years

●Pre-existing liver disease

●Prior hepatic SOS

●Recent treatment with gemtuzumab ozogamicin or inotuzumab ozogamicin

●Second myeloablative HCT, in particular with busulfan conditioning

●Underlying disease

•High-risk neuroblastoma undergoing autologous HCT

•Hemophagocytic lymphohistiocytosis

•Osteopetrosis

•Myelodysplastic syndromes/neoplasms

•Juvenile myelomonocytic leukemia

•Iron overload (eg, high liver iron content and/or bridging fibrosis) due to thalassemia or chronic transfusion therapy

Some experts apply different criteria for the implementation of prophylaxis for hepatic SOS.

Association of SOS with these features is discussed above. (See 'Risk factors' above.)

Choice of agent for prophylaxis is discussed below. (See 'Choice of agent' below.)

Choice of agent — Neither defibrotide nor ursodeoxycholic acid (UDCA) is approved by the US Food and Drug Administration (FDA) or the European Medicines Agency (EMA) for prophylaxis of hepatic SOS.

Nevertheless, for prophylaxis in children at high risk for VOD/SOS, we suggest defibrotide rather than UDCA based on the favorable balance of efficacy and toxicity.

Defibrotide has little toxicity, and it is the only prophylaxis that has been associated with reduced incidence and severity of SOS in children [58]. Most reports of prophylaxis using UDCA studied adults rather than children, and there is no proven benefit for prophylaxis using other agents [59-62].

Some experts consider UDCA acceptable for prophylaxis in children, based on studies of adults. Use of UDCA for prophylaxis and treatment of SOS is discussed separately. (See "Hepatic sinusoidal obstruction syndrome (veno-occlusive disease) in adults".)

Defibrotide is a mixture of single-stranded oligodeoxyribonucleotides derived from porcine intestinal mucosa that is used for prophylaxis and treatment of hepatic SOS [63].

●Administration – Defibrotide is generally administered 6.25 mg/kg intravenously every six hours and continued for ≥21 days post-transplant.

When SOS develops despite defibrotide prophylaxis, management is discussed below. (See 'Treatment of hepatic SOS' below.)

●Toxicity – The most common adverse effects (AEs) are mild hypotension, diarrhea, nausea, vomiting, and epistaxis; hemorrhage is less common, but it can be life threatening.

Hypersensitivity reactions occurred in <2 percent of patients treated with defibrotide. Defibrotide should be discontinued permanently in patients with a severe or life-threatening hypersensitivity reaction.

●Outcomes – Defibrotide prophylaxis can reduce the incidence, severity, and mortality attributable to SOS, and it is well tolerated; prophylaxis with defibrotide has not been proven to increase overall survival (OS) in transplanted children.

•Systematic analysis – A systematic analysis reported that defibrotide prophylaxis was associated with a significantly lower incidence of hepatic SOS [64]. The study included 3005 patients from 20 studies and reported SOS in 16 percent of control patients; for patients who received defibrotide prophylaxis, the risk ratio for developing SOS was 0.30 (95% CI 0.12-0.71) [64].

•Randomized trials – Results from phase 3 trials of defibrotide are mixed, but this may reflect differences in trial design, endpoints, and patient populations.

-An international trial reported no difference in OS, but prophylaxis reduced the incidence of hepatic SOS in 356 children (<18 years) with high risk for SOS who were randomly assigned to receive or not receive defibrotide [58]. Defibrotide prophylaxis achieved a lower incidence of SOS at day 30 after HCT (12 versus 20 percent; risk difference -7.7 percent [95% CI -15.3 to -0.1]). Defibrotide was given to all children who subsequently developed SOS (irrespective of trial arm); this crossover trial design may have confounded the survival data. Severity of acute GVHD at day 30 and at day 100 was also lower in the defibrotide arm. There was no difference between trial arms in the incidence of chronic GVHD or AEs.

-A phase 3 trial (Harmony) that randomly assigned 280 children and adults to defibrotide prophylaxis versus best supportive care (BSC) was stopped early for futility at a planned interim analysis by the adjudication committee [65]. There was no difference in the composite endpoint of day 30 SOS-free survival between patients receiving prophylaxis (67 percent) versus BSC (73 percent). There was no difference in AEs between trial arms.

Some features of the Harmony trial may have contributed to the absence of a benefit from defibrotide prophylaxis [65]. Few children had the highest risk features and adults (who have a lower incidence of SOS than children) were included, while differences in survival were lessened by a crossover to defibrotide for patients who subsequently developed SOS.

TREATMENT OF HEPATIC SOS — For children with hepatic SOS (whether receiving defibrotide prophylaxis or not), we suggest defibrotide rather than ursodeoxycholic acid (UDCA), other treatments, or supportive care alone. Defibrotide is associated with improved survival and little toxicity in children with SOS. No other treatment has a proven benefit in children, and some are associated with severe adverse effects (AEs; eg, hemorrhage).

Administration, timing of therapy, and toxicity of defibrotide are discussed below. (See 'Defibrotide' below.)

●Outcomes with defibrotide – Defibrotide is associated with improved survival in patients with severe/very severe hepatic SOS, compared with supportive care alone. Because of the morbidity/mortality of severe SOS, there are no randomized trials of defibrotide versus no treatment or placebo.

A systematic analysis that included 2598 patients from 17 studies reported that defibrotide was associated with 54 percent day 100 overall survival (OS; range 35 to 79 percent) [66]. By comparison, severe SOS was previously associated with >80 percent mortality [2,67,68].

Other studies of defibrotide for treatment of severe SOS in children include:

•A prospective multicenter study of 102 patients with severe SOS (43 percent were ≤16 years old) reported that defibrotide was associated with better survival than 32 matched historical control patients [69]. At day 100, defibrotide treatment was associated with a better OS (38 versus 25 percent) and complete response (CR; 26 versus 13 percent) compared with historical controls. Defibrotide was well tolerated, with rates of hemorrhagic AEs like those in the historical controls.

•A retrospective multicenter study that included 22 children and adolescents with severe SOS reported 50 percent CR and 36 percent day 100 OS; none of the long-term survivors died with SOS-related causes [70].

There are no reports of recurrent SOS in patients who achieve a CR with defibrotide, and most patients who recover from SOS regain normal liver function and do not develop sequelae, such as portal hypertension or esophageal varices.

●Other treatments – No other treatment has a proven benefit for children with hepatic SOS. Other treatments that have been evaluated include heparin, tissue plasminogen activator, antithrombin III, glucocorticoids, and prostaglandin E1.

A Cochrane analysis [59] and other studies reported that other agents had no benefit for treating SOS [6,11,71].

Defibrotide — Defibrotide is a mixture of single-stranded oligodeoxyribonucleotides used for prophylaxis and treatment of hepatic SOS [63]. Prophylaxis with defibrotide is discussed above. (See 'Choice of agent' above.)

Our approach to treatment of hepatic SOS with defibrotide follows:

Initiation of therapy — For transfusion-refractory thrombocytopenia or other symptoms of hepatic SOS, we suggest prompt initiation of defibrotide rather than waiting for the development of severe disease. Prompt treatment is associated with improved survival in children with hepatic SOS.

Transfusion-refractory thrombocytopenia is the first manifestation of SOS in most children; other aspects of the clinical presentation of SOS are discussed above. (See 'Clinical presentation' above.)

Recognition and diagnosis of hepatic SOS in children should be based on pediatric EBMT (European Society for Blood and Marrow Transplantation) criteria [1], as discussed above. (See 'Diagnosis' above.)

In untreated children, 30 to 60 percent with mild/moderate SOS progress to multiorgan dysfunction/failure [2,49,58,69]. Waiting to satisfy all diagnostic criteria delays initiation of therapy and increases the risk of multiorgan dysfunction/failure and death because many children with SOS initially manifest only thrombocytopenia, present late (≥21 days after hematopoietic cell transplantation [HCT]), or are anicteric.

A study of 282 children who underwent HCT from 2016 to 2019 reported that prompt initiation of defibrotide (based on EBMT criteria) was associated with improved outcomes compared with historical controls at the same institution [72]. Using EBMT criteria, 9 percent of children were diagnosed with SOS, compared with 2 and 6 percent diagnosed using Baltimore and modified Seattle criteria, respectively (which were used in the historical control patients). Compared with modified Seattle criteria, diagnosis was three days earlier using EBMT criteria. Initiation of defibrotide based on EBMT criteria was associated with 88 percent OS, 96 percent response rate (compared with 74 percent response in historical controls), and shorter hospitalization for patients with SOS (a 12-day median reduction). Duration of defibrotide treatment ranged from 4 to 34 days (a median of 17 days) in study patients.

Administration and toxicity

●Administration – Defibrotide is generally given 6.25 mg/kg every six hours intravenously and continued until the resolution of dynamic criteria for hepatic SOS, rather than a fixed duration of treatment. Duration of defibrotide therapy for the treatment of hepatic SOS is described below. (See 'Duration of therapy' below.)

For children who develop SOS while receiving UDCA prophylaxis, we stop UDCA while treating with defibrotide.

Defibrotide should be discontinued at least two hours prior to invasive procedures and can be resumed after the resolution of procedure-related risk of bleeding. There is no known reversal agent for defibrotide, but the half-life of elimination is <2 hours.

Defibrotide is approved for treatment of severe hepatic SOS by the US Food and Drug Administration (FDA) and European Medicines Agency (EMA).

●Toxicity – Defibrotide is well tolerated and was not associated with increased AEs compared with supportive care in clinical studies.

Defibrotide should be discontinued permanently in patients with a severe or life-threatening hypersensitivity reaction.

Duration of therapy — The duration of defibrotide therapy in patients with hepatic SOS should be individualized according to disease severity and treatment response.

●Milestones for discontinuation – We begin to taper defibrotide when there is improvement in the dynamic criteria for hepatic SOS. This reflects cessation of active disease and includes reduced frequency of platelet transfusions, improvement of coagulopathy, reduced ascites drainage, and unexpected weight loss accompanied by a significant diuresis despite unchanged fluid management.

Static criteria (ie, flow reversal, if present, hyperbilirubinemia, hepatomegaly) are consequences of post-endothelial damage; they do not reflect active disease as directly as the dynamic criteria and may take longer to resolve.

●Taper schedule – We use a "wean and wait" strategy for discontinuing defibrotide. We reduce the dose by 25 percent every three to five days with careful attention to dynamic criteria, which should continuously decrease. Any rise warrants an arrest of the tapering or even a return to the original dosage with a timely restart of the weaning process.

Because the defibrotide taper is based on improvement in dynamic criteria, the duration of therapy is the same whether a child was already receiving defibrotide prophylaxis.

Management for children with an inadequate response after ≥21 days (refractory SOS) is discussed below. (See 'Refractory SOS' below.)

Most patients require >21 days of defibrotide therapy. However, treatment for 21 days is not necessary if dynamic criteria are met before then; this is more likely when treatment is triggered by treatment-refractory thrombocytopenia or other pediatric EBMT criteria, as described above. (See 'Diagnosis' above.)

Supportive care — All patients should receive supportive care to mitigate hepatic SOS, with attention to fluid balance, maintaining intravascular volume and renal perfusion, and limiting third-space fluid collection.

The following supportive measures should be considered:

●Maintain euvolemia – Daily weights and recording of fluid intake and output are critical to maintaining euvolemia in patients with SOS; optimizing fluid balance is important for maintaining adequate renal perfusion while avoiding restriction of lung function due to ascites.

Fluid restriction and diuresis should be implemented when intake greatly exceeds output, but it is important to avoid overly aggressive fluid management that can lead to prerenal failure due to third spacing.

●Minimize hepatotoxic agents – Medications associated with liver injury should be avoided. Pain control may require narcotics to avoid excessive use of acetaminophen.

●Paracentesis – Patients may require serial paracentesis or continuous drainage for ascites that cause discomfort or pulmonary compromise. For children, the relief of pulmonary compromise is important to avoid assisted ventilation. The amount of fluid removed at each session should be limited to maintain renal perfusion. Decreasing ascitic loss may be an indication of response to treatment.

REFRACTORY SOS — Hepatic SOS that responded slowly and/or inadequately to ≥21 days of defibrotide is considered refractory disease. No pharmacologic agent has proven benefit for patients with refractory hepatic SOS.

For patients with refractory SOS, we continue defibrotide until there is an improvement in the dynamic criteria (which reflect sinusoidal recovery), as described above. (See 'Defibrotide' above.)

Other treatments that have been used to treat refractory SOS include:

●High-dose glucocorticoid – High-dose methylprednisolone has been used for the initial treatment of SOS, but a role for the treatment of refractory SOS is uncertain. It should be used with caution in this setting due to the increased risk of infection [73].

●Transjugular intrahepatic portosystemic stent-shunt – Insertion of a transjugular intrahepatic portosystemic stent-shunt (TIPS) was performed in small numbers of patients with SOS; some patients had a regression of the hepatic and renal symptoms [74-76]. Patients with milder disease may be more likely to respond; long-term survival appears to be uncommon but has been reported [76-78]. (See "Overview of transjugular intrahepatic portosystemic shunts (TIPS)".)

●Liver transplantation – Orthotopic liver transplantation has been successfully performed in small numbers of patients with SOS [79,80]. However, most patients with severe SOS are not medically fit to undergo such a rigorous surgical procedure. In many centers, patients at risk for recurrent malignancy are low-priority candidates for liver transplant. (See "Liver transplantation in adults: Patient selection and pretransplantation evaluation".)

SUMMARY AND RECOMMENDATIONS

●Description – Hepatic sinusoidal obstruction syndrome (SOS; also called veno-occlusive disease [VOD]), is a life-threatening condition involving injury to liver sinusoidal endothelium that manifests as refractory thrombocytopenia, abdominal pain, hepatomegaly, ascites, jaundice, and/or multiorgan failure. SOS develops in up to one-third of children who undergo hematopoietic cell transplantation (HCT) and arises rarely after other types of liver injury.

●Causes – Risk factors for SOS after HCT include pretransplant patient characteristics (eg, age, underlying disease, liver disease) and transplantation-related factors (eg, conditioning regimen, graft source, graft-versus-host disease [GVHD] prophylaxis) (table 1). Less often, SOS occurs without HCT (eg, after therapeutic immunoconjugates, chemotherapy, or radiation therapy). (See 'Risk factors' above.)

●Clinical presentation – Most children present ≤21 days after HCT with thrombocytopenia refractory to platelet transfusions, weight gain, ascites, jaundice, and/or firm, tender hepatomegaly; some cases present >30 days after HCT and/or without jaundice. (See 'Clinical presentation' above.)

●Evaluation – Clinical evaluation, laboratory studies, and abdominal ultrasound. (See 'Evaluation' above.)

All children should have a baseline abdominal ultrasound prior to HCT.

●Diagnosis – Diagnosis requires a high index of suspicion and should be considered in a child with transfusion-refractory thrombocytopenia, fluid overload, hepatomegaly, abdominal pain, weight gain, and/or ascites after HCT.

Diagnosis in children is based on EBMT (European Bone Marrow Transplantation) criteria, which include, as described above (see 'Diagnosis' above):

•Transfusion-refractory thrombocytopenia

•Weight gain

•Hepatomegaly

•Ascites

•Bilirubin increase

Liver biopsy is not required for diagnosis and is very risky in this setting.

Other diagnostic models for SOS (eg, Seattle or Baltimore models) should not be used in children.

●Differential diagnosis – Includes other causes of liver dysfunction, refractory thrombocytopenia, fluid overload, or multiorgan failure after HCT, such as engraftment syndrome, capillary leak syndrome, peri-engraftment respiratory distress syndrome, acute GVHD, and other infectious and chemical causes of liver dysfunction.

●Prevention.

•Mitigate risk factors – Factors that increase risk for SOS should be mitigated, when possible. (See 'Modifiable risk factors' above.)

•Prophylaxis.

-We suggest prophylaxis for children with at least one major risk factor for SOS (Grade 2C). (See 'Selection of patients' above.)

-For prophylaxis in children, we suggest defibrotide, rather than ursodeoxycholic acid (Grade 2C).

Defibrotide prophylaxis is administered for ≥21 days.

●Treatment. (See 'Treatment of hepatic SOS' above.)

•Choice of agent – For children with hepatic SOS, we suggest prompt initiation of defibrotide (whether receiving prophylaxis) rather than waiting for the development of severe disease or providing supportive care alone (Grade 2C). (See 'Defibrotide' above.)

•Duration of therapy – Tapering of defibrotide can begin when there is improvement in the dynamic criteria of hepatic SOS, as described above. (See 'Defibrotide' above.)

•Supportive care – All patients require careful attention to fluid balance. (See 'Supportive care' above.)

●Refractory SOS – We continue supportive care and defibrotide beyond 21 days for children who are slowly/incompletely responding to treatment. No other treatment has proven benefit in this setting, but other treatments are discussed above. (See 'Refractory SOS' above.)