Clinics and Research in Hepatology and Gastroenterology (2016) 40, 378—385

Available online at

ScienceDirect
www.sciencedirect.com

MINI REVIEW

Sinusoidal obstruction syndrome

Dominique-Charles Valla a,b,d,∗, Dominique Cazals-Hatem a,c,d

a
Département hospitalo-universitaire UNITY, 92118 Clichy, France
b
Service d’hépatologie, hôpital Beaujon, AP—HP, 100, boulevard Leclerc, 92118 Clichy, France
c
Département de pathologie, hôpital Beaujon, AP—HP, Clichy, France
d
CRI, UMR1149, université Paris Diderot and Inserm, Paris, France

Available online 30 March 2016

Summary Sinusoidal obstruction syndrome (SOS) is characterized by damage to small hepatic

vessels affecting particularly sinusoidal endothelium. Damaged sinusoids can be associated
with a partial or complete occlusion of small hepatic veins, hence the previous denomina-
tion of hepatic veno-occlusive disease (VOD). Exposure to certain exogenous toxins appears to
be specific to this condition and is frequently included in its definition. Typical histopathological
features of SOS in a liver biopsy specimen are presented in the text. The purpose of this article
is to provide an overview on the different entities corresponding to this general definition. Such
entities include: (i) liver disease related to pyrrolizidine alcaloids; (ii) liver injury related to
conditioning for hematopoietic stem cell transplantation; (iii) vascular liver disease occurring
in patients treated with chemotherapy for liver metastasis of colorectal cancer; and (iv) other
liver diseases related to toxic agents.
© 2016 Elsevier Masson SAS. All rights reserved.

Liver disease related to pyrrolizidine alcaloids edema and fibrosis [1]. It was soon recognized that the dis-
ease was associated to the consumption of ‘‘bush tea’’.
This entity was first recognized in South Africa in 1920 A relationship with a similar lesions encountered in cat-
as cirrhosis resulting from Senecio poisoning in humans. It tle exposed to pyrrolizidine alkaloid-containing plants was
was further characterized in the West Indies in the 1950s, rapidly established. Several variants of clinical presentation
based on conspicuous congestion of sinusoids and hemor- were described, including an acute presentation with rapid
rhagic necrosis in centrilobular area while large hepatic and massive abdominal swelling and pain associated with
veins were patent and there was a nonthrombotic occlusion hemorrhagic centrilobular necrosis; a subacute presentation
of central and sublobular hepatic veins by subendothelial with recurrent ascites, splenomegaly and hepatomegaly,
associated with extensive fibrosis in centrilobular areas; and
a chronic variant indistinguishable at bedside from cirrhosis
of other origin, but showing a venocentric type of cirrho-
∗ Corresponding author. Hôpital Beaujon, 100, boulevard Leclerc, sis at histological examination. Young children and adults
92118 Clichy, France. were both affected. A possible full clinical, biochemical and
E-mail address: [email protected] (D.-C. Valla). pathological recovery was recorded in half the patients, a

http://dx.doi.org/10.1016/j.clinre.2016.01.006
2210-7401/© 2016 Elsevier Masson SAS. All rights reserved.
Sinusoidal obstruction syndrome 379

rapid death in 20% of patients, and the development of proposed for the entity, in order to better account for dam-
decompensated liver disease in the rest. In 1970, ultra- age to sinusoidal endothelium rather than occlusion of the
structural studies on liver biopsy specimens from 6 children central vein as a primary event. Coagulative necrosis of
revealed the extensive ‘‘devastation’’ of the endothelium, hepatocytes was also found to occur later than endothelial
the extravasation of erythrocytes the space of Disse, and injury [8]. This model further provided clues to understand
massive dropout of parenchymal cells [2]. how toxic metabolites, which are produced in the hepa-
Pyrrolizidine alcaloids are many, varying in structure and tocytes, induce more severe damage to endothelial cells
origin. They are mostly found in plants, of several families, than hepatocytes. Indeed, decrease in glutathion content
including about 3% of the world’s flowering plants. Following was more profound in endothelial cells than in hepatocytes.
ingestion, they are absorbed from the gut and transformed Moreover, supply in exogenous GSH through the portal vein
by hepatic cytochrome P450, CYP3A and CYP2B into so- protected from SOS/VOD [9]. Further experiments showed
called DHP esters. These metabolites react rapidly with that the earliest changes detectable in sinusoidal endothe-
functional groups on DNA, proteins or glutathione to form lium, the rounding up of sinusoidal endothelial cell was
DHP adducts. Rapid spontaneous hydrolysis of DHP esters dependent on the production of matrix metalloproteinase 9
form less reactive intermediates that can diffuse outside the and 2 by endothelial cells [10], which could be induced by a
liver. DHP adducts may further induce toxic effects. Inter- decreased production of nitric oxide [11]. This experimental
species differences in metabolism to toxic intermediates model also allowed for showing that bone marrow-derived
likely explain part of the differences in species susceptibility progenitors replace sinusoidal and central venous endothe-
to pyrrolizidine alcaloids [3]. lial cells after injury, and that monocrotaline suppresses
In the 1970s, besides the endemic cases related to con- endothelial cell progenitors in the bone marrow and circu-
sumption of bush teas, epidemic forms of the disease have lation [12]. Thus, SOS/VOD appears to be a disease resulting
been described as a result of consuming products made from from 2 combined mechanisms:
wheat contaminated with seeds of pyrrolizidine alkaloid-
containing plants [4]. These epidemics occurred in a context • toxic injury to sinusoidal/central venous endothelial cells;
of war or drought modifying normal harvesting and allow- • toxic injury to bone marrow progenitors preventing the
ing for toxic plants to contaminate crops. Attack rates replacement of the injured endothelial cells in sinusoids
have been up to 30% based on clinical examination, and and central veins.
associated to fatalities or complete recovery within the
same epidemics. Where histopathological studies could be These concepts are highly relevant to hematopoietic
obtained, findings have been similar to those from endemic stem cell transplantation.
cases, spanning from acute centrilobular congestion with
occlusion of central veins to cirrhosis. Such outbreaks of
VOD/SOS have still been reported into the 1990s. Toxic liver injury related to conditioning for
Since the mid-1980s, a number of sporadic cases related hematopoietic stem cell transplantation
to the consumption of herbal remedies have been reported
from western countries or China. Attention has recently SOS/VOD was first and simultaneously reported by sev-
been drawn to the possible SOS/VOD occurring as a result of eral groups in 1979—1980 as a fatal complication of
erroneous substitution of pyrrolizidine alkaloid-containing hematopoietic stem cell transplantation [13]. Because
to non-pyrrolizidine alkaloid in herbal remedy [5]. Recent thrombocytopenia related to myeloablation precluded per-
epidemiological studies have focused on low-level dietary cutaneous liver biopsy to be performed, pathological studies
exposure to pyrrolizidine alcaloids, e.g. consuming honey were based on autopsy material, which limits the interpre-
from plants containing pyrrolizidine alcaloids [6]. Evidence tation of the data because findings were not synchronous to
for significant toxicity from such low-level dietary exposure clinical manifestations, and because of compounding fac-
is still lacking but difficult to check in humans [3]. They tors contributing to death. As a result, histopathological
could explain in part some endemic/epidemic toxic liver definition has mostly included the late finding of fibrous
injury related to co-exposure to DTT, as recently reported obliteration of the central veins, rather than the ear-
for the Hirmi valley disease [7]. Recent efforts have also lier lesions of central hemorrhagic necrosis, or sinusoidal
focused on identifying accurate biomarkers for exposure to changes. Still, clinical criteria for diagnosis were identified
pyrrolizidine alkaloids. Serum pyrrole-protein adducts may on the basis of these autopsy data [14]. Two widely used
prove valuable in this regards [5]. clinical criteria are presented in Table 1. The accuracy of
The long used administration of pyrrolizidine alkaloids to these criteria is limited by the need for a careful exclusion of
animals of various species has allowed a demonstration of a alternative diagnoses, particularly viral hepatitis, bacterial
direct responsibility in inducing the liver changes. A repro- infections, graft versus host disease, and other drug reac-
ducible rat model was eventually developed consisting of tions, all of which are common and frequently combined
gavage with monocrotaline for 1 to 10 days before sacrifice in the setting of hepatopoietic stem cell transplantation.
[8]. This model showed early injury to sinusoidal and central When clinical diagnosis was compared to that reached
vein endothelium, preceding the development of veno- through transvenous liver biopsy, a high rate of false positive
occlusive lesions. The latter could actually be explained diagnosis of SOS/VOD was found, as well as an underestima-
by the contiguity of venous subendothelial areas with the tion of associated conditions [15,16]. Furthermore, when
space of Disse, where endothelial denudation allowed accu- pre-transplantation and post-transplantation liver biopsy
mulation of erythrocytes and cellular or non-cellular debris. could be compared in the same patient, many features
These findings lead the new denomination of ‘‘SOS’’ being of VOD found after transplantation were already present
380 D.-C. Valla, D. Cazals-Hatem

was 1% of the whole series, 18% of SOS/VOD patients, and

Table 1 Two widely used sets of criteria for a clinical diag-
67% of severe SOS/VOD [21]. Baseline factors predicting the
nosis of veno-occlusive disease.
occurrence of severe clinical forms have not been well char-
Seattle criteria Baltimore criteria acterized. In patients developing SOS/VOD, features of more
severe liver disease (increased serum bilirubin or transami-
Within 20 days of Within 30 days of nase levels, increased hepatic venous pressure gradient) are
transplantation, two of transplantation, associated with a poorer outcome [14].
three findings among the bilirubinemia > 34.2 mol/L Prophylaxis with defibrotide has been recommended in
following (2 mg/dL), plus 2 other children at high risk of SOS/VOD. Its use has also been sug-
findings among the gested in such adults [23,24]. The recommendation is based
following on a positive randomized controlled trial in 356 children
Bilirubin > 34.2 mol/L Hepatomegaly, usually [25]. Actually, 22 (12%) patients receiving defibrotide devel-
(2 mg/dL) painful oped clinically defined SOS/VOD by day 30, versus 35 (20%) in
Hepatomegaly or RUQ > 5% weight gain the control group (P = 0.05). The design was that of an open-
pain of liver origin label trial. Non-relapse mortality was 9% in both groups.
> 2% weight gain due to Ascites Therefore, the clinical impact of prophylactic defibrotide
fluid accumulation administration appears to be limited. The drug however,
appears to be well tolerated [26]. Other agents proposed
for prophylaxis have included prostaglandin E1, pentoxy-
before, reflecting the importance of taking into account fylline, heparin and antithrombin but none of these are
pre-transplant condition and management [17]. Seattle and recommended or suggested [23]. Ursodeoxycholic acid has
Baltimore criteria yield differing estimates for the incidence been suggested based on randomized trials showing both
of SOS/VOD [18,19]. Taken together, these data suggest decreased and unchanged incidence of SOS/VOD, but with-
that the widely used clinical criteria for SOS/VOD point out any impact on mortality [23].
to a syndrome of liver dysfunction where SOS/VOD plays Data on treatment for established SOS/VOD from ran-
a role together with many other factors, rather than to a domized controlled trials are lacking. Most studies had a
discrete clinicopathological entity. A hepatic venous pres- retrospective design or, when prospective, used histori-
sure gradient above 9 or 10 mmHg appears to be 85—90% cal controls. Most studies reported apparently improved
specific for finding SOS/VOD lesions at liver biopsy [15,16]. outcomes. Tolerance appears to be good. As a result, defi-
Abdominal imaging with ultrasound, CT scan or magnetic brotide has been recommended for treatment of SOS/VOD
resonance show thickening of the gallbladder, enlarged liver despite a poor methodological quality of the trials [23].
and spleen, ascites, decreased or reversed flow in the portal Tissue plasminogen activator is not recommended as sev-
vein, all of which being non-specific features of liver disease eral fatalities related to the use of this agent have been
are of little help in establishing diagnosis [20]. reported. The use of N-acetyl cysteine proved inefficient in
Depending on clinical diagnostic criteria and risk factors, a randomized controlled trial [23].
the incidence of SOS/VOD has ranged between 0 and 50% Therefore, the most effective means to control SOS/VOD
among transplantation units, and the fatality rate between associated to hematopoietic stem cell transplantation has
3 and 45% [14]. In a large multicentric prospective survey, been to reduce the exposure to chemotherapy and radio-
the incidence was 5.3% [21]. Major risk factors for the occur- therapy, including conditioning for transplantation and
rence of this clinical SOS/VOD include: prior therapies for underlying blood disease. Attempts at
adjusting the dose of myeloablative agents to individual
• the intensity of the conditioning regimen; polymorphism in drug metabolism has thus far not been
• the autologous or allogeneic type of hematopoietic stem clearly associated with a decreased incidence of SOS/VOD
cell transplantation; [27].
• exposure to estroprogestatives in women; In summary, SOS/VOD related to hematopoietic stem cell
• second myeloablative transplantation; transplantation is a condition decreasing in incidence and
• preexistent liver disease (as indicated by raised serum severity due to increased attention to a reduced exposure
transaminase activity) [14]. to toxic chemotherapeutic agents. Unmet needs consist of
accurate biomarkers for early diagnosis and evaluation of
High intensity regimens include those using both busul- patients at risk, and well-designed therapeutic trials for
fan and cyclophosphamide, or cyclophosphamide with total prevention and management of the disease.
body irradiation > 12 Gy, or cyclophosphamide and BCNU and
etoposide; or those with an equivalent intensity. Based on
ex vivo studies, cyclophosphamide toxicity to sinusoidal Sinusoidal changes after chemotherapy for
endothelial cells appears to be mediated by metabolism liver metastasis
to toxic intermediates in hepatocytes and to depletion in
glutathione in sinusoidal endothelial cells [22]. In patients with liver metastasis from colorectal cancer,
In the multicentric European survey mentioned above, preoperative chemotherapy is widely recommended, based
SOS/VOD was classified as mild (i.e. self-limited) in 8%; mod- on a well-established association with improved survival
erate (i.e. with complete resolution but necessitating ther- [28]. As a consequence, an abundance of liver resection
apy) in 64%; and severe (i.e. leading to death or not resolving specimens have become available for a comprehensive
by day 100) in 28% of cases. SOS/VOD related mortality evaluation of the changes in non-tumorous areas. Indeed,
Sinusoidal obstruction syndrome 381

Figure 1 Sinusoidal obstruction syndrome related to oxaliplatin use for colo-rectal metastatic liver disease (A: fresh cut section of
hepatectomy for metastasis, B: Masson trichrome, C: hematein eosin-safran, D: argentation stain): Grossly, diffuse liver congestion
has a typical nutmeg aspect corresponding on histology to extensive veno-venous dilatation and congestion around small hepatic
veins*; centrilobular hepatocytes plates are atrophic or disrupted and a sinusoidal fibrosis is frequently observed without venous
obliteration.

specific sinusoidal changes resembling those of SOS/VOD In this model, however, sinusoidal dilatation and hepato-
have been described of such patients and illustrated in cyte atrophy are marked but endothelial destruction and
Fig. 1. They consist of congestive sinusoidal dilatation pre- hepatocyte necrosis are inconspicuous. There is evidence
dominating in centrilobular area in 30—65% of patients, for an activation of VEGF and IL6 pathway mediators both
perisinusoidal fibrosis in 35—40%, centrilobular fibrosis in in patients and in experimental models [36—38]. There is
30%, atrophy of liver cell plate but few necrotic cells bor- also evidence that changes induced in the tumor itself by
dering dilated areas, and low-grade nodular regenerative chemotherapeutic agents participate in inducing sinusoidal
changes in 12—20% [29,30]. According to a widely used changes through an activation of pro-inflammatory pathways
classification, the sinusoidal lesions are mild (less than [39]. Therefore, it is still unclear whether the sinusoidal
1/3 of the lobule), moderate (1/3 to 2/3 of the lobule), changes observed in oxaliplatin based chemotherapy are
and severe (extending in the whole lobule) in about 30% identical to SOS/VOD observed after hematopoietic stem
of patients each. Only in occasional patients with severe cell transplantation or pyrrolizidine alkaloid exposure, or
lesions was extravasation of erythrocytes in the space of are mostly related to pro-inflammatory mechanisms [40].
Disse described [31,32]. Similar lesions were not found in Clinical manifestations associated with sinusoidal
control patients receiving resection without preoperative changes in patients with liver metastasis from colorectal
chemotherapy [31]. carcinoma appear to be mild or absent. However, patients
There is a clear association with oxaliplatin-based with clinically conspicuous toxicity from chemotherapy
chemotherapy, where the incidence of such sinusoidal could have been denied resection and thus could have
changes is up to 75%. Still, irinotecan-based chemotherapy is been excluded from analyses due to lack of histological
also associated with a 25—40% incidence of similar sinusoidal data. Actually, reports of clinically significant liver injury
changes [29,30]. Actually 5-fluorouracile also was found in unoperated or operated patients are scarce. They were
to be independently associated with the development of exclusively observed in patients receiving intrahepatic arte-
sinusoidal changes [33]. Co-administration of bevacizumab rial administration of high doses of oxaliplatin combined
has been shown to reduce the incidence and severity of with systemic or hepatic arterial infusion of other agents
sinusoidal changes [32—35]. A murine model of FOLFOX- [41,42]. Therefore, the frequency of pathological findings
associated sinusoidal alterations has been established [36]. contrasts with the rarity of the clinical manifestations.
382 D.-C. Valla, D. Cazals-Hatem

The impact of sinusoidal changes associated with Similarity with pyrrolizidine alkaloid or hematopoietic stem
chemotherapy on the outcome of patients has been unclear. cell transplantation is only partial, probably related to the
Increased postoperative morbidity after major resection has use of clinically less toxic doses in most protocols than
been reported by some [43,44] but not all [30] investiga- those used for preparation to hematopoietic stem cell
tors. Interestingly, the occurrence of sinusoidal changes has transplantation.
been associated with a decreased tumor response [29,45].
However, there is currently no evidence for an increased
mortality in patients with sinusoidal changes [29]. Sinusoidal changes related to other drugs or
Determinants for the development of sinusoidal changes conditions
after chemotherapy for colorectal cancer have not been
clearly identified yet. Genetic polymorphism in drug metab-
Table 2 presents the agents which have been reported to be
olizing enzymes, such as glutathione S-transferase, has been
associated with SOS/VOD. All of these agents share causing
incriminated [46]. Several features have been reported
bone marrow suppression in addition to SOS/VOD.
as predictive for a finding of severe sinusoidal changes
The relationship of azathioprine to SOS/VOD is particu-
in resected livers, including APRI score [44,47], serum
larly unclear for several reasons:
hyaluronic acid [47], and a reticular pattern of enhance-
ment at CT or MRI abdominal imaging following injection of
vascular contrast media [48,49]. Of note, particular focal • there is no animal model;
lesions simulating metastasis can be attributed to sinusoidal • dose relationship is not obvious;
changes. A distinctive feature of such focal lesions is their • similar to oxaliplatin, lesions have been described under
ill-defined margin on hepatobiliary phase of gadoxetic acid various denomination including SOS/VOD but also peliosis,
enhanced and diffusion weighed MR imaging [49]. sinusoidal dilatation and nodular regenerative hyperpla-
In summary, oxaliplatin-associated sinusoidal lesions sia;
is a pathological entity with minimal clinical expression, • conditions for which azathioprine was administered have
which is not entirely specific for oxaliplatin in the context been reported to be associated with sinusoidal changes
of chemotherapy for liver metastasis of colorectal cancer. including Crohn’s disease [40], renal transplantation and
The impact on patients’ outcome is uncertain or minimal. liver transplantation [50,51].

Figure 2 Acute sinusoidal obstruction syndrome related to gemtuzumab use (Mylotarg) after haematopoietic stem cell transplan-
tation (A & B: Masson trichrome, C: hematein eosin-safran; D: argentation stain): dilatation and congestion of sinusoids are limited
to centrilobular zones around the terminal hepatic vein *; endothelial cells of veins and sinusoids are damaged, leading to a huge
hematic deposition in Disse space and to hepatocyte necrosis around the central veins.
Sinusoidal obstruction syndrome 383

result from similar mechanisms despite a much less severe

Table 2 Drugs reported to be associated with SOS/VOD.
clinical and histological expression, or may stem from dif-
6-mercaptopurine ferent pathways implicating proinflammatory mediators.
6-thioguanine Decreasing the intensity of chemotherapeutic regimen is
Actinomycin D currently the only proven means to prevent or decrease
Azathioprine the severity of sinusoidal changes and their clinical expres-
Busulfan sion. Long-term effect of low intensity regimens will have
Cytosine arabinoside to be assessed. Indeed, portal hypertension due to nodu-
Cyclophosphamide lar regenerative hyperplasia may represent a more common
Dacarbazine sequela of infraclinical sinusoidal damage than currently
Gemtuzumab-ozogamicin recognized [55]. Elucidation of the mechanisms under-
Melphalan lying the recently reported SP110 mutations associated
Oxaliplatin familial SOS/VOD occurring in the absence of exposure to
Urethane chemotherapy or related agents [52] will shed light on the
mechanisms involved when such agents.

Recently, an inherited condition combining VOD and

immunodeficiency (so-called VODI) associated with muta- Disclosure of interest
tions in Sp110 [52] has been described. The mechanism
explaining VOD in this condition remains unknown. One The authors declare that they have no competing interest.
among the possible explanations could be a concurrent
opportunistic viral infection affecting the endothelium of
sinusoids or central vein. References
Gemtuzumab-ozogamycin is a monoclonal antibody tar-
geting CD34 on the surface of myeloid cells but also [1] Stuart KL, Bras G. Veno-occlusive disease of the liver. QJM
sinusoidal endothelial cells. This antibody is conjugated with 1957;26:291—315.
a potent cytotoxic agent. According to a recent report on a [2] Brooks SE, Miller CG, McKenzie K, Audretsch JJ, Bras G. Acute
large survey, the incidence of SOS/VOD was about 9% when veno-occlusive disease of the liver. Fine structure in Jamaican
this agent was used for the treatment of acute myeloid children. Arch Pathol 1970;89:507—20.
leukemia, and the mortality rate related to SOS/VOD was [3] Edgar JA, Molyneux RJ, Colegate SM. Pyrrolizidine alkaloids:
2.7% [53]. Histopathological findings in such a patient are potential role in the etiology of cancers, pulmonary hyperten-
presented in Fig. 2. sion, congenital anomalies, and liver disease. Chem Res Toxicol
2015;28:4—20.
In summary, drugs which are simultaneously toxic for
[4] Tandon RK, Tandon BN, Tandon HD, Bhatia ML, Bhargava S, Lal
hepatic endothelia and hematopoietic cells appear to be at
P, et al. Study of an epidemic of venoocclusive disease in India.
risk of inducing SOS/VOD. However, conditions character- Gut 1976;17:849—55.
ized by a dysregulated immunity, and therefore indications [5] Lin G, Wang JY, Li N, Li M, Gao H, Ji Y, et al. Hepatic sinusoidal
for immunosuppressive agents including azathioprine also obstruction syndrome associated with consumption of Gynura
appear to be risk factors for sinusoidal changes, which segetum. J Hepatol 2011;54:666—73.
makes the relationship with immunosuppressive agents dif- [6] Edgar JA, Roeder E, Molyneux RJ. Honey from plants containing
ficult to establish. It can be suggested that the combination pyrrolizidine alkaloids: a potential threat to health. J Agric
of several factors is relevant to the pathogenesis of these Food Chem 2002;50:2719—30.
uncommon lesions: an immune-related injury and a drug [7] Robinson O, Want E, Coen M, Kennedy R, van den Bosch C,
Gebrehawaria Y, et al. Hirmi Valley liver disease: a disease
induced injury.
associated with exposure to pyrrolizidine alkaloids and DDT.
J Hepatol 2014;60:96—102.
Conclusions [8] DeLeve LD, McCuskey RS, Wang X, Hu L, McCuskey MK, Epstein
RB, et al. Characterization of a reproducible rat model of
hepatic veno-occlusive disease. Hepatology 1999;29:1779—91.
Sinusoidal changes induced by pyrrolizidine alkaloids and [9] Wang X, Kanel GC, DeLeve LD. Support of sinusoidal endothelial
myeloablative agents are likely the results of a combination cell glutathione prevents hepatic veno-occlusive disease in the
of: i) a direct, dose-related toxicity to sinusoidal and central rat. Hepatology 2000;31:428—34.
venous endothelial cell, and ii) the effect of chemotherapy [10] Deleve LD, Wang X, Tsai J, Kanel G, Strasberg S, Tokes ZA. Sinu-
on bone marrow progenitors of endothelial cells that are soidal obstruction syndrome (veno-occlusive disease) in the rat
mobilized for the repair of sinusoidal lesions. It is difficult to is prevented by matrix metalloproteinase inhibition. Gastroen-
ascertain whether the associated liver cell damage is a pure terology 2003;125:882—90.
result of sinusoidal injury, or a result of combined sinusoidal [11] DeLeve LD, Wang X, Kanel GC, Ito Y, Bethea NW, McCuskey MK,
injury and direct toxicity on liver cells. There is evidence for et al. Decreased hepatic nitric oxide production contributes
to the development of rat sinusoidal obstruction syndrome.
a generalized alteration of endothelia in animal models of
Hepatology 2003;38:900—8.
pyrrolizidine alkaloid toxicity but also in recipients of HSCT, [12] Harb R, Xie G, Lutzko C, Guo Y, Wang X, Hill CK, et al. Bone mar-
when they develop SOS/VOD [54]. These changes include row progenitor cells repair rat hepatic sinusoidal endothelial
features of endothelial activation and apoptosis [54]. Sinu- cells after liver injury. Gastroenterology 2009;137:704—12.
soidal changes found in patients receiving chemotherapy [13] Shulman HM, McDonald GB, Matthews D, Doney KC, Kopecky KJ,
for hepatic metastasis of colorectal adenocarcinoma may Gauvreau JM, et al. An analysis of hepatic venocclusive disease
384 D.-C. Valla, D. Cazals-Hatem

and centrilobular hepatic degeneration following bone marrow colorectal liver metastases before the era of antiangiogenics.
transplantation. Gastroenterology 1980;79:1178—91. Int J Hepatol 2013;2013:314868.
[14] DeLeve LD, Valla DC, Garcia-Tsao G. Vascular disorders of the [31] Rubbia-Brandt L, Audard V, Sartoretti P, Roth AD, Breza-
liver. Hepatology 2009;49:1729—64. ult C, Le Charpentier M, et al. Severe hepatic sinusoidal
[15] Shulman HM, Gooley T, Dudley MD, Kofler T, Feldman R, Dwyer obstruction associated with oxaliplatin-based chemotherapy in
D, et al. Utility of transvenous liver biopsies and wedged patients with metastatic colorectal cancer. Ann Oncol 2004;15:
hepatic venous pressure measurements in sixty marrow trans- 460—6.
plant recipients. Transplantation 1995;59:1015—22. [32] Rubbia-Brandt L, Lauwers GY, Wang H, Majno PE, Tanabe K, Zhu
[16] Carreras E, Granena A, Navasa M, Bruguera M, Marco V, Sierra J, AX, et al. Sinusoidal obstruction syndrome and nodular regen-
et al. Transjugular liver biopsy in BMT. Bone Marrow Transplant erative hyperplasia are frequent oxaliplatin-associated liver
1993;11:21—6. lesions and partially prevented by bevacizumab in patients with
[17] Azar N, Valla D, Abdel-Samad I, Hoang C, Fretz C, Sutton L, hepatic colorectal metastasis. Histopathology 2010;56:430—9.
et al. Liver dysfunction in allogeneic bone marrow transplan- [33] Hubert C, Sempoux C, Humblet Y, van den Eynde M, Zech
tation recipients. Transplantation 1996;62:56—61. F, Leclercq I, et al. Sinusoidal obstruction syndrome (SOS)
[18] Blostein MD, Paltiel OB, Thibault A, Rybka WB. A comparison related to chemotherapy for colorectal liver metastases: fac-
of clinical criteria for the diagnosis of veno-occlusive disease tors predictive of severe SOS lesions and protective effect of
of the liver after bone marrow transplantation. Bone Marrow bevacizumab. HPB (Oxford) 2013.
Transplant 1992;10:439—43. [34] Robinson SM, Wilson CH, Burt AD, Manas DM, White SA.
[19] Carreras E, Granena A, Navasa M, Bruguera M, Marco V, Sierra Chemotherapy-associated liver injury in patients with colorec-
J, et al. On the reliability of clinical criteria for the diagnosis of tal liver metastases: a systematic review and meta-analysis.
hepatic veno-occlusive disease. Ann Hematol 1993;66:77—80. Ann Surg Oncol 2012;19:4287—99.
[20] Mahgerefteh SY, Sosna J, Bogot N, Shapira MY, Pappo O, Bloom [35] van der Pool AE, Marsman HA, Verheij J, Ten Kate FJ, Egger-
AI. Radiologic imaging and intervention for gastrointestinal and mont AM, Ijzermans JN, et al. Effect of bevacizumab added
hepatic complications of hematopoietic stem cell transplanta- preoperatively to oxaliplatin on liver injury and complications
tion. Radiology 2011;258:660—71. after resection of colorectal liver metastases. J Surg Oncol
[21] Carreras E, Bertz H, Arcese W, Vernant JP, Tomas JF, Hagglund 2012;106:892—7.
H, et al. Incidence and outcome of hepatic veno-occlusive [36] Robinson SM, Mann J, Vasilaki A, Mathers J, Burt AD, Oakley
disease after blood or marrow transplantation: a prospec- F, et al. Pathogenesis of FOLFOX induced sinusoidal obstruc-
tive cohort study of the European Group for Blood and tion syndrome in a murine chemotherapy model. J Hepatol
Marrow Transplantation. European Group for Blood and Mar- 2013;59:318—26.
row Transplantation Chronic Leukemia Working Party. Blood [37] Rubbia-Brandt L, Tauzin S, Brezault C, Delucinge-Vivier C,
1998;92:3599—604. Descombes P, Dousset B, et al. Gene expression profiling
[22] DeLeve LD, Wang X, Kuhlenkamp JF, Kaplowitz N. Toxic- provides insights into pathways of oxaliplatin-related sinu-
ity of azathioprine and monocrotaline in murine sinusoidal soidal obstruction syndrome in humans. Mol Cancer Ther
endothelial cells and hepatocytes: the role of glutathione 2011;10:687—96.
and relevance to hepatic venoocclusive disease. Hepatology [38] Agostini J, Benoist S, Seman M, Julie C, Imbeaud S,
1996;23:589—99. Letourneur F, et al. Identification of molecular pathways
[23] Dignan FL, Wynn RF, Hadzic N, Karani J, Quaglia A, Pagliuca involved in oxaliplatin-associated sinusoidal dilatation. J Hep-
A, et al. BCSH/BSBMT guideline: diagnosis and management atol 2012;56:869—76.
of veno-occlusive disease (sinusoidal obstruction syndrome) [39] Robinson SM, Mann DA, Manas DM, Oakley F, Mann J, White SA.
following haematopoietic stem cell transplantation. Br J The potential contribution of tumour-related factors to the
Haematol 2013;163:444—57. development of FOLFOX-induced sinusoidal obstruction syn-
[24] Zhang L, Wang Y, Huang H. Defibrotide for the prevention drome. Br J Cancer 2013;109:2396—403.
of hepatic veno-occlusive disease after hematopoietic stem [40] Marzano C, Cazals-Hatem D, Rautou PE, Valla DC. The sig-
cell transplantation: a systematic review. Clin Transplant nificance of nonobstructive sinusoidal dilatation of the liver:
2012;26:511—9. impaired portal perfusion or inflammatory reaction syndrome.
[25] Corbacioglu S, Cesaro S, Faraci M, Valteau-Couanet D, Gruhn Hepatology 2015.
B, Rovelli A, et al. Defibrotide for prophylaxis of hepatic [41] Arotcarena R, Cales V, Berthelemy P, Parent Y, Malet M,
veno-occlusive disease in paediatric haemopoietic stem-cell Etcharry F, et al. Severe sinusoidal lesions: a serious and over-
transplantation: an open-label, phase 3, randomised controlled looked complication of oxaliplatin-containing chemotherapy?
trial. Lancet 2012;379:1301—9. Gastroenterol Clin Biol 2006;30:1313—6.
[26] Richardson PG, Corbacioglu S, Ho VT, Kernan NA, Lehmann L, [42] van Iersel LB, de Leede EM, Vahrmeijer AL, Tijl FG, den
Maguire C, et al. Drug safety evaluation of defibrotide. Expert Hartigh J, Kuppen PJ, et al. Isolated hepatic perfusion with
Opin Drug Saf 2013;12:123—36. oxaliplatin combined with 100 mg melphalan in patients with
[27] Valla DC. Budd-Chiari syndrome and veno-occlusive dis- metastases confined to the liver: A phase I study. Eur J Surg
ease/sinusoidal obstruction syndrome. Gut 2008;57:1469—78. Oncol 2014;40:1557—63.
[28] Van Cutsem E, Cervantes A, Nordlinger B, Arnold D. Metastatic [43] Lehmann K, Rickenbacher A, Weber A, Pestalozzi BC, Clavien
colorectal cancer: ESMO Clinical Practice Guidelines for diag- PA. Chemotherapy before liver resection of colorectal metas-
nosis, treatment and follow-up. Ann Oncol 2014;25(Suppl. tases: friend or foe? Ann Surg 2012;255:237—47.
3):iii1—9. [44] Soubrane O, Brouquet A, Zalinski S, Terris B, Brezault C, Mallet
[29] Vigano L, Capussotti L, De Rosa G, De Saussure WO, Mentha V, et al. Predicting high grade lesions of sinusoidal obstruc-
G, Rubbia-Brandt L. Liver resection for colorectal metastases tion syndrome related to oxaliplatin-based chemotherapy for
after chemotherapy: impact of chemotherapy-related liver colorectal liver metastases: correlation with post-hepatectomy
injuries, pathological tumor response, and micrometastases outcome. Ann Surg 2010;251:454—60.
on long-term survival. Ann Surg 2013;258:731—40 [discussion [45] Vreuls CP, Van Den Broek MA, Winstanley A, Koek GH, Wisse
741-732]. E, Dejong CH, et al. Hepatic sinusoidal obstruction syndrome
[30] Nguyen-Khac E, Lobry C, Chatelain D, Fuks D, Joly JP, Brevet (SOS) reduces the effect of oxaliplatin in colorectal liver metas-
M, et al. A reappraisal of chemotherapy-induced liver injury in tases. Histopathology 2012;61:314—8.
Sinusoidal obstruction syndrome 385

[46] Vreuls CP, Olde Damink SW, Koek GH, Winstanley A, Wisse disease/sinusoidal obstruction syndrome after liver transplan-
E, Cloots RH, et al. Glutathione S-transferase M1-null geno- tation. Liver Transpl 2011;17:798—808.
type as risk factor for SOS in oxaliplatin-treated patients [51] Takamura H, Nakanuma S, Hayashi H, Tajima H, Kakinoki K,
with metastatic colorectal cancer. Br J Cancer 2013;108: Kitahara M, et al. Severe veno-occlusive disease/sinusoidal
676—80. obstruction syndrome after deceased-donor and living-donor
[47] van den Broek MA, Vreuls CP, Winstanley A, Jansen RL, van liver transplantation. Transplant Proc 2014;46:3523—35.
Bijnen AA, Dello SA, et al. Hyaluronic acid as a marker [52] Wang T, Ong P, Roscioli T, Cliffe ST, Church JA. Hepatic veno-
of hepatic sinusoidal obstruction syndrome secondary to occlusive disease with immunodeficiency (VODI): first reported
oxaliplatin-based chemotherapy in patients with colorectal case in the U.S. and identification of a unique mutation in
liver metastases. Ann Surg Oncol 2013;20:1462—9. Sp110. Clin Immunol 2012;145:102—7.
[48] Shin NY, Kim MJ, Lim JS, Park MS, Chung YE, Choi JY, et al. Accu- [53] Tallman MS, McDonald GB, DeLeve LD, Baer MR, Cook MN, Grae-
racy of gadoxetic acid-enhanced magnetic resonance imaging pel GJ, et al. Incidence of sinusoidal obstruction syndrome
for the diagnosis of sinusoidal obstruction syndrome in patients following Mylotarg (gemtuzumab ozogamicin): a prospective
with chemotherapy-treated colorectal liver metastases. Eur observational study of 482 patients in routine clinical practice.
Radiol 2012;22:864—71. Int J Hematol 2013;97:456—64.
[49] Han NY, Park BJ, Sung DJ, Kim MJ, Cho SB, Lee CH, et al. [54] Vion AC, Rautou PE, Durand F, Boulanger CM, Valla DC. Interplay
Chemotherapy-induced focal hepatopathy in patients with of inflammation and endothelial dysfunction in bone mar-
gastrointestinal malignancy: gadoxetic acid–enhanced and row transplantation: Focus on hepatic veno-occlusive disease.
diffusion-weighted MR imaging with clinical-pathologic corre- Semin Thromb Hemost 2015.
lation. Radiology 2014;271:416—25. [55] Snover DC, Weisdorf S, Bloomer J, McGlave P, Weisdorf D. Nodu-
[50] Sebagh M, Azoulay D, Roche B, Hoti E, Karam V, Teicher lar regenerative hyperplasia of the liver following bone marrow
E, et al. Significance of isolated hepatic veno-occlusive transplantation. Hepatology 1989;9:443—8.