- Email: [email protected]
Obliterative portal venopathy: Portal hypertension is not always present at diagnosis
Research Article
Obliterative portal venopathy: Portal hypertension is not always present at diagnosis Dominique Cazals-Hatem1,⇑, , Sophie Hillaire2, , Marika Rudler2, Aurélie Plessier2, Valerie Paradis1, Bertrand Condat2, Claire Francoz2, Marie-Hélène Denninger3, François Durand2, Pierre Bedossa1, Dominique C. Valla2 1
Service d’Anatomie-Pathologique, Hôpital Beaujon AP-HP, Clichy and Institut National de la Santé et de la Recherche Médicale U773, Centre de Recherche Biomédicale Bichat Beaujon CRB3, Université Denis Diderot-Paris 7, France; 2Centre de Référence Maladies Rares, Maladies Vasculaires du Foie, Service d’Hépatologie, Hôpital Beaujon AP-HP, Clichy and Institut National de la Santé et de la Recherche Médicale U773, Centre de Recherche Biomédicale Bichat Beaujon CRB3, Université Denis Diderot-Paris 7, France; 3Service d’Hématologie biologique, Hôpital Beaujon AP-HP, Clichy and Institut National de la Santé et de la Recherche Médicale U773, Centre de Recherche Biomédicale Bichat Beaujon CRB3, Université Denis Diderot-Paris 7, France
Background & Aims: Previous studies on obliterative portal venopathy (OPV) have been biased due to the selection of patients with non-cirrhotic portal hypertension. The aim of this study was to clarify the characteristics of OVP diagnosed by liver biopsy. Methods: Fifty-nine consecutive patients with OPV were retrospectively selected on strict histological criteria. Clinical, laboratory, portal vein patency, and associated disorders potentially involving vascular alterations were analyzed. The occurrence of complications was recorded during follow-up. Results: Mean age at diagnosis was 38.5 ± 15 years old. Initial presentation was portal hypertension (64% of patients) and/or extrahepatic portal vein thrombosis (EHPVT) (22%) or isolated abnormal laboratory tests (20%). Associated diseases found at diagnosis were: prothrombotic disorders (30% of patients) and immune-mediated disorders (17%); 53% of patients had no causal factor (idiopathic OPV). During follow-up (median 8.6 years, range 1–23 years), features of portal hypertension worsened in 46% of patients; EHPVT and portal hypertension were finally found in 44% and 88% of patients. Anti-coagulation and betablockers were administered in 47% and 59% of patients, respectively. Severe complications (liver transplantation and/or death) occurred in 11 (19%) patients, 8 had idiopathic OPV. Patients with prothrombotic disorders received earlier anticoagulation therapy; all survived without transplantation.
Keywords: Vascular liver disease; Idiopathic portal hypertension; Hepatoportal sclerosis; Portal vein thrombosis; Liver biopsy. Received 20 March 2010; received in revised form 2 July 2010; accepted 5 July 2010; available online 28 October 2010 ⇑ Corresponding author. Tel.: +33 1 40 87 58 61; fax: +33 1 40 87 00 77. E-mail address: [email protected] (D. Cazals-Hatem). These authors contributed equally to this work. Abbreviations: OPV, obliterative portal venopathy; NCPH, non-cirrhotic portal hypertension; CPT, complete portal tract; NRH, nodular regenerative hyperplasia; HIV, human immunodeficiency virus; EHPVT, extrahepatic portal venous thrombosis; OLT, orthotopic liver transplantation; ALT, alanine aminotransferase; GGT, gamma glutamyl transferase; ULN, upper limit of normal value; JAK, Janus kinase.
Conclusions: A confident diagnosis of OPV can be done by biopsy and is conceivable in patients under 40 years without clinically significant portal hypertension. Poor outcome was noted in 19% of patients, most of them affected with idiopathic OPV. Patients with prothrombotic disorders received early anticoagulation and appeared to have a better outcome despite a high proportion of EHPVT. Ó 2010 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.
Introduction Obliterative portal venopathy (OPV) was coined to describe primary occlusion of the intrahepatic portal veins in the absence of cirrhosis, inflammation, and hepatic neoplasia [1,2]. This disease has been reported in patients with non-cirrhotic portal hypertension (NCPH) and because of the lack of extrahepatic portal vein obstruction, intrahepatic portal veins were studied. This entity was first described as an intrahepatic portal phlebosclerosis and the term hepatoportal sclerosis was used [3]. Portal vasculopathy predominated in medium-sized and preterminal veins (0.2–3 mm in diameter) which are easily accessible on liver biopsies [4,5]. However, in the literature, OPV has mainly been described in autopsies and livers explanted for NCPH or cryptogenic cirrhosis without portal vein thrombosis [6–9]. A positive diagnosis of OPV on biopsy requires specific histological findings composed of a spectrum of lesions that have been well described [10,11]. We have previously reported this disease diagnosed by biopsy in Caucasian patients, but all patients were selected based on the presence of portal hypertension preventing a relevant analysis of disease progression [12]. OPV probably exists before the signs of portal hypertension appear and it seems important to recognize it beforehand and to follow patients afterwards; many aspects of OPV diagnosed earlier on needle biopsies remain unclear and need to be re-evaluated.
Journal of Hepatology 2011 vol. 54 j 455–461
Research Article The aim of this study was to assess the clinical and histological features at presentation, the associated disorders, and the outcome of patients in whom a diagnosis of OPV was obtained via biopsy.
Patients and methods Patient selection From January 1987 to December 2007, all liver biopsies with a diagnosis of ‘‘primary OPV’’ were retrospectively selected from the database of our pathology department. Tissues were fixed in 10% formalin solution and paraffin sections were stained with H&E, trichrome, and argentation stains. The number of complete portal tracts (CPT) was counted in each specimen and portal veins were assessed. Ninety-eight specimens were available from 76 patients (18 patients underwent several biopsies). Specimens, reviewed by the same pathologist (DCH), had to fulfill 3 criteria: (i) to be longer than 1 cm and to contain more than 5 CPT; (ii) to have an alternation of CPT and centrilobular veins to exclude cirrhosis; (iii) to have more than 2/3 (66%) of the CPT harbouring abnormal portal venules, defined as absent or clearly reduced in caliber with sclerosis or thickening of the smooth muscle wall [4,5]. According to Krasinskas et al. [13], abnormal portal vessels corresponding to those described in NCPH can be present in greater than 25% of CPT in ‘‘normal livers with normal portal pressure’’. Therefore our latter criterion allowed the diagnosis of OPV to be made with confidence. Seventeen patients (22% of the entire cohort) were excluded as their needle biopsies contained less than 6 CPT (mean CPT: 3 [2–5]) and/or measured less than 1 cm in length (mean 7 mm [4–14]). Therefore, 59 patients were eligible corresponding to 82 biopsy specimens. Histological evaluation Additional morphological changes classically reported with OPV were recorded to assess their prevalence on biopsy specimens [4,5,8,10,11]: (a) nodular regenerative hyperplasia (NRH) defined as irregular thickened hepatic plates with a nodular appearance alternating with atrophic compressed plates without fibrous septa (b) sinusoidal dilatation; (d) perisinusoidal fibrosis detected on argentation stain; (e) aberrant vessels located close to or distant from portal tracts welldescribed by Ohbu et al. [15]; and (f) extensive portal fibrosis (>grade F1 adapted from METAVIR classification). Explanted livers available during follow-up were processed routinely: gross features and microscopic features were recorded the same way as for biopsies, in addition to specific data: weight of liver explants, dysmorphic appearance of livers (defined by atrophy and/or hypertrophy of a lobe and/or nodular contours), endophlebitis of large portal veins (>3 mm in diameter) and of large hepatic veins (>2 mm, hepatic venopathy), and macroscopic nodules (larger than 5 mm).
Results Initial presentation At the diagnosis of OPV, the 59 selected patients were 38.5 ± 15 years (range 7–77 years), 20 women (33%), and 39 men (66%). All patients had patent hepatic veins at Doppler-ultrasound and no evidence of neoplasia. The distribution of patients according to the type of initial presentation is depicted in Fig. 1. A clinically significant portal hypertension was present in 38 patients (64%) and absent in 21 patients (36%). EHPVT was present in 13 patients (22%) at diagnosis, including seven patients with acute thrombosis, and six with cavernous transformation; all six patients with portal cavernoma and one patient with acute portal vein thrombosis also had features of portal hypertension. Portal hypertension and EHPVT were absent in 15 patients (25%) at the initial diagnosis. The main indicators for biopsy were the appearance of portal hypertension (38 patients, detailed Fig. 1* according to the predominant symptoms initially presented), recent detection of EHPVT generally associated with
N = 59 Patients with OPV diagnosed by biopsy
N = 38 Portal Hypertension *
Collection of other data Clinical, laboratory, and imaging features at the time of initial diagnosis and during follow-up and treatments were recorded. Conditions considered to cause OPV were then identified from the patient’s chart based on conventional diagnostic criteria: professional exposure to copper sulphate or vinyl chloride monomer; previous exposure to thorium sulphate, Spanish toxic oil or arsenic salts, regular vitamin A supplements, autoimmune or collagen vascular diseases, primary biliary cirrhosis or primary sclerosing cholangitis, schistosomiasis, sarcoidosis, abdominal or systemic sepsis and prothrombotic disorders. Tests for antithrombin, protein S, protein C deficiencies, for the detection of antiphospholipid antibodies, hyperhomocysteinemia had been routinely performed according to previously reported methods [16]. Activated protein C resistance, Factor V Leiden (G1691A factor V gene mutation), G202100A Factor II gene mutation, and V617 JAK2 mutation were also assessed shortly after they were described. Patients seen before these molecular tests became available were tested later on at follow-up visits or using frozen stored blood samples. HIV serology status was determined. The presence of a congenital/genetic syndrome was explored. Additional causes of chronic liver diseases were identified according to conventional diagnostic criteria (chronic viral hepatitis, alcoholic liver disease, non-alcoholic steatohepatitis, haemochromatosis, autoimmunue hepatitis, Wilson disease, and biliary diseases). All patients who underwent a liver biopsy have been systematically explored by Doppler ultrasound. An imaging study completed with contrast-enhanced computerized tomography (CT) scan was routinely performed at the time of liver disease diagnosis: they evaluated portal vein patency size and flow, gross collat-
456
eral portosystemic circulation, ascites, and spleen enlargement defined as a spleen size >12 cm. They were routinely repeated during follow-up based on our previously reported data [12]. Upper endoscopy was also performed at the time of diagnosis and during follow-up to check for the deterioration of portal hypertension. Clinically significant portal hypertension was defined as: (i) oesophageal varices assessed on upper endoscopy (with or without bleeding) or (ii) gross collateral portosystemic circulation or (iii) splenomegaly with portal vein dilatation at imaging, or (iii) ascites or pleural effusion, or (iv) portosystemic encephalopathy. Deterioration of portal hypertension was defined as the occurrence of varices if previously absent or if increased in size by one grade as assessed by upper endoscopy (with or without bleeding), presence of ascites or spleen enlargement if absent at the previous examination.
N = 21 No Portal Hypertension *
N=7 N=6 Extra-hepatic Portal Vein Thrombosis (EHPVT) 7 Portal Cavernoma 6 Acute Portal Vein Thrombosis
Portal hypertension w/o EHPVT N = 31 (64%)
EHPVT N = 13 (22%)
No Portal hypertension and no EHPVT N = 15 (25%)
Fig. 1. Flowchart of the clinical presentation of the 59 patients at the time of initial diagnosis of OPV. *Main clinical manifestations of portal hypertension initially presented were: Varices bleeding n = 12, non-bleeding varices n = 7, splenomegaly n = 10, ascitis or pleural effusion n = 8, encephalopathy n = 1.
Journal of Hepatology 2011 vol. 54 j 455–461
JOURNAL OF HEPATOLOGY abnormal liver tests and/or discovered with prothrombotic disorders (six patients) or unexplained and isolated chronic liver test abnormalities (12 patients, 2 had cholestasis during pregnancy). The need for biopsy was unclear for three patients. Patients with portal hypertension were younger than those without it at diagnosis (mean age 30 ± 16 years versus 44 ± 14 years, p = 0.01). Most patients (n = 46, 78%) had laboratory abnormalities although marked abnormalities were uncommon (Table 1): only 14% of patients had serum alanine aminotransferase (ALT) >2 times the upper limit of normal value (ULN) or alkaline phosphatases >2 ULN or a prothrombin index <50% of normal. Most patients (56%) had a platelet count <150,000/mm3; while 10% of patients had thrombocytosis due to a myeloproliferative disease. Total bilirubin was normal (<17 lm/L) in 70% of patients (median 16 lm/L, [range 16–190]). Normal laboratory tests were observed in 13 of the 59 patients at diagnosis (22%). Associated disorders After a complete screening, associated disorders potentially involving vascular alterations were found in 28 patients (47%) divided into two groups according to the existence of prothrombotic (n = 18) or immune-mediated disorders (n = 10). Clinical and biological findings at the initial diagnosis and during follow-up were detailed for the entire cohort (n = 59) and according to associated disorders (Table 2). Ten patients (17%) had concomitant immune-mediated disorders (Table 3): four patients had autoimmune hepatitis identified by biopsies before the diagnosis of OPV (no extensive fibrosis at that time) and treated successfully with corticoids (n = 2) and azathioprine (n = 2). One patient had primary biliary cirrhosis without extensive fibrosis that was treated successfully with ursodeoxycholic acid and associated with an antiphospholipid syndrome. Three patients had rheumatoid arthritis (one with autoimmune hepatitis). A common variable immunodeficiency (treated with iterative immunoglobulin injections), a Wegener Table 1. Laboratory data at the time of histological diagnosis.
Biology
ALT
Alkaline phosphatase (N = 55)
GGT
Platelets
Prothrombin index
Values
Patients N = 59
%
28 23 7 1 24 23 8 18 21 20 22 11 20 6 8 33 18
47 39 12 2 44 42 14 31 36 34 37 19 34 10 14 56 30
<1 ULN 1-2 ULN 2-10 ULN >10 ULNN <1 ULN 1-2 ULN >2 N <1 ULN 1-4 ULN >4 N <90,000/mm3 90-150,000/mm 3 150-350,000/mm 3 >350,000/mm3 <50% 50-80% >80%
ALT, alanine aminotransferase; GGT, gamma glutamyl transferase; ULN, times upper the limit of normal value.
disease, and a Sharp syndrome were identified in one patient respectively. Anticardiolipid antibodies were found in six patients and two presented with an antiphospholipid syndrome. Investigations for prothrombotic disorders were obtained in 55 of the 59 patients: 18 patients (33% of patients tested) had prothrombotic disorders (Table 4); four patients had at least two prothrombotic disorders. The V617K JAK2 mutation was detected in 6 of the 30 patients tested. An isolated protein S deficiency was strongly suspected in three patients but could not be confirmed because a family study was impossible. One woman had both a factor II gene mutation and Turner’s syndrome. One patient had both an autoimmune disease and a protein S deficiency. No prothrombotic disorder or immune-mediated disorder could be identified in 53% of the patients (n = 31). None was HIV+ and no systemic or intra-abdominal infection was found. Repeated interviews excluded exposure to toxins potentially involving vascular alterations. Two patients had stigma of hepatitis C viral infection and presented with portal vein thrombosis without cryoglobulinemia: OPV was diagnosed on histology when the METAVIR evaluations were A1F1 in both. Treatment and outcome As shown in Table 2, during a median follow-up of 8.6 years [range 1–23], 40% of the 15 patients without initial portal hypertension or EHPVT, developed portal hypertension (varices in five, ascites in one), associated with an acute EHPVT in two patients; two of them were transplanted. Thirteen of the 46 patients (28%) without initial EHPVT developed an EHPVT detected at imaging: four patients had a prothrombotic disorder, three had an immune-mediated disorder. The deterioration or occurrence of portal hypertension was observed in 27 (46%) patients with variceal bleeding in seven (varices in 16, ascites in 11). At the end of follow-up, 77% of patients with thrombophilia developed EHPVT, versus 40% of patients with immune-mediated disorders and versus 26% in patients with no evidence of associated disorders. Lastly, clinically significant portal hypertension and EHPVT could be detected in 88% and 44% of all patients, respectively. In the entire cohort only seven patients had no evidence of portal hypertension and portal vein thrombosis at the end of followup, and all had a normal spleen at imaging. Thirty-five patients (59%) were treated with beta-blockers and this was combined with endoscopic therapy for five patients to check for gastrointestinal bleeding. Beta-blockers were discontinued in two of these patients because of its side-effects. Three patients also underwent surgical portocaval shunting for recurrent bleeding. Twenty-eight patients (47%) received anticoagulant therapy: indications for treatment were EHPVT in 24 patients, discovered at the initial diagnosis (n = 12) or during follow-up (n = 12), and a prothrombotic disorder without obvious thrombosis in the remaining four patients. Severe complications were observed in two patients (recurrent gastrointestinal bleeding and lethal hemorrhagic stroke). The interval between the diagnosis of OPV and anticoagulation varied according to the underlying disease, being shorter in patients with prothrombotic disorders than in others (Table 2). Nine patients (15%) underwent orthotopic liver transplantation (OLT) a median of 8.5 years [1–16] after OPV diagnosis at a median age of 49 years (27–59). All transplanted patients had severe portal hypertension which was associated with severe
Journal of Hepatology 2011 vol. 54 j 455–461
457
Research Article Table 2. Clinical findings at initial diagnosis and during follow-up of the entire cohort and of patients distributed according to associated disorders potentially involved in vascular alterations.
Associated disorders N(%)
All patients N = 59
Prothrombotic disorders 18 (30)
Immune- mediated disorders 10 (17)
Other patients 31 (53)
Initial Presentation Mean Age (Years old) PHT w/o Portal Thrombosis Extrahepatic Portal VeinThrombosis Isolated Abnormal Hepatic Tests Abdominal Pain
38 ± 15 31 (53) 13 (22) 12*(20) 3 (5)
40 ± 17 5 (28) 10 (55) 3 (17) 0
41 ± 9 4 (40) 1 (10 ) 4 (40) 1 (10)
36 ± 9 22 (71) 2 (6) 5 (16) 2 (6)
Outcome Follow-up (Years) Extrahepatic Portal Vein Thrombosis Deterioration/occurrence of PHT Liver Transplantation Death
8.6 ± 7.8 13/46 (28) 27 (46) 9 (15) 5 (8)
5 ± 5.8 4/8 (50) 6 (33) 0 0
13 ± 7 3/9 (33) 6 (60 ) 1 (10 ) 2 (20)
9±8 6/29 (21) 15 (48) 8 (26) 3 (10)
Treatment Anticoagulant (AC) Delay diagnosis/ AC therapy (Years)
28 (47) 2.3
13 (72) 0.7
5 (50) 4.6
10 (33) 3.6
PHT, portal hypertension; *two patients had cholestasis of pregnancy.
Table 3. Immune-mediated disorders associated with OPV.
Immune-mediated disorders Autoimmune hepatitis Primary biliary cirrhosis Rheumatoid arthritis Common variable immunodeficiency Wegener disease Sharp syndrome Total
Number of patients (%) 4* 1 3 1 1 1 10 / 59 (17%)
*One patient had also rheumatoid arthritis.
Table 4. Results of investigations for prothrombotic disorders performed in 55 out of the 59 patients.
Prothrombotic disorders Myeloproliferative disease (MPD)† Factor II deficiency Protein S deficiency Protein C deficiency MTHFR homozygote mutation Total
Number of patients (%) 10 3 3 3 2 18*/55 tested (33%)
V617F JAK2 mutation in six patients. *Four patients had more than one prothrombotic disorder: 2 had MPD and MTHFR homozygote mutation, 1 had MPD and protein S deficiency, 1 had a factor II mutation and a protein S deficiency.
liver dysfunction in seven of those patients and hepatopulmonary syndrome in two of the patients. The liver failure occurred in the context of sepsis/or oesophageal bleeding in five of the seven patients. Only one patient had an EHPVT treated with anticoagulation. Eight of the nine transplanted patients had idiopathic OPV: no etiological factor or associated disorder potentially involving vascular alterations was identified (one of them underwent both renal and hepatic transplantation for familial idiopathic renal 458
insufficiency); the last patient transplanted had an immunerelated disorder. During follow-up, five patients died (median age of 51 years old (29–62)): three of them were recently transplanted and died in the year post-OLT due to hepatopulmonary syndrome, sepsis, and multiorgan failure. One patient died with sepsis and hepatocellular deficiency and the last one with a hemorrhagic stroke after anticoagulation. Finally 11 of the 59 patients (19%) had an unfavorable outcome (OLT in nine, death in five, and both OLT and death in three) at a median age of 51 years [27–62]. Among them nine patients had no anticoagulation and two were receiving anticoagulation. Pathological data Among the 82 liver specimens obtained from the 59 patients selected, 73 were needle biopsies and nine surgical biopsies. Eighteen patients (30.5%) needed several biopsies to confirm the diagnosis of OPV. Transjugular biopsies were performed in 17 patients (29%) and a hemodynamic study was therefore available in these 17 patients. The median hepatic venous pressure gradient was 7 mm Hg [2–17] and 35% of patients had a normal pressure gradient (<5 mm Hg). The median length of needle biopsy specimens was 13.7 mm [5–30]. An average of 11 CPTs (6–40) were present per biopsy. Overall, 703 (79%) of the 885 CPTs analyzed as a whole contained abnormal portal venules. Most CPTs were small, sclerotic, and round without inflammation (Fig. 2). The argentation stain was essential to assess inconspicuous NRH changes and architectural disturbances. All biopsies had at least one additional lesion and 76% had at least two lesions as shown in Table 5 and Fig. 2. At the end of follow-up, 10 patients (17%) developed macronodules at imaging (1–3 each), most <3 cm in diameter and having the appearance of focal nodular hyperplasia (biopsy in seven patients). None developed hepatocellular carcinoma. The explanted livers of the nine patients transplanted during follow-up were fully examined: most were atrophic (median
Journal of Hepatology 2011 vol. 54 j 455–461
JOURNAL OF HEPATOLOGY Table 5. Histological features observed on biopsies and explants.
Histological lesions
Biopsy specimen N = 82 (%) Mean size°= 13.7mm [5-30]
Obliteration of portal venules Nodular regenerative hyperplasia Perisinusoidal fibrosis Sinusoidal dilatation Aberrant vessels Extensive portal fibrosis ¤ (>F1) Dysmorphic liver Obliterated large portal veins Hepatic venopathy
79* (96) 57 (70) 47 (57) 34 (41) 31 (38) 25 (30)
Explanted livers N = 9 (%) Mean weight= 837g [610-1,640] 9 (100) 9 (100) 9 (100) 6 (67) 5 (55) 8 (89) 6 (67) 6 (67) 2 (22)
Needle biopsies only (surgical biopsies not included); *3 needle biopsies displayed a nodular regenerative hyperplasia aspect with scarce portal tracts: repeated biopsies done in these patients asserted obliterative portal venopathy (OPV) diagnosis; ¤evaluation adapted from METAVIR classification.
Fig. 2. Histological features of obliterative portal venopathy diagnosed on needle biopsies. (A–C) Same specimen, trichrome stains 25, 100, 200. (A) Representative hepatic sample displaying >66% of complete portal tract with abnormal venules, (B) sclerotic portal venules with occlusion of the portal vein radicle, (C) a aberrant vessel present around the portal tract. (D and E) Same specimen, trichrome stains 25, 100. (D) Sample showing alternation of small fibrous portal tracts prolonged by thin septa and dilated veins, (E) most of the dilated veins were aberrant vessels. (F,G, and H) Same specimen, trichrome 25, 100, argentation stains 100). (F) Sample showing extensive portal fibrosis, (G) mixed with nodular regenerative hyperplasia in non-fibrous zones, (H) with sinusoidal fibrosis, thin septa, and aberrant vessels.
weight: 837 g (610–1,640)) and six were dysmorphic (Table 5 and Fig. 3). Only one liver exhibited two macronodules (<1 cm large) without neoplasia. Microscopically, all livers displayed diffuse OPV without cirrhosis. Extensive portal fibrosis was present in 8 of the 9 explants, more pronounced in perihilar (grade F2 in 6 and grade F3 in 2). Hepatic venopathy was found in 22% of livers. Large portal vein endophlebitis related to preexistent parietal thrombosis were present in 67% of livers. A ductopenia with bilirubin accumulation was observed in two patients: one of them was in an immune-related context and the other had idiopathic
Fig. 3. Gross and microscopic findings of livers explanted for obliterative portal venopathy. (A and B) same explant. (A) Gross view showing no dysmorphy; (B) trichrome stain 100 showing hepatoportal sclerosis with extensive fibrosis (F2) and sclerotic portal veins. (C,D) same explant. (C) Macroscopic view of a liver cut section showing perihilar nodular transformation without cirrhotic aspect, (D) trichrome stain 100 showing small sclerotic portal tracts and periportal nodular regenerative changes.
OPV combined with renal insufficiency. No active cholangitis or hepatitis was identified.
Discussion The present study handled a large cohort of patients affected with a rare and under-appreciated vascular disease: it provided histological, clinical, and prognostic features of OPV diagnosed by liver biopsy. Our inclusion criteria were based on the histological identification of primary occlusion of the portal microvasculature after
Journal of Hepatology 2011 vol. 54 j 455–461
459
Research Article ruling out cirrhosis, active inflammation, and/or metabolic liver diseases, regardless of initial manifestations. Although it was frequently associated, NRH per se was not accepted as a criterion of portal vasculopathy since it’s a secondary change and other vascular lesions have been proposed in its pathogenesis [5]. However, NRH changes constituted a useful guidance for the diagnosis of OPV since it was present in 70% of specimens. The 17 patients excluded because of insufficient CPT displayed NRH and were managed as the 59 patients described in the study: their features at presentation and during follow-up were similar to those of the 59 included patients (see Table 6 in supplementary data). In this series 30.5% of patients underwent several biopsies to meet our requirements as far as the number or CPT is concerned for a fair evaluation of portal veins. Examinations of explanted livers confirmed OPV in all patients, supporting the diagnostic criteria applied to biopsy. The complete histological work-up corroborated the heterogeneity of lesions in a single liver and between individuals as previously described [8,9,11,14]; overlaps between NRH and hepatoportal sclerosis reinforce the assumption that both belong to a same spectrum with OPV as a pivotal component. The first aim of our study was to describe the clinical presentation of OPV diagnosed by biopsy. Manifestations were heterogeneous and clinically significant portal hypertension was lacking in 36% of patients: only 17% and 10% of patients had splenomegaly or ascites, respectively, which contrasts with previous cohorts on NCPH clinically defined by portal hypertension and massive splenomegaly [6,7]. Patients with initial portal hypertension were younger than those without, which eliminated the effect of age at diagnosis. The fact that 25% of our patients initially presented without evidence for portal hypertension or portal thrombosis was also unusual compared to previous data on NCPH [7,12,17]. Many patients underwent a liver biopsy for slight-to-moderate liver test alterations and the diagnosis of OPV occurred by chance. Documented portal thrombosis was the initial sign in 55% of the patients with underlying thrombophilia, which should have prompted the systematic screening for causes. Secondly, we identified disorders potentially involved in vasculopathy in 47% of our patients: concomitant thrombotic or immune-mediated disorders were found in 30% and 17% of patients, respectively. These figures are higher than that expected in the general population (<15% and <12%, respectively) [18,19]. These findings were primarily descriptive but did not provide a mechanistic approach, which constitutes a limitation. The role of thrombophilia in the mechanism of OPV is well established [5,12]. On the other hand, the role of immune-mediated disorder in OPV is not well-documented in the process of portal vasculopathy but frequently proposed in the process of NRH [4,20,21]. A Japanese study on 160 patients with NCPH reported up to 11.9% of associated non-hepatic autoimmune disorders: thyroiditis, systemic lupus, and Sjogren syndrome were the most frequent [22]. OPV and more commonly NRH have been described in association with chronic inflammatory and immune-disorders disorders such as sarcoidosis, primary biliary cirrhosis, HIV infection, and common variable immunodeficiency [21,23–25]. The mechanisms of OPV have not yet been clarified: an inflammatory reaction close to the portal endothelium could cause endothelitis, thrombosis, and venular sclerosis in inflammatory disease such as primary biliary cirrhosis or sarcoidosis; otherwise no sign of vascularitis has been described. Thrombosis could be favored by
460
hyperviscosity induced by the marked hypergammaglobulinemia observed in autoimmune diseases. Two patients in our cohort were treated with azathioprine for autoimmune hepatitis or Sharp syndrome; but the role of this agent in OPV development remains unclear. Azathioprine treatment may generate NRH in inflammatory bowel disease but its role in OPV has not yet been confirmed [26]. Finally, 53% of our patients presented with ‘‘idiopathic OPV’’: histological examination of liver explants did not provide etiological indications. Clinically, they were younger (mean age at presentation = 36) and had severe portal hypertension requiring OTL in a quarter of them before the age of 50. In a recently reported series, the mean age of patients transplanted for progressive OPV varied from 47 years (n = 19) to 60 years (n = 8) [8,9]. The young age of our patients transplanted for idiopathic OPV might be due to complex pathological processes other than hypercoagulability, including genetic or congenital malformations. Indeed, OPV has been described in Turner’s syndrome suggesting the role of abnormal development or vascular malformations [27]; one patient in this series had Turner’s syndrome but she also had a prothrombotic disorder. Limited family studies have suggested the existence of familial hepatoportal sclerosis, incomplete septal cirrhosis, or NRH [28] but no familial clustering was found in the present cohort. Our last aim was to describe the natural course of OPV diagnosed by biopsy. After a mean follow-up of 8.6 years, the clinical course of the disease was severe in 19% of patients leading to transplantation or death. Deterioration or occurrence of portal hypertension was observed in nearly half of the patients, and at the end of follow-up evidence of portal hypertension and EHPVT were finally observed in 88% and 44% of patients, respectively. However, this study had limitations since it was retrospective, thus treatment had affected outcome; patients were heterogeneous and managed differently according to their etiology. Globally, the clinical course of patients was favorable thanks to appropriate management with betablockers. Patients with underlying prothrombotic disorders experienced portal thrombosis (77%) but had no severe complications since none needed transplantation or died. This favorable course could be influenced by the rapid administration of anticoagulants. By contrast, 25% of patients with idiopathic OPV evolved unfavorably without clear explanation; they received no anticoagulants but a causal link between both looked hazardous since patients were too heterogeneous to be compared each other. Our data do not allow to establish whether different disorders or the use of anticoagulation therapy can explain the differences in outcome between the ‘‘idiopathic’’ and the ‘‘prothrombotic OVP’’. Most transplant patients did not show EHPVT at imaging and therefore did not receive anticoagulation before OLT. Nevertheless, histological findings found extensive endophlebitis of the large portal veins in two third of the explants with impaired perfusion leading to liver atrophy and liver failure, probably due to inadequate arterial compensation. Therefore detection of thrombosis at imaging was underestimated and maintaining portal vein patency should be a major goal. Unfortunately, the overall survival post-OLT was poor in our cohort while patients had good outcome after OLT in other series [8,29]. To improve prognosis a trial of anticoagulation therapy could be warranted in all patients whatever the causal factor, after considering the benefit/risk ratio of such a treatment.
Journal of Hepatology 2011 vol. 54 j 455–461
JOURNAL OF HEPATOLOGY Macronodules detected in 17% patients were all benign as previously reported [7,14]. Regular survey is recommended since neoplastic transformation has been described in other chronic vascular diseases [30]. In conclusion, a confident OPV diagnosis can be obtained via biopsy; copious biopsies should be required for adequate portal vessel assessment. OPV had heterogeneous presentations and portal hypertension may be absent at diagnosis in 36% of patients. Early recognition of OPV by pathologists should prompt clinicians to search for immune-mediating or prothrombotic conditions; the latter appeared to have a better outcome when receiving early anticoagulation despite a high proportion of EHPVT. Nevertheless poor prognosis was noted in 19% of patients affected with idiopathic OPV. Latent congenital malformations or genetic susceptibilities may be involved. Published familial forms of OPV should be considered for preventive familial screening.
Conflict of interest The authors who have taken part in this study declared that they do not have anything to disclose regarding funding or conflict of interest with respect to this paper.
Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jhep.2010.07.038.
References [1] Nayak NC, Ramalingaswami V. Obliterative portal venopathy of the liver. Associated with so-called idiopathic portal hypertension or tropical splenomegaly. Arch Pathol 1969;87:359–369. [2] Kameda H. Obliterative portal venopathy: a comparison study of 184 cases of extrahepatic portal obstruction and 469 cases of idiopathic portal hypertension. J Gastroenterol Hepatol 1986;1:139–149. [3] Mikkelsen WP, Edmondson HA, Peters RL, Redeker AG, Reynolds TB. Extraand intrahepatic portal hypertension without cirrhosis (hepatoportal sclerosis). Ann Surg 1965;162:602–608. [4] Wanless IR, Godwin TA, Allen F, Feder A. Nodular regenerative hyperplasia of the liver in hematologic disorders: a possible response to obliterative portal venopathy. A morphometric study of nine cases with hypothesis on the pathogenesis. Medicine 1980;59:367–379. [5] Wanless IR, Peterson P, Das A, Boitnott JK, Moore W, Bernier V. Hepatic vascular disease and portal hypertension in polycythemia vera and agnogenic myeloid metaplasia: a clinicopathological study of 145 patients examined at autopsy. Hepatology 1990;12:1166–1174. [6] Kingham JGC, Levison DA, Stansfeld AG, Dawson AM. Non-cirrhotic intrahepatic portal hypertension: a long term follow-up study. Q J Med 1981;199:259–268. [7] Dhiman RK, Chawla Y, Vasishta RK, Kakkar N, Dilawari JB, Trehan MS, et al. Non-cirrhotic portal fibrosis: experience with 151 patients and a review of the literature. J Gastroenterol Hepatol 2002;17:1–5. [8] Krasinskas AM, Eghtesad B, Kamath PS, Demetris AJ, Abraham SC. Liver transplantation for severe intrahepatic noncirrhotic portal hypertension. Liver transplant 2005;6:627–634.
[9] Isabel Fiel M, Thung SN, Hytiroglou P, Emre S, Schiano TD. Liver failure and need for liver transplantation in patients with advanced hepatoportal sclerosis. Am J Surg Pathol 2007;31:607–614. [10] Ludwig J, Hashimoto E, Obata H, Baldus WP. Idiopathic portal hypertension: a histopathological study of 26 Japanese cases. Histopathology 1993;22:227–234. [11] Bioulac-Sage P, Le Bail B, Bernard PH, Balabaud C. Hepatoportal sclerosis. Semin Liver Dis 1995;15:329–339. [12] Hillaire S, Bonte E, Denninger MH, Casadevall N, Cadranel JF, Lebrec D, et al. Idiopathic non-cirrhotic intrahepatic portal hypertension in the West: a reevaluation in 28 patients. Gut 2002;51:275–280. [13] Krasinskas AM, Goldsmith JD, Burke A, Furth EE. Abnormal intrahepatic portal vasculature in native and allograft liver biopsies. A comparative analysis. Am J Surg Pathol 2005;29:1382–1388. [14] Ibarrola C, Colina F. Clinicopathological features of nine cases of noncirrhotic portal hypertension: current definitions and criteria are inadequate. Histopathology 2003;42:251–264. [15] Ohbu M, Okudaira M, Watanabe K, et al. Histological study of intrahepatic aberrant vessels in cases of noncirrhotic portal hypertension. Hepatology 1994;20:302–308. [16] Denninger MH, Chaït Y, Casadevall N, Hillaire S, Guillin MC, Bezeaud A, et al. Cause of portal or hepatic venous thrombosis in adults: the role of multiple concurrent factors. Hepatology 2000;31:587–591. [17] Okuda K, Kono K, Ohnishi K, Kimura K, Omata M, Koen H, et al. Clinical study of eighty-six cases of idiopathic portal hypertension and comparison with cirrhosis with splenomegaly. Gastroenterology 1984;86:600–610. [18] Rosandaal FR. Venous thrombosis: a multicausal disease. Lancet 1999;353:1167–1173. [19] Watt FE, James OFW, Jones DEJ. Patterns of autoimmunity in primary biliary cirrhosis patients and their families: a population-based cohort study. Q J Med 2004;97:397–406. [20] Naber AH, Van Haelst U, Yap SH. Nodular regenerative hyperplasia of the liver: an important cause of portal hypertension in non-cirrhotic patients. J Hepatol 1991;12:94–99. [21] Malamut G, Ziol M, Suarez F, Beaugrand M, Viallard JF, Lascaux AS, et al. Nodular regenerative hyperplasia: the main liver disease in patients with primary hypogammaglobulinemia and hepatic abnormalities. J Hepatol 2008;48:74–82. [22] Saito K, Nakanuma Y, Takegoshi K, Ohta G, Obata Y, Okuda K, et al. Nonspecific immunological abnormalities and association of autoimmune diseases in idiopathic portal hypertension. A study by questionnaire. Hepato-Gastoenterol 1993;40:163–166. [23] Valla D, Pessegueiro-Miranda H, Degott C, Lebrec D, Rueff B, Benhamou JP. Hepatic sarcoidosis with portal hypertension. A report of seven cases with a review of the literature. Q J Med 1987;63:531–544. [24] Abraham SC, Kamath PS, Eghtesad B, Demetris AJ, Krasinskas AM. Liver transplantation in precirrhotic biliary tract disease: portal hypertension is frequently associated with nodular regenerative hyperplasia and obliterative portal venopathy. Am J Surg Pathol 2006;30:1454–1461. [25] Schiano TD, Kotler DP, Ferran E, Fiel IM. Hepatoportal sclerosis as a cause of noncirrhotic portal hypertension in patients in patients with HIV. Am J Gastroenterol 2007;102:2536–2540. [26] Vernier-Massouille G, Cosne J, Lemann M, Marteau P, Reinisch W, Laharie D, et al. Nodular regenerative hyperplasia in patients with inflammatory bowel disease treated with azathioprine. Gut 2007;56:1404–1409. [27] Roulot D, Degott C, Chazouillère O, Oberti F, Calès P, Carbonell N, et al. Vascular involvement of the liver in Turner’s syndrome. Hepatology 2004;39:239–247. [28] Sarin SK, Mehra NK, Agarwal A, Malhotra V, Anand BS, Phil D, et al. Familial aggregation in noncirrhotic portal fibrosis: a report of four families. Am J Gastroenterol 1997;82:1130–1133. [29] Dumortier J, Bizollon T, Scoazec JY, Chevallier M, Bancel B, Berger F, et al. Orthotopic transplantation for idiopathic portal hypertension: indications and outcome. J Scand Gastroenterol 2001;36:417–422. [30] Moucari R, Rautou PE, Cazals-Hatem D, Geara A, Bureau C, Consigny Y, et al. Hepatocellular carcinoma in Budd-Chiari syndrome: characteristics and risk factors. Gut 2008;57:828–835.
Journal of Hepatology 2011 vol. 54 j 455–461
461
Obliterative portal venopathy: Portal hypertension is not always present at diagnosis Dominique Cazals-Hatem1,⇑, , Sophie Hillaire2, , Marika Rudler2, Aurélie Plessier2, Valerie Paradis1, Bertrand Condat2, Claire Francoz2, Marie-Hélène Denninger3, François Durand2, Pierre Bedossa1, Dominique C. Valla2 1
Service d’Anatomie-Pathologique, Hôpital Beaujon AP-HP, Clichy and Institut National de la Santé et de la Recherche Médicale U773, Centre de Recherche Biomédicale Bichat Beaujon CRB3, Université Denis Diderot-Paris 7, France; 2Centre de Référence Maladies Rares, Maladies Vasculaires du Foie, Service d’Hépatologie, Hôpital Beaujon AP-HP, Clichy and Institut National de la Santé et de la Recherche Médicale U773, Centre de Recherche Biomédicale Bichat Beaujon CRB3, Université Denis Diderot-Paris 7, France; 3Service d’Hématologie biologique, Hôpital Beaujon AP-HP, Clichy and Institut National de la Santé et de la Recherche Médicale U773, Centre de Recherche Biomédicale Bichat Beaujon CRB3, Université Denis Diderot-Paris 7, France
Background & Aims: Previous studies on obliterative portal venopathy (OPV) have been biased due to the selection of patients with non-cirrhotic portal hypertension. The aim of this study was to clarify the characteristics of OVP diagnosed by liver biopsy. Methods: Fifty-nine consecutive patients with OPV were retrospectively selected on strict histological criteria. Clinical, laboratory, portal vein patency, and associated disorders potentially involving vascular alterations were analyzed. The occurrence of complications was recorded during follow-up. Results: Mean age at diagnosis was 38.5 ± 15 years old. Initial presentation was portal hypertension (64% of patients) and/or extrahepatic portal vein thrombosis (EHPVT) (22%) or isolated abnormal laboratory tests (20%). Associated diseases found at diagnosis were: prothrombotic disorders (30% of patients) and immune-mediated disorders (17%); 53% of patients had no causal factor (idiopathic OPV). During follow-up (median 8.6 years, range 1–23 years), features of portal hypertension worsened in 46% of patients; EHPVT and portal hypertension were finally found in 44% and 88% of patients. Anti-coagulation and betablockers were administered in 47% and 59% of patients, respectively. Severe complications (liver transplantation and/or death) occurred in 11 (19%) patients, 8 had idiopathic OPV. Patients with prothrombotic disorders received earlier anticoagulation therapy; all survived without transplantation.
Keywords: Vascular liver disease; Idiopathic portal hypertension; Hepatoportal sclerosis; Portal vein thrombosis; Liver biopsy. Received 20 March 2010; received in revised form 2 July 2010; accepted 5 July 2010; available online 28 October 2010 ⇑ Corresponding author. Tel.: +33 1 40 87 58 61; fax: +33 1 40 87 00 77. E-mail address: [email protected] (D. Cazals-Hatem). These authors contributed equally to this work. Abbreviations: OPV, obliterative portal venopathy; NCPH, non-cirrhotic portal hypertension; CPT, complete portal tract; NRH, nodular regenerative hyperplasia; HIV, human immunodeficiency virus; EHPVT, extrahepatic portal venous thrombosis; OLT, orthotopic liver transplantation; ALT, alanine aminotransferase; GGT, gamma glutamyl transferase; ULN, upper limit of normal value; JAK, Janus kinase.
Conclusions: A confident diagnosis of OPV can be done by biopsy and is conceivable in patients under 40 years without clinically significant portal hypertension. Poor outcome was noted in 19% of patients, most of them affected with idiopathic OPV. Patients with prothrombotic disorders received early anticoagulation and appeared to have a better outcome despite a high proportion of EHPVT. Ó 2010 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.
Introduction Obliterative portal venopathy (OPV) was coined to describe primary occlusion of the intrahepatic portal veins in the absence of cirrhosis, inflammation, and hepatic neoplasia [1,2]. This disease has been reported in patients with non-cirrhotic portal hypertension (NCPH) and because of the lack of extrahepatic portal vein obstruction, intrahepatic portal veins were studied. This entity was first described as an intrahepatic portal phlebosclerosis and the term hepatoportal sclerosis was used [3]. Portal vasculopathy predominated in medium-sized and preterminal veins (0.2–3 mm in diameter) which are easily accessible on liver biopsies [4,5]. However, in the literature, OPV has mainly been described in autopsies and livers explanted for NCPH or cryptogenic cirrhosis without portal vein thrombosis [6–9]. A positive diagnosis of OPV on biopsy requires specific histological findings composed of a spectrum of lesions that have been well described [10,11]. We have previously reported this disease diagnosed by biopsy in Caucasian patients, but all patients were selected based on the presence of portal hypertension preventing a relevant analysis of disease progression [12]. OPV probably exists before the signs of portal hypertension appear and it seems important to recognize it beforehand and to follow patients afterwards; many aspects of OPV diagnosed earlier on needle biopsies remain unclear and need to be re-evaluated.
Journal of Hepatology 2011 vol. 54 j 455–461
Research Article The aim of this study was to assess the clinical and histological features at presentation, the associated disorders, and the outcome of patients in whom a diagnosis of OPV was obtained via biopsy.
Patients and methods Patient selection From January 1987 to December 2007, all liver biopsies with a diagnosis of ‘‘primary OPV’’ were retrospectively selected from the database of our pathology department. Tissues were fixed in 10% formalin solution and paraffin sections were stained with H&E, trichrome, and argentation stains. The number of complete portal tracts (CPT) was counted in each specimen and portal veins were assessed. Ninety-eight specimens were available from 76 patients (18 patients underwent several biopsies). Specimens, reviewed by the same pathologist (DCH), had to fulfill 3 criteria: (i) to be longer than 1 cm and to contain more than 5 CPT; (ii) to have an alternation of CPT and centrilobular veins to exclude cirrhosis; (iii) to have more than 2/3 (66%) of the CPT harbouring abnormal portal venules, defined as absent or clearly reduced in caliber with sclerosis or thickening of the smooth muscle wall [4,5]. According to Krasinskas et al. [13], abnormal portal vessels corresponding to those described in NCPH can be present in greater than 25% of CPT in ‘‘normal livers with normal portal pressure’’. Therefore our latter criterion allowed the diagnosis of OPV to be made with confidence. Seventeen patients (22% of the entire cohort) were excluded as their needle biopsies contained less than 6 CPT (mean CPT: 3 [2–5]) and/or measured less than 1 cm in length (mean 7 mm [4–14]). Therefore, 59 patients were eligible corresponding to 82 biopsy specimens. Histological evaluation Additional morphological changes classically reported with OPV were recorded to assess their prevalence on biopsy specimens [4,5,8,10,11]: (a) nodular regenerative hyperplasia (NRH) defined as irregular thickened hepatic plates with a nodular appearance alternating with atrophic compressed plates without fibrous septa (b) sinusoidal dilatation; (d) perisinusoidal fibrosis detected on argentation stain; (e) aberrant vessels located close to or distant from portal tracts welldescribed by Ohbu et al. [15]; and (f) extensive portal fibrosis (>grade F1 adapted from METAVIR classification). Explanted livers available during follow-up were processed routinely: gross features and microscopic features were recorded the same way as for biopsies, in addition to specific data: weight of liver explants, dysmorphic appearance of livers (defined by atrophy and/or hypertrophy of a lobe and/or nodular contours), endophlebitis of large portal veins (>3 mm in diameter) and of large hepatic veins (>2 mm, hepatic venopathy), and macroscopic nodules (larger than 5 mm).
Results Initial presentation At the diagnosis of OPV, the 59 selected patients were 38.5 ± 15 years (range 7–77 years), 20 women (33%), and 39 men (66%). All patients had patent hepatic veins at Doppler-ultrasound and no evidence of neoplasia. The distribution of patients according to the type of initial presentation is depicted in Fig. 1. A clinically significant portal hypertension was present in 38 patients (64%) and absent in 21 patients (36%). EHPVT was present in 13 patients (22%) at diagnosis, including seven patients with acute thrombosis, and six with cavernous transformation; all six patients with portal cavernoma and one patient with acute portal vein thrombosis also had features of portal hypertension. Portal hypertension and EHPVT were absent in 15 patients (25%) at the initial diagnosis. The main indicators for biopsy were the appearance of portal hypertension (38 patients, detailed Fig. 1* according to the predominant symptoms initially presented), recent detection of EHPVT generally associated with
N = 59 Patients with OPV diagnosed by biopsy
N = 38 Portal Hypertension *
Collection of other data Clinical, laboratory, and imaging features at the time of initial diagnosis and during follow-up and treatments were recorded. Conditions considered to cause OPV were then identified from the patient’s chart based on conventional diagnostic criteria: professional exposure to copper sulphate or vinyl chloride monomer; previous exposure to thorium sulphate, Spanish toxic oil or arsenic salts, regular vitamin A supplements, autoimmune or collagen vascular diseases, primary biliary cirrhosis or primary sclerosing cholangitis, schistosomiasis, sarcoidosis, abdominal or systemic sepsis and prothrombotic disorders. Tests for antithrombin, protein S, protein C deficiencies, for the detection of antiphospholipid antibodies, hyperhomocysteinemia had been routinely performed according to previously reported methods [16]. Activated protein C resistance, Factor V Leiden (G1691A factor V gene mutation), G202100A Factor II gene mutation, and V617 JAK2 mutation were also assessed shortly after they were described. Patients seen before these molecular tests became available were tested later on at follow-up visits or using frozen stored blood samples. HIV serology status was determined. The presence of a congenital/genetic syndrome was explored. Additional causes of chronic liver diseases were identified according to conventional diagnostic criteria (chronic viral hepatitis, alcoholic liver disease, non-alcoholic steatohepatitis, haemochromatosis, autoimmunue hepatitis, Wilson disease, and biliary diseases). All patients who underwent a liver biopsy have been systematically explored by Doppler ultrasound. An imaging study completed with contrast-enhanced computerized tomography (CT) scan was routinely performed at the time of liver disease diagnosis: they evaluated portal vein patency size and flow, gross collat-
456
eral portosystemic circulation, ascites, and spleen enlargement defined as a spleen size >12 cm. They were routinely repeated during follow-up based on our previously reported data [12]. Upper endoscopy was also performed at the time of diagnosis and during follow-up to check for the deterioration of portal hypertension. Clinically significant portal hypertension was defined as: (i) oesophageal varices assessed on upper endoscopy (with or without bleeding) or (ii) gross collateral portosystemic circulation or (iii) splenomegaly with portal vein dilatation at imaging, or (iii) ascites or pleural effusion, or (iv) portosystemic encephalopathy. Deterioration of portal hypertension was defined as the occurrence of varices if previously absent or if increased in size by one grade as assessed by upper endoscopy (with or without bleeding), presence of ascites or spleen enlargement if absent at the previous examination.
N = 21 No Portal Hypertension *
N=7 N=6 Extra-hepatic Portal Vein Thrombosis (EHPVT) 7 Portal Cavernoma 6 Acute Portal Vein Thrombosis
Portal hypertension w/o EHPVT N = 31 (64%)
EHPVT N = 13 (22%)
No Portal hypertension and no EHPVT N = 15 (25%)
Fig. 1. Flowchart of the clinical presentation of the 59 patients at the time of initial diagnosis of OPV. *Main clinical manifestations of portal hypertension initially presented were: Varices bleeding n = 12, non-bleeding varices n = 7, splenomegaly n = 10, ascitis or pleural effusion n = 8, encephalopathy n = 1.
Journal of Hepatology 2011 vol. 54 j 455–461
JOURNAL OF HEPATOLOGY abnormal liver tests and/or discovered with prothrombotic disorders (six patients) or unexplained and isolated chronic liver test abnormalities (12 patients, 2 had cholestasis during pregnancy). The need for biopsy was unclear for three patients. Patients with portal hypertension were younger than those without it at diagnosis (mean age 30 ± 16 years versus 44 ± 14 years, p = 0.01). Most patients (n = 46, 78%) had laboratory abnormalities although marked abnormalities were uncommon (Table 1): only 14% of patients had serum alanine aminotransferase (ALT) >2 times the upper limit of normal value (ULN) or alkaline phosphatases >2 ULN or a prothrombin index <50% of normal. Most patients (56%) had a platelet count <150,000/mm3; while 10% of patients had thrombocytosis due to a myeloproliferative disease. Total bilirubin was normal (<17 lm/L) in 70% of patients (median 16 lm/L, [range 16–190]). Normal laboratory tests were observed in 13 of the 59 patients at diagnosis (22%). Associated disorders After a complete screening, associated disorders potentially involving vascular alterations were found in 28 patients (47%) divided into two groups according to the existence of prothrombotic (n = 18) or immune-mediated disorders (n = 10). Clinical and biological findings at the initial diagnosis and during follow-up were detailed for the entire cohort (n = 59) and according to associated disorders (Table 2). Ten patients (17%) had concomitant immune-mediated disorders (Table 3): four patients had autoimmune hepatitis identified by biopsies before the diagnosis of OPV (no extensive fibrosis at that time) and treated successfully with corticoids (n = 2) and azathioprine (n = 2). One patient had primary biliary cirrhosis without extensive fibrosis that was treated successfully with ursodeoxycholic acid and associated with an antiphospholipid syndrome. Three patients had rheumatoid arthritis (one with autoimmune hepatitis). A common variable immunodeficiency (treated with iterative immunoglobulin injections), a Wegener Table 1. Laboratory data at the time of histological diagnosis.
Biology
ALT
Alkaline phosphatase (N = 55)
GGT
Platelets
Prothrombin index
Values
Patients N = 59
%
28 23 7 1 24 23 8 18 21 20 22 11 20 6 8 33 18
47 39 12 2 44 42 14 31 36 34 37 19 34 10 14 56 30
<1 ULN 1-2 ULN 2-10 ULN >10 ULNN <1 ULN 1-2 ULN >2 N <1 ULN 1-4 ULN >4 N <90,000/mm3 90-150,000/mm 3 150-350,000/mm 3 >350,000/mm3 <50% 50-80% >80%
ALT, alanine aminotransferase; GGT, gamma glutamyl transferase; ULN, times upper the limit of normal value.
disease, and a Sharp syndrome were identified in one patient respectively. Anticardiolipid antibodies were found in six patients and two presented with an antiphospholipid syndrome. Investigations for prothrombotic disorders were obtained in 55 of the 59 patients: 18 patients (33% of patients tested) had prothrombotic disorders (Table 4); four patients had at least two prothrombotic disorders. The V617K JAK2 mutation was detected in 6 of the 30 patients tested. An isolated protein S deficiency was strongly suspected in three patients but could not be confirmed because a family study was impossible. One woman had both a factor II gene mutation and Turner’s syndrome. One patient had both an autoimmune disease and a protein S deficiency. No prothrombotic disorder or immune-mediated disorder could be identified in 53% of the patients (n = 31). None was HIV+ and no systemic or intra-abdominal infection was found. Repeated interviews excluded exposure to toxins potentially involving vascular alterations. Two patients had stigma of hepatitis C viral infection and presented with portal vein thrombosis without cryoglobulinemia: OPV was diagnosed on histology when the METAVIR evaluations were A1F1 in both. Treatment and outcome As shown in Table 2, during a median follow-up of 8.6 years [range 1–23], 40% of the 15 patients without initial portal hypertension or EHPVT, developed portal hypertension (varices in five, ascites in one), associated with an acute EHPVT in two patients; two of them were transplanted. Thirteen of the 46 patients (28%) without initial EHPVT developed an EHPVT detected at imaging: four patients had a prothrombotic disorder, three had an immune-mediated disorder. The deterioration or occurrence of portal hypertension was observed in 27 (46%) patients with variceal bleeding in seven (varices in 16, ascites in 11). At the end of follow-up, 77% of patients with thrombophilia developed EHPVT, versus 40% of patients with immune-mediated disorders and versus 26% in patients with no evidence of associated disorders. Lastly, clinically significant portal hypertension and EHPVT could be detected in 88% and 44% of all patients, respectively. In the entire cohort only seven patients had no evidence of portal hypertension and portal vein thrombosis at the end of followup, and all had a normal spleen at imaging. Thirty-five patients (59%) were treated with beta-blockers and this was combined with endoscopic therapy for five patients to check for gastrointestinal bleeding. Beta-blockers were discontinued in two of these patients because of its side-effects. Three patients also underwent surgical portocaval shunting for recurrent bleeding. Twenty-eight patients (47%) received anticoagulant therapy: indications for treatment were EHPVT in 24 patients, discovered at the initial diagnosis (n = 12) or during follow-up (n = 12), and a prothrombotic disorder without obvious thrombosis in the remaining four patients. Severe complications were observed in two patients (recurrent gastrointestinal bleeding and lethal hemorrhagic stroke). The interval between the diagnosis of OPV and anticoagulation varied according to the underlying disease, being shorter in patients with prothrombotic disorders than in others (Table 2). Nine patients (15%) underwent orthotopic liver transplantation (OLT) a median of 8.5 years [1–16] after OPV diagnosis at a median age of 49 years (27–59). All transplanted patients had severe portal hypertension which was associated with severe
Journal of Hepatology 2011 vol. 54 j 455–461
457
Research Article Table 2. Clinical findings at initial diagnosis and during follow-up of the entire cohort and of patients distributed according to associated disorders potentially involved in vascular alterations.
Associated disorders N(%)
All patients N = 59
Prothrombotic disorders 18 (30)
Immune- mediated disorders 10 (17)
Other patients 31 (53)
Initial Presentation Mean Age (Years old) PHT w/o Portal Thrombosis Extrahepatic Portal VeinThrombosis Isolated Abnormal Hepatic Tests Abdominal Pain
38 ± 15 31 (53) 13 (22) 12*(20) 3 (5)
40 ± 17 5 (28) 10 (55) 3 (17) 0
41 ± 9 4 (40) 1 (10 ) 4 (40) 1 (10)
36 ± 9 22 (71) 2 (6) 5 (16) 2 (6)
Outcome Follow-up (Years) Extrahepatic Portal Vein Thrombosis Deterioration/occurrence of PHT Liver Transplantation Death
8.6 ± 7.8 13/46 (28) 27 (46) 9 (15) 5 (8)
5 ± 5.8 4/8 (50) 6 (33) 0 0
13 ± 7 3/9 (33) 6 (60 ) 1 (10 ) 2 (20)
9±8 6/29 (21) 15 (48) 8 (26) 3 (10)
Treatment Anticoagulant (AC) Delay diagnosis/ AC therapy (Years)
28 (47) 2.3
13 (72) 0.7
5 (50) 4.6
10 (33) 3.6
PHT, portal hypertension; *two patients had cholestasis of pregnancy.
Table 3. Immune-mediated disorders associated with OPV.
Immune-mediated disorders Autoimmune hepatitis Primary biliary cirrhosis Rheumatoid arthritis Common variable immunodeficiency Wegener disease Sharp syndrome Total
Number of patients (%) 4* 1 3 1 1 1 10 / 59 (17%)
*One patient had also rheumatoid arthritis.
Table 4. Results of investigations for prothrombotic disorders performed in 55 out of the 59 patients.
Prothrombotic disorders Myeloproliferative disease (MPD)† Factor II deficiency Protein S deficiency Protein C deficiency MTHFR homozygote mutation Total
Number of patients (%) 10 3 3 3 2 18*/55 tested (33%)
V617F JAK2 mutation in six patients. *Four patients had more than one prothrombotic disorder: 2 had MPD and MTHFR homozygote mutation, 1 had MPD and protein S deficiency, 1 had a factor II mutation and a protein S deficiency.
liver dysfunction in seven of those patients and hepatopulmonary syndrome in two of the patients. The liver failure occurred in the context of sepsis/or oesophageal bleeding in five of the seven patients. Only one patient had an EHPVT treated with anticoagulation. Eight of the nine transplanted patients had idiopathic OPV: no etiological factor or associated disorder potentially involving vascular alterations was identified (one of them underwent both renal and hepatic transplantation for familial idiopathic renal 458
insufficiency); the last patient transplanted had an immunerelated disorder. During follow-up, five patients died (median age of 51 years old (29–62)): three of them were recently transplanted and died in the year post-OLT due to hepatopulmonary syndrome, sepsis, and multiorgan failure. One patient died with sepsis and hepatocellular deficiency and the last one with a hemorrhagic stroke after anticoagulation. Finally 11 of the 59 patients (19%) had an unfavorable outcome (OLT in nine, death in five, and both OLT and death in three) at a median age of 51 years [27–62]. Among them nine patients had no anticoagulation and two were receiving anticoagulation. Pathological data Among the 82 liver specimens obtained from the 59 patients selected, 73 were needle biopsies and nine surgical biopsies. Eighteen patients (30.5%) needed several biopsies to confirm the diagnosis of OPV. Transjugular biopsies were performed in 17 patients (29%) and a hemodynamic study was therefore available in these 17 patients. The median hepatic venous pressure gradient was 7 mm Hg [2–17] and 35% of patients had a normal pressure gradient (<5 mm Hg). The median length of needle biopsy specimens was 13.7 mm [5–30]. An average of 11 CPTs (6–40) were present per biopsy. Overall, 703 (79%) of the 885 CPTs analyzed as a whole contained abnormal portal venules. Most CPTs were small, sclerotic, and round without inflammation (Fig. 2). The argentation stain was essential to assess inconspicuous NRH changes and architectural disturbances. All biopsies had at least one additional lesion and 76% had at least two lesions as shown in Table 5 and Fig. 2. At the end of follow-up, 10 patients (17%) developed macronodules at imaging (1–3 each), most <3 cm in diameter and having the appearance of focal nodular hyperplasia (biopsy in seven patients). None developed hepatocellular carcinoma. The explanted livers of the nine patients transplanted during follow-up were fully examined: most were atrophic (median
Journal of Hepatology 2011 vol. 54 j 455–461
JOURNAL OF HEPATOLOGY Table 5. Histological features observed on biopsies and explants.
Histological lesions
Biopsy specimen N = 82 (%) Mean size°= 13.7mm [5-30]
Obliteration of portal venules Nodular regenerative hyperplasia Perisinusoidal fibrosis Sinusoidal dilatation Aberrant vessels Extensive portal fibrosis ¤ (>F1) Dysmorphic liver Obliterated large portal veins Hepatic venopathy
79* (96) 57 (70) 47 (57) 34 (41) 31 (38) 25 (30)
Explanted livers N = 9 (%) Mean weight= 837g [610-1,640] 9 (100) 9 (100) 9 (100) 6 (67) 5 (55) 8 (89) 6 (67) 6 (67) 2 (22)
Needle biopsies only (surgical biopsies not included); *3 needle biopsies displayed a nodular regenerative hyperplasia aspect with scarce portal tracts: repeated biopsies done in these patients asserted obliterative portal venopathy (OPV) diagnosis; ¤evaluation adapted from METAVIR classification.
Fig. 2. Histological features of obliterative portal venopathy diagnosed on needle biopsies. (A–C) Same specimen, trichrome stains 25, 100, 200. (A) Representative hepatic sample displaying >66% of complete portal tract with abnormal venules, (B) sclerotic portal venules with occlusion of the portal vein radicle, (C) a aberrant vessel present around the portal tract. (D and E) Same specimen, trichrome stains 25, 100. (D) Sample showing alternation of small fibrous portal tracts prolonged by thin septa and dilated veins, (E) most of the dilated veins were aberrant vessels. (F,G, and H) Same specimen, trichrome 25, 100, argentation stains 100). (F) Sample showing extensive portal fibrosis, (G) mixed with nodular regenerative hyperplasia in non-fibrous zones, (H) with sinusoidal fibrosis, thin septa, and aberrant vessels.
weight: 837 g (610–1,640)) and six were dysmorphic (Table 5 and Fig. 3). Only one liver exhibited two macronodules (<1 cm large) without neoplasia. Microscopically, all livers displayed diffuse OPV without cirrhosis. Extensive portal fibrosis was present in 8 of the 9 explants, more pronounced in perihilar (grade F2 in 6 and grade F3 in 2). Hepatic venopathy was found in 22% of livers. Large portal vein endophlebitis related to preexistent parietal thrombosis were present in 67% of livers. A ductopenia with bilirubin accumulation was observed in two patients: one of them was in an immune-related context and the other had idiopathic
Fig. 3. Gross and microscopic findings of livers explanted for obliterative portal venopathy. (A and B) same explant. (A) Gross view showing no dysmorphy; (B) trichrome stain 100 showing hepatoportal sclerosis with extensive fibrosis (F2) and sclerotic portal veins. (C,D) same explant. (C) Macroscopic view of a liver cut section showing perihilar nodular transformation without cirrhotic aspect, (D) trichrome stain 100 showing small sclerotic portal tracts and periportal nodular regenerative changes.
OPV combined with renal insufficiency. No active cholangitis or hepatitis was identified.
Discussion The present study handled a large cohort of patients affected with a rare and under-appreciated vascular disease: it provided histological, clinical, and prognostic features of OPV diagnosed by liver biopsy. Our inclusion criteria were based on the histological identification of primary occlusion of the portal microvasculature after
Journal of Hepatology 2011 vol. 54 j 455–461
459
Research Article ruling out cirrhosis, active inflammation, and/or metabolic liver diseases, regardless of initial manifestations. Although it was frequently associated, NRH per se was not accepted as a criterion of portal vasculopathy since it’s a secondary change and other vascular lesions have been proposed in its pathogenesis [5]. However, NRH changes constituted a useful guidance for the diagnosis of OPV since it was present in 70% of specimens. The 17 patients excluded because of insufficient CPT displayed NRH and were managed as the 59 patients described in the study: their features at presentation and during follow-up were similar to those of the 59 included patients (see Table 6 in supplementary data). In this series 30.5% of patients underwent several biopsies to meet our requirements as far as the number or CPT is concerned for a fair evaluation of portal veins. Examinations of explanted livers confirmed OPV in all patients, supporting the diagnostic criteria applied to biopsy. The complete histological work-up corroborated the heterogeneity of lesions in a single liver and between individuals as previously described [8,9,11,14]; overlaps between NRH and hepatoportal sclerosis reinforce the assumption that both belong to a same spectrum with OPV as a pivotal component. The first aim of our study was to describe the clinical presentation of OPV diagnosed by biopsy. Manifestations were heterogeneous and clinically significant portal hypertension was lacking in 36% of patients: only 17% and 10% of patients had splenomegaly or ascites, respectively, which contrasts with previous cohorts on NCPH clinically defined by portal hypertension and massive splenomegaly [6,7]. Patients with initial portal hypertension were younger than those without, which eliminated the effect of age at diagnosis. The fact that 25% of our patients initially presented without evidence for portal hypertension or portal thrombosis was also unusual compared to previous data on NCPH [7,12,17]. Many patients underwent a liver biopsy for slight-to-moderate liver test alterations and the diagnosis of OPV occurred by chance. Documented portal thrombosis was the initial sign in 55% of the patients with underlying thrombophilia, which should have prompted the systematic screening for causes. Secondly, we identified disorders potentially involved in vasculopathy in 47% of our patients: concomitant thrombotic or immune-mediated disorders were found in 30% and 17% of patients, respectively. These figures are higher than that expected in the general population (<15% and <12%, respectively) [18,19]. These findings were primarily descriptive but did not provide a mechanistic approach, which constitutes a limitation. The role of thrombophilia in the mechanism of OPV is well established [5,12]. On the other hand, the role of immune-mediated disorder in OPV is not well-documented in the process of portal vasculopathy but frequently proposed in the process of NRH [4,20,21]. A Japanese study on 160 patients with NCPH reported up to 11.9% of associated non-hepatic autoimmune disorders: thyroiditis, systemic lupus, and Sjogren syndrome were the most frequent [22]. OPV and more commonly NRH have been described in association with chronic inflammatory and immune-disorders disorders such as sarcoidosis, primary biliary cirrhosis, HIV infection, and common variable immunodeficiency [21,23–25]. The mechanisms of OPV have not yet been clarified: an inflammatory reaction close to the portal endothelium could cause endothelitis, thrombosis, and venular sclerosis in inflammatory disease such as primary biliary cirrhosis or sarcoidosis; otherwise no sign of vascularitis has been described. Thrombosis could be favored by
460
hyperviscosity induced by the marked hypergammaglobulinemia observed in autoimmune diseases. Two patients in our cohort were treated with azathioprine for autoimmune hepatitis or Sharp syndrome; but the role of this agent in OPV development remains unclear. Azathioprine treatment may generate NRH in inflammatory bowel disease but its role in OPV has not yet been confirmed [26]. Finally, 53% of our patients presented with ‘‘idiopathic OPV’’: histological examination of liver explants did not provide etiological indications. Clinically, they were younger (mean age at presentation = 36) and had severe portal hypertension requiring OTL in a quarter of them before the age of 50. In a recently reported series, the mean age of patients transplanted for progressive OPV varied from 47 years (n = 19) to 60 years (n = 8) [8,9]. The young age of our patients transplanted for idiopathic OPV might be due to complex pathological processes other than hypercoagulability, including genetic or congenital malformations. Indeed, OPV has been described in Turner’s syndrome suggesting the role of abnormal development or vascular malformations [27]; one patient in this series had Turner’s syndrome but she also had a prothrombotic disorder. Limited family studies have suggested the existence of familial hepatoportal sclerosis, incomplete septal cirrhosis, or NRH [28] but no familial clustering was found in the present cohort. Our last aim was to describe the natural course of OPV diagnosed by biopsy. After a mean follow-up of 8.6 years, the clinical course of the disease was severe in 19% of patients leading to transplantation or death. Deterioration or occurrence of portal hypertension was observed in nearly half of the patients, and at the end of follow-up evidence of portal hypertension and EHPVT were finally observed in 88% and 44% of patients, respectively. However, this study had limitations since it was retrospective, thus treatment had affected outcome; patients were heterogeneous and managed differently according to their etiology. Globally, the clinical course of patients was favorable thanks to appropriate management with betablockers. Patients with underlying prothrombotic disorders experienced portal thrombosis (77%) but had no severe complications since none needed transplantation or died. This favorable course could be influenced by the rapid administration of anticoagulants. By contrast, 25% of patients with idiopathic OPV evolved unfavorably without clear explanation; they received no anticoagulants but a causal link between both looked hazardous since patients were too heterogeneous to be compared each other. Our data do not allow to establish whether different disorders or the use of anticoagulation therapy can explain the differences in outcome between the ‘‘idiopathic’’ and the ‘‘prothrombotic OVP’’. Most transplant patients did not show EHPVT at imaging and therefore did not receive anticoagulation before OLT. Nevertheless, histological findings found extensive endophlebitis of the large portal veins in two third of the explants with impaired perfusion leading to liver atrophy and liver failure, probably due to inadequate arterial compensation. Therefore detection of thrombosis at imaging was underestimated and maintaining portal vein patency should be a major goal. Unfortunately, the overall survival post-OLT was poor in our cohort while patients had good outcome after OLT in other series [8,29]. To improve prognosis a trial of anticoagulation therapy could be warranted in all patients whatever the causal factor, after considering the benefit/risk ratio of such a treatment.
Journal of Hepatology 2011 vol. 54 j 455–461
JOURNAL OF HEPATOLOGY Macronodules detected in 17% patients were all benign as previously reported [7,14]. Regular survey is recommended since neoplastic transformation has been described in other chronic vascular diseases [30]. In conclusion, a confident OPV diagnosis can be obtained via biopsy; copious biopsies should be required for adequate portal vessel assessment. OPV had heterogeneous presentations and portal hypertension may be absent at diagnosis in 36% of patients. Early recognition of OPV by pathologists should prompt clinicians to search for immune-mediating or prothrombotic conditions; the latter appeared to have a better outcome when receiving early anticoagulation despite a high proportion of EHPVT. Nevertheless poor prognosis was noted in 19% of patients affected with idiopathic OPV. Latent congenital malformations or genetic susceptibilities may be involved. Published familial forms of OPV should be considered for preventive familial screening.
Conflict of interest The authors who have taken part in this study declared that they do not have anything to disclose regarding funding or conflict of interest with respect to this paper.
Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jhep.2010.07.038.
References [1] Nayak NC, Ramalingaswami V. Obliterative portal venopathy of the liver. Associated with so-called idiopathic portal hypertension or tropical splenomegaly. Arch Pathol 1969;87:359–369. [2] Kameda H. Obliterative portal venopathy: a comparison study of 184 cases of extrahepatic portal obstruction and 469 cases of idiopathic portal hypertension. J Gastroenterol Hepatol 1986;1:139–149. [3] Mikkelsen WP, Edmondson HA, Peters RL, Redeker AG, Reynolds TB. Extraand intrahepatic portal hypertension without cirrhosis (hepatoportal sclerosis). Ann Surg 1965;162:602–608. [4] Wanless IR, Godwin TA, Allen F, Feder A. Nodular regenerative hyperplasia of the liver in hematologic disorders: a possible response to obliterative portal venopathy. A morphometric study of nine cases with hypothesis on the pathogenesis. Medicine 1980;59:367–379. [5] Wanless IR, Peterson P, Das A, Boitnott JK, Moore W, Bernier V. Hepatic vascular disease and portal hypertension in polycythemia vera and agnogenic myeloid metaplasia: a clinicopathological study of 145 patients examined at autopsy. Hepatology 1990;12:1166–1174. [6] Kingham JGC, Levison DA, Stansfeld AG, Dawson AM. Non-cirrhotic intrahepatic portal hypertension: a long term follow-up study. Q J Med 1981;199:259–268. [7] Dhiman RK, Chawla Y, Vasishta RK, Kakkar N, Dilawari JB, Trehan MS, et al. Non-cirrhotic portal fibrosis: experience with 151 patients and a review of the literature. J Gastroenterol Hepatol 2002;17:1–5. [8] Krasinskas AM, Eghtesad B, Kamath PS, Demetris AJ, Abraham SC. Liver transplantation for severe intrahepatic noncirrhotic portal hypertension. Liver transplant 2005;6:627–634.
[9] Isabel Fiel M, Thung SN, Hytiroglou P, Emre S, Schiano TD. Liver failure and need for liver transplantation in patients with advanced hepatoportal sclerosis. Am J Surg Pathol 2007;31:607–614. [10] Ludwig J, Hashimoto E, Obata H, Baldus WP. Idiopathic portal hypertension: a histopathological study of 26 Japanese cases. Histopathology 1993;22:227–234. [11] Bioulac-Sage P, Le Bail B, Bernard PH, Balabaud C. Hepatoportal sclerosis. Semin Liver Dis 1995;15:329–339. [12] Hillaire S, Bonte E, Denninger MH, Casadevall N, Cadranel JF, Lebrec D, et al. Idiopathic non-cirrhotic intrahepatic portal hypertension in the West: a reevaluation in 28 patients. Gut 2002;51:275–280. [13] Krasinskas AM, Goldsmith JD, Burke A, Furth EE. Abnormal intrahepatic portal vasculature in native and allograft liver biopsies. A comparative analysis. Am J Surg Pathol 2005;29:1382–1388. [14] Ibarrola C, Colina F. Clinicopathological features of nine cases of noncirrhotic portal hypertension: current definitions and criteria are inadequate. Histopathology 2003;42:251–264. [15] Ohbu M, Okudaira M, Watanabe K, et al. Histological study of intrahepatic aberrant vessels in cases of noncirrhotic portal hypertension. Hepatology 1994;20:302–308. [16] Denninger MH, Chaït Y, Casadevall N, Hillaire S, Guillin MC, Bezeaud A, et al. Cause of portal or hepatic venous thrombosis in adults: the role of multiple concurrent factors. Hepatology 2000;31:587–591. [17] Okuda K, Kono K, Ohnishi K, Kimura K, Omata M, Koen H, et al. Clinical study of eighty-six cases of idiopathic portal hypertension and comparison with cirrhosis with splenomegaly. Gastroenterology 1984;86:600–610. [18] Rosandaal FR. Venous thrombosis: a multicausal disease. Lancet 1999;353:1167–1173. [19] Watt FE, James OFW, Jones DEJ. Patterns of autoimmunity in primary biliary cirrhosis patients and their families: a population-based cohort study. Q J Med 2004;97:397–406. [20] Naber AH, Van Haelst U, Yap SH. Nodular regenerative hyperplasia of the liver: an important cause of portal hypertension in non-cirrhotic patients. J Hepatol 1991;12:94–99. [21] Malamut G, Ziol M, Suarez F, Beaugrand M, Viallard JF, Lascaux AS, et al. Nodular regenerative hyperplasia: the main liver disease in patients with primary hypogammaglobulinemia and hepatic abnormalities. J Hepatol 2008;48:74–82. [22] Saito K, Nakanuma Y, Takegoshi K, Ohta G, Obata Y, Okuda K, et al. Nonspecific immunological abnormalities and association of autoimmune diseases in idiopathic portal hypertension. A study by questionnaire. Hepato-Gastoenterol 1993;40:163–166. [23] Valla D, Pessegueiro-Miranda H, Degott C, Lebrec D, Rueff B, Benhamou JP. Hepatic sarcoidosis with portal hypertension. A report of seven cases with a review of the literature. Q J Med 1987;63:531–544. [24] Abraham SC, Kamath PS, Eghtesad B, Demetris AJ, Krasinskas AM. Liver transplantation in precirrhotic biliary tract disease: portal hypertension is frequently associated with nodular regenerative hyperplasia and obliterative portal venopathy. Am J Surg Pathol 2006;30:1454–1461. [25] Schiano TD, Kotler DP, Ferran E, Fiel IM. Hepatoportal sclerosis as a cause of noncirrhotic portal hypertension in patients in patients with HIV. Am J Gastroenterol 2007;102:2536–2540. [26] Vernier-Massouille G, Cosne J, Lemann M, Marteau P, Reinisch W, Laharie D, et al. Nodular regenerative hyperplasia in patients with inflammatory bowel disease treated with azathioprine. Gut 2007;56:1404–1409. [27] Roulot D, Degott C, Chazouillère O, Oberti F, Calès P, Carbonell N, et al. Vascular involvement of the liver in Turner’s syndrome. Hepatology 2004;39:239–247. [28] Sarin SK, Mehra NK, Agarwal A, Malhotra V, Anand BS, Phil D, et al. Familial aggregation in noncirrhotic portal fibrosis: a report of four families. Am J Gastroenterol 1997;82:1130–1133. [29] Dumortier J, Bizollon T, Scoazec JY, Chevallier M, Bancel B, Berger F, et al. Orthotopic transplantation for idiopathic portal hypertension: indications and outcome. J Scand Gastroenterol 2001;36:417–422. [30] Moucari R, Rautou PE, Cazals-Hatem D, Geara A, Bureau C, Consigny Y, et al. Hepatocellular carcinoma in Budd-Chiari syndrome: characteristics and risk factors. Gut 2008;57:828–835.
Journal of Hepatology 2011 vol. 54 j 455–461
461