Hepatic alcoholic liver disease - From BMJ and ACP · of alcoholic hepatitis and termed this...
5
J Clin Pathol 1986;39:63-67 Hepatic vein lesions in alcoholic liver disease: retrospective biopsy and necropsy study AD BURT, RNM MacSWEEN University Department of Pathology Western Infirmary, Glasgow SUMMARY Obliteration of the terminal hepatic venules with perivenular fibrosis (phlebosclerosis) is a well recognised feature in alcoholic liver disease. Veno-occlusive lesions with intimal obliteration of hepatic veins and a lymphocytic phlebitis of hepatic veins may also be present. We looked for these lesions in 256 liver biopsies and 50 livers obtained at necropsy from patients with alcoholic liver disease. Phlebosclerosis was a universal finding in alcoholic hepatitis and cirrhosis and showed increasing severity with progressive liver injury. Veno-occlusive lesions, however, were found in only 25 of 256 (9-8%) of biopsies and 11 of 50 (22%) of livers obtained at necropsy, showing alcoholic hepatitis or cirrhosis: lymphocytic phlebitis was found in 10 of 256 (3 9%) and two of 50 (4%), respectively. Moreover, veno-occlusive lesions were generally mild. The prevalence of veno-occlusive lesions and lymphocytic phlebitis was considerably less than has been previously documented. Phlebosclerosis may have a different mechanism and be a more important contributory factor in progressive liver injury. The histological features of alcoholic liver disease have been reviewed in several recent publications.` 1- The presence of hepatic vein lesions, however, has been little emphasised. Edmondson et al emphasised the importance of the obliteration of hepatic vein branches by a process of perivenular fibrosis.6 In their review of alcoholic liver disease Baptista et al noted that the venous radicles eventually become obliterated in alcoholic hepatitis and may, consequently, be difficult to identify.4 Venous occlusion by a process of intimal proliferation was noted by Patrick and McGee,7 but Popper et al considered this lesion to be rare and seldom of substantial functional importance.3 Goodman and Ishak, however, in a retrospective necropsy study showed these veno-occlusive lesions in 25 of 48 (52-1%) cases of alcoholic hepatitis and in 103 of 139 (74%) cases with established alcoholic cirrhosis. In addition, they reported on the universal prevalence of perivenular phlebosclerotic lesions in both hepatitis and cirrhosis and described a lymphocytic phlebitis in eight (16 7%) cases with alcoholic hepatitis. In this retrospective study we reviewed a series of 256 liver biopsies and 50 livers obtained at necropsy from patients with alcoholic liver disease: we specifically looked for hepatic vein lesions as defined by Goodman and Ishak.8 Accepted for publication 17 September 1985 Material and methods The material comprised 256 liver biopsies submitted to the department of pathology at this hospital over five years (from 1977 to 1982). This was a consecutive series selected on the basis of: histological diagnosis; a documented clinical history of alcohol abuse; and clinical exclusion of any other known cause of liver injury. There was a male to female ratio of 1.56:1 with a mean age at the time of biopsy of 51 1 years (range 20-76) and 51-8 (range 31-76), respectively. Liver tissue obtained after death was examined from a similarly selected group of 50 cases, comprising 37 men and 13 women (mean age 56 3 and 58-3 years, respectively). Thirty eight (78%) of the patients had died from complications related to their liver disease; the proximate causes of death in eight of the remaining patients were: acute pancreatitis (3), pneumonia (3), fungal septicaemia (1), and Gram negative peritonitis (1). In the final four death was attributed to congestive cardiac failure (rheumatic heart disease (2), congestive cardiomyopathy (1), and cor pulmonale (1)): in all four cases there was established cirrhosis, but cardiac failure was not thought to have been a major contributor to the liver disease. Sections (5 pm) from all specimens were routinely stained with haematoxylin and eosin, Masson's trichrome, Gordon and Sweet's reticulin, Shikata's 63 on August 7, 2020 by guest. Protected by copyright. http://jcp.bmj.com/ J Clin Pathol: first published as 10.1136/jcp.39.1.63 on 1 January 1986. Downloaded from
Transcript of Hepatic alcoholic liver disease - From BMJ and ACP · of alcoholic hepatitis and termed this...
J Clin Pathol 1986;39:63-67
Hepatic vein lesions in alcoholic liver disease:retrospective biopsy and necropsy studyAD BURT, RNM MacSWEEN
University Department of Pathology Western Infirmary, Glasgow
SUMMARY Obliteration of the terminal hepatic venules with perivenular fibrosis (phlebosclerosis) isa well recognised feature in alcoholic liver disease. Veno-occlusive lesions with intimal obliterationof hepatic veins and a lymphocytic phlebitis of hepatic veins may also be present. We looked forthese lesions in 256 liver biopsies and 50 livers obtained at necropsy from patients with alcoholicliver disease. Phlebosclerosis was a universal finding in alcoholic hepatitis and cirrhosis and showedincreasing severity with progressive liver injury. Veno-occlusive lesions, however, were found inonly 25 of 256 (9-8%) of biopsies and 11 of 50 (22%) of livers obtained at necropsy, showingalcoholic hepatitis or cirrhosis: lymphocytic phlebitis was found in 10 of 256 (3 9%) and two of 50(4%), respectively. Moreover, veno-occlusive lesions were generally mild. The prevalence ofveno-occlusive lesions and lymphocytic phlebitis was considerably less than has been previouslydocumented. Phlebosclerosis may have a different mechanism and be a more importantcontributory factor in progressive liver injury.
The histological features of alcoholic liver diseasehave been reviewed in several recent publications.`1-The presence of hepatic vein lesions, however, hasbeen little emphasised. Edmondson et al emphasisedthe importance of the obliteration of hepatic veinbranches by a process of perivenular fibrosis.6 In theirreview of alcoholic liver disease Baptista et al notedthat the venous radicles eventually becomeobliterated in alcoholic hepatitis and may,consequently, be difficult to identify.4 Venousocclusion by a process of intimal proliferation wasnoted by Patrick and McGee,7 but Popper et alconsidered this lesion to be rare and seldom ofsubstantial functional importance.3 Goodman andIshak, however, in a retrospective necropsy studyshowed these veno-occlusive lesions in 25 of 48(52-1%) cases of alcoholic hepatitis and in 103 of 139(74%) cases with established alcoholic cirrhosis. Inaddition, they reported on the universal prevalence ofperivenular phlebosclerotic lesions in both hepatitisand cirrhosis and described a lymphocytic phlebitis ineight (16 7%) cases with alcoholic hepatitis.
In this retrospective study we reviewed a series of256 liver biopsies and 50 livers obtained at necropsyfrom patients with alcoholic liver disease: wespecifically looked for hepatic vein lesions as definedby Goodman and Ishak.8
Accepted for publication 17 September 1985
Material and methods
The material comprised 256 liver biopsies submittedto the department of pathology at this hospital overfive years (from 1977 to 1982). This was a consecutiveseries selected on the basis of: histological diagnosis; adocumented clinical history of alcohol abuse; andclinical exclusion of any other known cause of liverinjury. There was a male to female ratio of 1.56:1 witha mean age at the time of biopsy of 51 1 years (range20-76) and 51-8 (range 31-76), respectively.
Liver tissue obtained after death was examinedfrom a similarly selected group of 50 cases,comprising 37 men and 13 women (mean age 56 3 and58-3 years, respectively). Thirty eight (78%) of thepatients had died from complications related to theirliver disease; the proximate causes of death in eight ofthe remaining patients were: acute pancreatitis (3),pneumonia (3), fungal septicaemia (1), and Gramnegative peritonitis (1). In the final four death wasattributed to congestive cardiac failure (rheumaticheart disease (2), congestive cardiomyopathy (1), andcor pulmonale (1)): in all four cases there wasestablished cirrhosis, but cardiac failure was notthought to have been a major contributor to the liverdisease.
Sections (5 pm) from all specimens were routinelystained with haematoxylin and eosin, Masson'strichrome, Gordon and Sweet's reticulin, Shikata's
63
on August 7, 2020 by guest. P
rotected by copyright.http://jcp.bm
j.com/
J Clin P
athol: first published as 10.1136/jcp.39.1.63 on 1 January 1986. Dow
nloaded from
Burt, MacSween
Fig. la Moderate (grade 2+) veno-occlusive lesion with accompanying phlebitis andperivenularfibrosis in alcoholichepatitis. (Massons trichrome.) x 300. (b) Serial section stained with Shikata's orcein showing presence ofintimalproliferation within elastic lamina. x 300.
orcein, periodic acid Schiff (before and after diastasetreatment), and Perls's reaction. In selected casesadditional sections were stained with elastic vanGieson.
Vascular lesions were assessed according to thecriteria of Goodman and Ishak8:Veno-occlusive lesions These were characterised byintimal proliferation and fibrosis occurring within the
elastic lamina of the vein radicles, ranging from mild(1+), intimal thickening with one third occlusion,through to moderate (2 +), with two thirds occlusion,to severe (3 +), when there was more than two thirdsocclusion to complete obliteration (Figs. 1 and 2).Individual cases were scored on the basis of the mostsevere lesion present.Phlebosclerosis This consisted of compression of
42Xi*; i.S & X s'm 2 s ' >- i ' 8 -
M @ . *,0 o
*ND.s .re'w4 a* W,.st-F
Fig. 2 Severe (grade 3 +) veno-occlusive lesion in cirrhoticliver affecting interlobular vein and tributary. Necropsyseries. (Elastic van Gieson.) x 140.
Fig. 3 Mild (grade 1+) phlebosclerosis in alcoholichepatitis. (Elastic van Gieson.) x 180.
64
on August 7, 2020 by guest. P
rotected by copyright.http://jcp.bm
j.com/
J Clin P
athol: first published as 10.1136/jcp.39.1.63 on 1 January 1986. Dow
nloaded from
Hepatic vein lesions in alcoholic liver disease
Fig. 4a Obliterative lesion in hepatic vein branch in cirrhotic liver at necropsy (Hamatoxylin and eosin.) x 375. (b)Elastic tissue admixed with collagen in obliterative lesion. (Elastic van Gieson.) x 375.
hepatic vein radicles by perivenular fibrosis and wasgraded as: mild (1+), perivenular fibrosis withminimal compression of the vein lumen (Fig. 3);moderate (2+), more marked fibrosis andcompression of the vein lumen; severe (3 +), completeobliteration of the vein radicle with extensiveperivenular fibrosis (Fig. 4). Individual cases werealso scored on the basis of the most severe lesionpresent.Lymphocytic phlebitis was defined as the presence of
sFo
* r" a X
%.,a
~#t #/ :*
r ; rs v
tu '.*4 A %w .>. ; t s->s** S ?40
Fig. 5 Mild lymnphocytic phlebitis affiecting interlobular veinin established cirrhosis. (Haemaroxylin and eosin.) x 280.
an inflammatory infiltrate, predominantlylymphocytic, within the walls of hepatic vein branches(Figs la and 5). Spillover of neutrophil polymorphsfrom foci of hepatocyte necrosis in the perivenularzone was excluded.
Results
The biopsy cases were classified as follows: fatty liver(71), fatty liver with pericellular or perivenularfibrosis, or both (20), alcoholic hepatitis (82), andcirrhosis (83). Of the 83 cases of cirrhosis, 76 showedevidence of continuing active alcoholic hepatitis.Cirrhosis was present in 44 livers obtained atnecropsy and alcoholic hepatitis in six.The Table summarises the incidence of hepatic vein
lesions in the various subgroups. Veno-occlusivelesions were found in only one case of fatty liver,although they were present in 13 (15-8%) of the 82biopsy specimens with alcoholic hepatitis and in 12(14-4%) of those with cirrhosis; they were generallymild (1 +). Phlebosclerotic lesions were much morecommon; lesions were found in 20 (22%) of the 91cases with fatty liver and in all but one instance theywere of mild degree and concomitant with mildpericellular fibrosis. When we compared thesubgroups with alcoholic hepatitis with these withcirrhosis, the severity of the phlebosclerotic lesion wasgreater in the group with cirrhosis; 65 of 83 (78%)were severe (3+) compared with four of 82 (5%) inthe group with alcoholic hepatitis. A lymphocyticphlebitis occurred in eight (9-7%) of the 82 cases ofalcoholic hepatitis and in two (2 4%) of the cases of
65
on August 7, 2020 by guest. P
rotected by copyright.http://jcp.bm
j.com/
J Clin P
athol: first published as 10.1136/jcp.39.1.63 on 1 January 1986. Dow
nloaded from
66
Table Prevalence ofhepatic vein lesions in subgroups ofalcoholic liver disease
Burt, MacSween
Veno-occlusive lesions Phlebosclerosis Lymphocytic phlebitisMild moderate severe Mild moderate severe (%)
Total (%) Total (%)
Fattyliver(n = 71) 1 0 0 1(1.4) 0 0 0 - 0Fatty liver with fibrosis (n = 20) 0 0 0 - 19 1 0 20(100) 0Alcoholic hepatitis:
Biopsy(n = 82) 9 3 1 13(15-8) 41 35 4 80(975) 8(9-7)Necropsy(n = 6) 0 1 2 3(50) 0 3 3 6(100) 1(16-6)
Cirrhosis:Biopsy(n = 83) 9 3 0 12(14-4) 1 17 65 83(100) 2(2.4)Necropsy (n = 44) 2 4 2 8(18-1) 0 0 44 44(150) 1 (2.3)
cirrhosis.In the necropsy group veno-occlusive lesions were
seen in three of six (50%) of those with alcoholichepatitis but were present in only eight of 44 (18%) ofthe cirrhotic livers. Phlebosclerosis was universallypresent and severe lymphocytic phlebitis was rare.
Discussion
Edmondson noted that obliterative sclerosis ofhepatic veins and perivenular sinusoids was a featureof alcoholic hepatitis and termed this "sclerosinghyaline necrosis."' This phlebosclerosis hassubsequently been referred to in several otheraccounts of alcoholic liver disease,1 5 but itsprevalence, pathogenesis, and its possiblecontribution to the progression of the liver injuryhave remained uncertain. Furthermore, in someaccounts distinction has been drawn betweenphlebosclerosis and veno-occlusive lesions. Theveno-occlusive lesions resemble those described byBras et al in their classical studies of pyrollizidinerelated liver disease.9 Goodman and Ishak in theirretrospective necropsy study also distinguishedbetween phlebosclerosis and veno-occlusive lesions,finding phlebosclerosis in all their cases of alcoholichepatitis or alcoholic cirrhosis and veno-occlusivelesions in 52% and 74%, respectively.8 The authorsalso described a lymphocytic phlebitis, which wasfound much less often.
In this study we looked for hepatic vein lesions asdefined by Goodman and Ishak. The veno-occlusivelesions were readily identified and usually affected thelarger hepatic vein branches. The prevalence of theselesions was, however, considerably less than thatnoted by Goodman and Ishak.The histological diagnosis of phlebosclerosis
presented some difficulties. In sections stained withhaematoxylin and eosin it was often difficult toidentify hepatic vein radicles within areas of fibrousscarring that topographically corresponded to thevenular and perivenular zones. Residual elasticlamina, however, could usually be shown within these
areas on either the orcein stain or the elastic vanGieson stain (Fig. 4). In a small number of biopsiessome hepatic vein branches showed features ofintimal veno-occlusion as defined by Goodman andIshak8 but accompanied by perivenular adventitialfibrosis. Possibly, progression of both the intimalproliferation and the adventitial fibrosis could haveproduced a totally obliterative lesion which we thenregarded as being phlebosclerotic. Nakanuma et alshowed transitions between veno-occlusive andphlebosclerotic lesions in cirrhosis on serialsectioning. 1 2The pathogenesis of veno-occlusion and
phlebosclerosis in alcoholic liver disease is unclear.These conditions may represent direct toxic injury tovascular endothelium in a manner similar to thatproposed for lesions seen with certain toxins andchemotherapeutic agents.9-1l Veno-occlusive diseaseproduced by pyrollizidine alkaloids, however, may berelated to injury at the sinusoidal level withsubsequent damage to both endothelial cells andhepatocytes.12 13 There is a consequent decrease inhepatic vein flow and a narrowing in the luminaldiameter to a size appropriate for the reduced bloodflow. In cirrhotic livers studied after death byNakanuma et al focal veno-occlusive lesions werefound in 75-90%, and the incidence and morphologyof the lesions seemed to be similar irrespective of theaetiological type of the cirrhosis-that is, alcoholic,primary biliary, posthepatitic, or cryptogenic. 14 Theysuggested that these focal lesions represented asecondary response to a reduced blood flow, thereduction resulting from microvascular disturbanceby fibrous scarring or nodular regeneration, or both.Although such an explanation could apply to the
veno-occlusive lesions which we saw in this study, webelieve that the phlebosclerotic lesions represent adifferent form of injury. They were seen at a stage ofalcoholic fatty liver in which there was no evidentswelling of the liver cells and when reduction invenous flow seemed improbable. These early lesionswere accompanied by a mild degree of pericellular
on August 7, 2020 by guest. P
rotected by copyright.http://jcp.bm
j.com/
J Clin P
athol: first published as 10.1136/jcp.39.1.63 on 1 January 1986. Dow
nloaded from
Hepatic vein lesions in alcoholic liver disease
fibrosis, which is being increasingly recognised as anearly component of liver disease induced by alcoholand which, indeed, may progress to cirrhosis withoutan intervening stage of alcoholic hepatitis.15 - 17
It seems likely to us that the early phleboscleroticlesion is part of the fibrosis induced by alcohol, whichcommences in the perivenular zone and which may bethe very earliest evidence of incipient and progressiveliver injury. It is evident from our results that theseverity of the phlebosclerosis increases withprogressive liver injury and may of itself be importantin the further evolution of the disease.
References
1Edmnondson HA. Pathology of alcoholism. Am J Clin Pathol1963;74:725-42.
2MacSween RNM. Alcoholic liver disease. In: Anthony PP, WoolfN, eds. Recent advances in histopathology. Edinburgh: ChurchillLivingstone, 1978:193-212.
3Popper H, Thung SN, Gerber MA. Pathology of alcoholic liverdiseases. Semin Liver Dis 1981;1:203-16.
'Baptista A, De Groote J, Bianchi L, et al. Alcoholic liver disease:morphological manifestations. Lancet 1981;i:707-1 I.
sFleming KA, McGee JO'D. Alcohol induced liver disease. J ClinPathol 1984;37:721-33.
6Edmondson HA, Peters RL, Reynolds TB, Kuyura OT. Sclerosinghyaline necrosis in the liver of the chronic alcoholic. A recog-nisable lesion. Ann Intern Med 1963;59:646-73.
67
'Patrick RS, McGee JO'D. Alcoholic liver disease. Biopsy pathologyof the liver. Philadelphia: JB Lippincott Co, 1980:119-39.
8Goodman ZD, Ishak KG. Occlusive venous lesions in alcoholicliver disease: a study of 200 cases. Gastroenterology1982;83:786-96.
9 Bras G, Jellifee DB, Stuart KL. Veno-occlusive disease of the liverwith non-portal type of cirrhosis occurring in Jamaica. Archivesof Pathology 1954;57:285-300.
'°Griner PF, Elbadawi A, Packman CH. Veno-occlusive disease ofthe liver after chemotherapy. Ann Intern Med 1976;85:578-82.
'Gill RA, Onstad GR, Cardainone JM. Hepatic veno-occlusivedisease caused by 6-thioguanine. Ann Intern Med 1982;96:58-60.
Stirling GA, Bras G, Urquhart AE. The early lesions in veno-occlusive disease of the liver. Arch Dis Child 1962;37:535-8.
13 Brooks SEH, Miller CG, McKenzie K, Audretsh JJ, Bras G. Acuteveno-occlusive disease of the liver. Archives of Pathology1970;89:507-20.
14Nakanuma Y, Ohta G, Doishita K. Quantitation and serial sectionobservations of focal veno-occlusive lesions of hepatic veins inliver cirrhosis. Virchows Arch (Pathol Anat) 1985;405:429-38.
"Van Waes L, Lieber CS. Early perivenular sclerosis in alcoholicfatty liver: an index of progressive liver injury. Gastroenterology1977;73:646-50.
16Nasrallah SM, Nassar VH, Galambos JT. Importance of terminalhepatic venule thickening. Arch Pathol Lab Med 1980;104:84-6.
7Nakano M, Worner TM, Lieber CS. Perivenular fibrosis inalcoholic liver injury: ultrastructure and histologic progression.Gastroenterology 1982;83:777-85.
Requests for reprints to: Professor RNM MacSween,University Department of Pathology, Western Infirmary,Glasgow GI 1 6NT, Scotland.
on August 7, 2020 by guest. P
rotected by copyright.http://jcp.bm
j.com/
J Clin P
athol: first published as 10.1136/jcp.39.1.63 on 1 January 1986. Dow
nloaded from
