Hepatology Research 33 (2005) 68–71

Pathology of nonalcoholic steatohepatitis

Elizabeth M. Brunt∗

Saint Louis University Liver Center, Saint Louis University School of Medicine, 3635 Vista Avenue, St. Louis, MO 63110, USA

Available online 7 October 2005

Summary

To date, histologic evaluation, most commonly in the form of liver biopsy, remains the gold standard in evaluation of nonalcoholic fatty liverdisease (NAFLD). Histologic evaluation was fundamental to the initial studies that introduced and defined the concept of fatty liver as a liverdisease. Currently, liver biopsy in NAFLD serves multiple roles: confirmation (or exclusion) of the diagnosis; distinction of steatohepatitisfrom “simple steatosis”; assessment of extent of necroinflammatory activity, fibrosis, and architectural alterations. Histopathologic studieshave underscored the fact that not all obese and/or diabetic individuals with elevated liver tests have fatty liver disease; for example, hepaticglycogenosis and hepatosclerosis have been described in diabetics, and other significant liver diseases have been documented. Likewisebiopsy studies have documented lesions of steatosis or steatohepatitis in unusual patient groups or clinical settings, such as lean individuals,individuals with normal liver tests, patients taking certain medications, patients with co-existent serologically-diagnosed liver disease, andp NASH onl SH. Liverb steatosis”i s in NASHw

Treatmentt elations ofp e currenr

nosis ands me studies;B tory lesions( e NIDDKN steatosis tos tohepatitis,t ither extantt ome) willb©

K

1

ei

theep-f thea-Wesev-iliar

1d

ediatric patients. Biopsy studies have shown that the lesions of NASH may or may not persist in cirrhosis; prior evidence ofiver biopsy serves as a benchmark for the concept that many cases of otherwise cryptogenic cirrhosis developed from NAFLD/NAiopsy remains a significant feature of studies delineating long-term outcome of NAFLD, some of which have shown that “simple

s not always non-progressive and benign. Finally, investigators have noted correlations of proposed pathophysiologic processeith particular histologic features.Therapeutic trials for NASH rely on histologic evaluation as the most sensitive analysis to document effects of treatment.

rials afford an opportunity to evaluate histologic features of resolution, and these trials have also provided an opportunity for corrarticular histologic lesions with clinical and laboratory features in well-characterized patient populations. These kinds of studies artlyelatively few, but results of a recent study have reinforced the concept of necessary criteria for diagnosis.

Current discussions in pathology include identification of lesions of concern for progression, reproducible methods of diagemiquantification of lesions, and appropriate nomenclature. Matteoni et al. proposed NAFLD types 1–4 based on long-term outcorunt et al. proposed a system of grading and staging for NASH that follows methods of separate assessment for necroinflamma

grade) and fibrosis (stage) accepted in other forms of non-biliary chronic liver disease. Recently, the Pathology Committee of thASH Clinical Research Network has proposed a system of evaluation that encompasses the entire spectrum of NAFLD fromteatohepatitis with fibrosis for use in upcoming treatment trials. And, just as the clinician cannot distinguish steatosis and steahe pathologist cannot discern if alcohol abuse may be an underlying cause of the lesions. Proposed nomenclature to align with eerminology in other forms of chronic liver disease, or to align with our knowledge of underlying cause(s) (such as metabolic syndre discussed.2005 Elsevier Ireland Ltd. All rights reserved.

eywords: Non-alcoholic steatohepatitis; Pathology; Steatosis; Ballooning; Mallory’s hyaline; Perisinusoidal fibrosis; Grading; Staging

. Introduction

For 30 years prior to this publication in 1980 by Ludwigt al. [1] there were several papers in the literature relat-

ng to the liver disease in obesity and in obesity in diabetes.

∗ Tel.: +1 314 577 8782; fax: +1 314 268 5120.E-mail address: bruntem@slu.edu.

However, it was the paper by Ludwig et al. that codifiedterm nonalcoholic steatohepatitis into the lexicon of hatology. The Ludwig paper was based on a search oMayo Clinic files for liver biopsies with alcohol-like fetures, but in patients who clearly were not alcoholic.can recognize not only the histologic findings, but alsoeral of the clinical associations with which we are famtoday.

386-6346/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.hepres.2005.09.006

E.M. Brunt / Hepatology Research 33 (2005) 68–71 69

Currently our working definition of nonalcoholic steato-hepatitis remains a clinicopathologic entity. Clinical fea-tures may be suggestive. However, there remain challengesin deciding which patients to include or exclude withcurrent definitions[2–4]. More troublesome, there are nodiagnostic laboratory or imaging tests to positively diag-nose this entity. Therefore, liver biopsy evaluation remainsthe gold standard in this disease[2–5]. Biopsy findingsare useful for diagnosis, for distinguishing steatosis fromsteatohepatitis, and to assess ongoing necroinflammatoryinjury, or activity, and fibrosis and architectural change, orstage.

2. Value of liver biopsy in NAFLD and NASH

Liver biopsy has been used extensively in the last 20 yearswith natural history studies[6,7], in the studies to predictfibrosis[8,9] to determine which patient would benefit frombiopsy [10], to associate with the clinical associations ofmetabolic syndromes[11–16], certain drugs[17], and to doc-ument the concurrence of nonalcoholic fatty liver disease andnonalcoholic steatohepatitis with other forms of chronic liverdisease such as hepatitis C[17–20]. Certainly liver biopsyremains useful in investigations into pathogenesis and willbecome even more useful as a measure of efficacy in ongo-i

tsf ther ofw or-m ain-i osism n/re-e rectt inedl iverd

3

on-s turnt st ma-t allo eld.I intot viewb rp gicc

olics ntlyi er-

ent metabolically, enzymatically, and has different oxygentension. Likewise studies have shown, as Dr. Seiki will bepresenting, studies for lipid peroxidation products and DNAdamage predominate also in zone 3, as does the increasedexpression of CYP-2E1[23,24].

Liver cell injury in the form of ballooning is also a sig-nificant feature in nonalcoholic steatohepatitis and, as canbe seen from the well known study of Matteoni et al.[25]in which they provided clinical correlations with types 1–4,assessed by histologic features, and showed increased inci-dence of cirrhosis in types 3 and 4, compared with type 1.The distinguishing lesion between type 3 and types 1 and 2was the presence of ballooning.

Lobular inflammation in nonalcoholic steatohepatitis istypically quite mild. However, it is also typically mixed andincludes a small number of lymphocytes, a small numberof macrophages, Kupffer cells, but also a small number ofpolymorphonuclear leukocytes. Portal inflammation is eithernon-existent or very mild.

Mallory’s hyaline is a lesion that may or may not be seen.Many pathologists, including the author, do not require it forthe diagnosis[26]. When present, MH is typically referredto as “poorly formed”, as it lacks the dense ropy qualityof Mallory’s hyaline seen in alcoholic hepatitis or chroniccholestasis.

Apoptosis is another form of liver cell injury that hasg o-h as.G hep-a n int pl on-d to atitis[ ronp oidall omY entsw

e 3,t dingt ne 3a ten-s mayb omec thehT d ina elle thea e ofc thei ortedb ir-r hasd

ng treatment trials[2].Skelly et al. [21] studied biopsies from 354 patien

or otherwise unexplained liver test abnormalities;esult was 66% of the patients had fatty liver, halfhom also had fibrosis. But interestingly 6% were nal, almost 10% remained cryptogenic and the rem

ng 13% actually were able to have a clinical diagnade based on the liver biopsy and then re-evaluatio

xamination of the patient. Therefore, it is really not coro assume without biopsy proof that all otherwise unexplaiver test abnormalities are steatohepatitis or fatty lisease.

. Histologic features of NAFLD and NASH

How do we differentiate simple steatosis from the ctellations of lesions known as steatohepatitis? If we reo the original paper of Ludwig et al.[1], we note the lesionhat he described—steatosis, ballooning, lobular inflamion, Mallory’s hyaline, perisinusoidal fibrosis; currently,f these are very familiar to those of us working in this fi

nterestingly, also in the last 20 years with investigationhe pathogenesis of NASH, as shown in this thorough rey Professor Day and Saksena[22], many of the known outative mechanisms for injury in NASH have histoloounterparts.

One of the first, most important findings in nonalcohteatohepatitis is the fact that this lesion is predominanitially in zone 3. Certainly zone 3 of the acinus is diff

ained attention recently[27]. Other lesions seen in nonalcolic steatohepatitis include fatty cysts and lipogranulomlycogenated nuclei are common in nonalcoholic steatotitis. Megamitochondria have gained significant attentio

his disease[28–30]. A recent study from Caldwell’s grouooked at the distribution and morphology of megamitochria by ultra-structural analysis[31]. Iron is actually noften discussed or looked for in nonalcoholic steatohep

32], but typically there may be very mild hepatocellular iresent or iron present in the reticuloendothelial or sinus

ining cells [32]. A recent study published from Japan framauchi et al. showed a very high proportion of patiith NASH with hepatocellular iron[33].Fibrosis begins in the perisinusoidal regions of zon

ypically as a very delicate strand of collagen spreahroughout the sinusoids; fibrosis can be present in zond the portal tract not be involved. With time more exive damage and fibrosis occurs and bridging fibrosise seen. Ultimately, of course, cirrhosis may result. Sases of cirrhosis related to NASH may actually retainistologic lesions of active nonalcoholic steatohepatitis[34].here is growing recognition of the concept mentioneseries by Powell et al.[6] and substantiated by Caldw

t al. [35] that with progression the histologic lesions,ctive lesions may actually burn out and leave a typirrhosis that gives absolutely no histologic clue as tonitiating events. Based on series of biopsies such as repy Abelmalek et al.[36], the concept that cryptogenic chosis may actually derive in a large percent from NASHeveloped.

70 E.M. Brunt / Hepatology Research 33 (2005) 68–71

4. Grading and staging the lesions of NASH

Based on a series of over 50 cases examined histologi-cally for severity our group proposed a grading and stagingsystem[37]. We chose to emphasize the lesions that areunique to steatohepatitis as opposed to other forms of chronicliver disease. And those lesions are steatosis, ballooning andinflammation—both in the lobules and portal inflammation.As noted, with progression of grade steatosis is not the deter-mining factor; rather the combination of liver cell injury andinflammation in the lobules and portal tracts are.

Likewise for stage we cannot simply use the staging sys-tems of chronic hepatitis as those are initiated in the portalareas. So the proposed staging system begins with focal orextensive zone 3 perisinusoidal fibrosis for stage 1. Stage 2includes the zone 3 perisinusoidal fibrosis but then in addi-tion adds the portal or periportal fibrosis. And then of course,the damage can progress to bridging fibrosis and ultimatelycirrhosis. Once bridging fibrosis and cirrhosis have occurredthe perisinusoidal fibrosis may or may not persist.

5. Is NASH really identical to ASH?

It is often stated in the clinical literature that part of thedefinition of nonalcoholic fatty liver disease is the lesions arei ll byt true.I thatw t thath ss oft otheri se.

6

ato-s Andt ef at wes dis-e d1 hep-a pati-t ostics heses uch aa olo-g ogists.

on-a ng tod iths to do

with alcohol. We really do not know yet what role alcoholmay or may not play in this entity and I would like to pro-pose that it’s time we move away from a non-association tosomething that we know. Pathologists can diagnosis steato-hepatitis, and then we rely on our clinicians to evaluate thepatient to determine the underlying etiology. Perhaps in thismanner, we can serve two purposes: to continue further studyof this very interesting liver disease and to study in a moreuniform fashion treatment options for the patients.

7. Conclusions

What do we see ahead in the future for the discussions inpathology? Believe it or not, we are still discussing diagnos-tic features[26]. I think it’s going to be very interesting toactually see more literature on NASH from Asia[23]. Therecertainly is NASH in Asia, and it will be interesting to studythe similarities and the differences. Another area of growth ispediatric nonalcoholic fatty liver disease and pediatric NASH[46,47]. And, finally, with more treatment trials ongoing nowwe are going to start to be able to learn from the lesions thatrepresent resolution of injury[48,49].

The NIDDK NASH Cooperative Research NetworkPathology Committee presented a group proposal for his-tologic assessment at AASLD in 2003 in Boston that willh

R

to-ease.

pati-logy

se.

ase.

risktroen-

JW,s: aogy

est

torsepa-

ght

[ ase:the

[ atic82Y

dentical to alcoholic liver disease. Obviously, you can tehe question that this author does not think this is alwaysn fact, there are several lesions in alcoholic liver diseasee do not see in nonalcoholic steatohepatitis, at leasave not been reported to date[38]. One of these includeclerosing hyaline necrosis with this obliterative lesionhe terminal hepatic venule and severe fibrosis, and thes the veno-occulsive lesion seen in alcoholic liver disea

. Terminology of NASH

On the other hand, in many biopsies with only steis, could a pathologist really tell the underlying cause?he answer of course is “no”[26,32,39,40]. Because of thesacts, this author and others have offered a proposal thtart thinking about changing the way we talk about thisase, at least in pathology[2,26,32,40]. In the 1960s an970s, chronic hepatitis was divided into chronic activetitis, chronic persistent hepatitis and chronic lobular he

is; these distinctions were thought to carry some prognignificance. But it was very confusing, in part because tame names were also used for specific liver diseases, sutoimmune hepatitis. A group of recognized liver pathists decided to diagnose what we actually see as pathol. . the entity of chronic hepatitis[41–45].

Furthermore, in nonalcoholic fatty liver disease and nlcoholic steatohepatitis we have spent a long time havieal with something that is a “non” and having to deal womething that we keep saying does not have anything

s

opefully be coming out in publication soon[50].

eferences

[1] Ludwig J, Viggiano TR, McGill DB, Ott BJ. Nonalcoholic steahepatitis: Mayo clinic experiences with a hitherto unnamed disMayo Clin Proc 1980;55:434–8.

[2] Neuschwander-Tetri BA, Caldwell SH. Nonalcoholic steatohetis: summary of an AASLD Single Topic Conference. Hepato2003;37:1202–19.

[3] Clark JM, Brancati FL, Diehl AM. Nonalcoholic fatty liver diseaGastroenterology 2002;122:1649–57.

[4] Sanyal AJ. AGA Technical review on nonalcoholic fatty liver diseGastroenterolgy 2002;123:1705–25.

[5] Saibara T. “Insufficient” leptin production for the fat mass: afactor for nonalcoholic steatohepatitis in obese patients? J Gasterol 2003;38:522–3.

[6] Powell EE, Cooksley WGE, Hanson R, Searle J, HallidayPowell LW. The natural history of nonalcoholic steatohepatitifollow-up study of forty-two patients for up to 21 years. Hepatol1990;11:74–80.

[7] Angulo, Lindor KD. Treatment of non-alcoholic steatohepatitis. BPract Res Clin Gastroenterol 2002;16:797–810.

[8] Angulo P, Keach JC, Batts KP, Linder KD. Independent predicof liver fibrosis in patients with nonalcoholic steatohepatitis. Htology 1999;30:1356–62.

[9] Ratziu V, Giral P, Charlotte F, et al. Liver fibrosis in overweipatients. Gastroenterology 2000;118:1117–23.

10] Dixon JB, Bhathal PS, O’Brien PE. Nonalcoholic fatty liver disepredictors of nonalcoholic steatohepatitis and liver fibrosis inseverely obese. Gastroenterology 2001;121:91–100.

11] Chitturi S, WEltman M, Farrell GC, et al. HFE mutations, hepiron, and fibrosis: ethnic-specific association of NASH with C2but not with fibrotic severity. Hepatology 2002;36:142–9.

E.M. Brunt / Hepatology Research 33 (2005) 68–71 71

[12] Bugianesi E, Leone N, Vanni E, et al. Expanding the natural his-tory of nonalcoholic steatohepatitis: from cryptogenic cirrhosis tohepatocellular carcinoma. Gastroenterology 2002;123:134–40.

[13] Bugianesi E, Manzini P, D’Antico S, et al. Relative contributionof iron burden, HFE mutations, and insulin resistance to fibrosis innonalcoholic fatty liver. Hepatology 2004;39:179–87.

[14] Marchesini G, Brizi M, Blanchi G, et al. Nonalcoholic fattyliver disease: a feature of the metabolic syndrome. Diabetes2001;50:1844–50.

[15] Marchesini G, Brizi M, Morselli-Labate AM, et al. Association ofnonalcoholic fatty liver disease with insulin resistance. Am J Med1999;107:450–5.

[16] Marchesini G, Bugianesi E, Forlani G, et al. Nonalcoholic fattyliver, steatohepatitis, and the metabolic syndrome. Hepatology2003;37:917–23.

[17] Farrell GC. Drugs and steatohepatitis. Semin Liver Dis2002;22:185–94.

[18] Brunt EM, Ramrakhiani S, Cordes BG, et al. Concurrence of his-tologic features of steathepatitis with other forms of chronic liverdisease. Mod Pathol 2003;16:49–56.

[19] Sanyal AJ, Contos MJ, Sterling RK, et al. Nonalcoholic fatty liverdisease in patients with hepatitis C is associated with features of themetabolic syndrome. Am J Gastroenterol 2003;98:2064–71.

[20] Ong JP, Younossi ZM, Speer C, Olano A, Gramlich T, BoparaiN. Chronic hepatitis C and superimposed nonalcoholic fatty liverdisease. Liver 2001;21:266–71.

[21] Skelly MM, James PD, Ryder SD. Findings on liver biopsy to inves-tigate abnormal liver function tests in the absence of diagnosticserology. J Hepatol 2001;35:195–9.

[22] Day CP, Saksena S. Nonalcoholic steatohepatitis: definitions andpathogenesis. J Gastroenterol Hepatol 2002;17:S377–84.

[ y isatol

[ a K.age

[ ul-and

[ Dis

[ andteato-

[ to-is

[ or-–4.

[ holicndrial

[31] Le TH, Caldwell SH, Redick JA, et al. The zonal distribution ofmegamitochondria with crystalline inclusions in nonalcoholic steato-hepatitis. Hepatology 2004;39:1423–9.

[32] Brunt EM. Nonalcoholic steatohepatitis: definition and pathology.Semin Liver Dis 2001;21:3–16.

[33] Yamauchi T, Kamon J, Waki H, et al. The fat-derived hormoneadiponectin reverses insulin resistance associated with both lipoatro-phy and obesity. Nat Med 2001;7:941–6.

[34] Shimada M, Hashimoto E, Taniai M, et al. Hepatocellular car-cinoma in patients with non-alcoholic steatohepatitis. J Hepatol2002;37:154–60.

[35] Caldwell SH, Oelsner DH, Iezzoni JC, Hespenheide EE, Battle EH,Driscoll CJ. Cryptogenic cirrhosis: clinical characterization and riskfactors for underlying disease. Hepatology 1999;29:664–9.

[36] Abdelmalek M, Ludwig J, Lindor KD. Two cases from the spectrumof nonalcoholic steatohepatitis. J Clin Gastroenterol 1995;20:127–30.

[37] Brunt EM, Janney CG, Di Bisceglie AM, Neuschwander-TetriBA, Bacon BR. Nonalcoholic steatohepatitis: a proposal for grad-ing and staging the histological lesions. Am J Gastroenterol1999;94:2467–74.

[38] Brunt EM. Alcoholic and nonalcoholic steatohepatitis. Clin LiverDis 2002;6:399–420.

[39] Burt AD, Mutton A, Day CP. Diagnosis and interpretation of steato-sis and steatohepatitis. Semin Liver Dis 1998;15:246–58.

[40] Brunt EM, Tiniakos DG. Pathology of steatohepatitis. Best Pract ResClin Gastroenterol 2002;16:691–707.

[41] Desmet VJ, Gerber M, Hoofnagle JH, Manns M, Scheuer PJ. Clas-sification of chronic hepatitis: diagnosis, grading and staging. Hep-atology 1994;19:1513–20.

[42] Ishak KG. Chronic hepatitis: morphology and nomenclature. ModPathol 1994;7:690–713.

[ for

[ ange.

[ ary.

[ R,logi-diatr

[ nal-

[ conment17.

[ oflogy

[ ionis-logy

23] Farrell GC. Nonalcoholic steatohepatitis: what is it, and whit important in the Asia-Pacific region? J Gastroenterol Hep2003;18:124–38.

24] Seki S, Kitada T, Yamada T, Sakaguchi H, Nakatani K, WakasIn situ detection of lipid peroxidation and oxidative DNA damin non-alcoholic fatty liver diseases. J Hepatol 2002;37:56–62.

25] Matteoni CA, Younossi ZM, Gramlich T, Borarai N, Yao CL, McClough AJ. Nonalcoholic fatty liver disease: a spectrum of clinicalpathological severity. Gastroenterology 1999;116:1413–9.

26] Brunt EM. Nonalcoholic steatohepatitis. Semin Liver2004;24:3–20.

27] Feldstein AE, Canbay A, Angulo P, et al. Hepatocyte apoptosisfas expression are prominent features of human nonalcoholic shepatitis. Gastroenterolgy 2003;125:437–43.

28] Caldwell SH, Chang CY, Nakamoto RK, Krugner-Higby L. Michondria in nonalcoholic fatty liver disease. Clin Liver D2004;8:595–617.

29] Caldwell SH, Swerdlow RH, Khan EM, et al. Mitochondrial abnmalities in non-alcoholic steatohepatitis. J Hepatol 1999;31:430

30] Sanyal AJ, Camapbell-Sargent C, Mirshahi F, et al. Nonalcosteatohepatitis: association of insulin resistance and mitochoabnormalities. Gastroenterology 2001;120:1183–92.

43] Scheuer PJ. Classification of chronic viral hepatitis: a needreassessment. J Hepatol 1991;13:372–4.

44] Scheuer PJ. The nomenclatrue of chronic hepatitis: time for a chJ Hepatol 1995;22:112–4 [Editorial].

45] Ludwig J. The nomenclature of chronic active hepatitis: an obituGastroenterolgy 1993;105:274–8.

46] Schwimmer JB, Deutsch R, Rauch JB, Behling C, NewburyLavine JE. Obesity, insulin resistance and other clinicopathocal correlates of pediatric nonalcoholic fatty liver disease. J Pe2003;143:500–5.

47] Schwimmer JB, et al. The histological features of pediatric nocoholic fatty liver disease (NAFLD). Hepatology 2002;36:412A.

48] Neuschwander-Tetri BA, Brunt EM, Wehmeier KR, Oliver D, BaBR. Improved nonalcholic steatohepatitis after 48 weeks of treatwith the PPAR-� ligand rosiglitazone. Hepatology 2003;38:1008–

49] Promrat K, Lutchman G, Uwaifo GI, et al. A pilot studypioglitazone treatment for nonalcoholic steatohepatitis. Hepato2004;39:188–96.

50] Kleiner DE, Brunt EM, Van Atta M, et al. Design and validatof a histologic scoring system for nonalcoholic fatty liver dease (NAFLD) and nonalcoholic steatohepatitis (NASH). Hepato2003;38:233A.