Nonalcoholic steatohepatitis: A study of 49 patients

Nonalcoholic Steatohepatitis: A Study of 49 Patients RANDALL G. LEE, MD Nonalcoholic steatohepatitis (NASH) refers to an alcoholic hepatitis-like histologic pattern found in nonalcoholic patients. A review of 543 liver biopsies diagnosed as alcoholic hepatitis yielded 49 cases of NASH. The patients were commonly middleaged women who were obese and often diabetic. NASH was usually discovered because of abnormal liver function tests or hepatomegaly noted during evaluation of other medical problems. Histologic examination revealed the same spectrum of changes found in alcoholic hepatitis, including cirrhosis in eight patients. Follow-up information was available for 39 patients after an average length of 3.8 years. Only one patient developed hepatic decompensation or died with liver failure or portal hypertension. Repeat histologic material was available for 13 patients after a mean 3.5 years of follow-up. Five patients showed progression of fibrosis, with cirrhosis developing in two, but the other eight patients demonstrated little morphologic change. These findings indicate that NASH is, in general, a clinically mild and biologically low-grade condition, but with the potential to progress and evolve into cirrhosis in some patients. The factors promoting progression are unclear. HUM PATHOL 20:59&598. 0 1989 by W.B. Saunders Company.

Alcoholic hepatitis presents a familiar and distinctive histologic pattern of liver injury resulting from chronic alcohol consumption. Its characteristic features include hepatocyte damage, often distinguished by Mallory bodies (alcoholic hyalin), an inflammatory infiltrate with neutrophils predominating, and fibrosis that may range from minimal perivascular foci to dense scarring and fibrous septa.1*2 Fatty change is a frequent accompaniment. Mild injury is located primarily in perivenular areas, but, when severe, the process may extend to involve portal tracts, creating necrotic and fibrotic bridges that can eventuate in cirrhosis3 However, a similar histologic picture is occasionally discovered in patients with no history of alcohol abuse and has been referred to as nonalcoholic steatohepatitis (NASH) or fatty liver hepatitis.4-6 First recognized as a complication of jejunoileal bypass procedures for morbid obesity,7,8 NASH has been associated with a variety of other clinical conditions (Table 1). Ludwig et al5 have distinguished primary NASH, which is associated with obesity, from second-

From the Department of Pathology, Oregon Health Sciences University, Portland. Accepted for publication January 3, 1989. Presented in part at the annual meeting of the United States and Canadian Academy of Pathology, Chicago, March 9, 1987. Kq words: nonalcoholic steatohepatitis, fatty liver hepatitis, alcoholic liver disease, steatohepatitis, diabetic hepatitis. Address correspondence and reprint requests to Randall G. Lee, MD, Department of Pathology UHS-19, Oregon Health Sciences University, 3181 SW Sam Jackson Park Rd, Portland, OR 97201. 0 1989 by W.B. Saunders Company. 0046-8177/89/2006-0009$5.00/O

594

ary NASH, which has a more specific cause such as jejunoileal bypass or drugs. The pathogenetic factors that underlie these diverse conditions and produce the hepatic damage of NASH are poorly understood. The natural history of NASH is largely undefined, as few follow-up data are available. The concurrence of cirrhosis in many reported cases suggests that NASH, like alcoholic hepatitis, progresses with increasing fibrosis and cirrhosis. However, the risk of developing cirrhosis is unclear and no prognostic indicators of progression have been identified. To examine these issues, patients with NASH were identified and their clinical features, pathologic characteristics, and follow-up status evaluated. MATERIALS AND METHODS A total of 543 patients had liver biopsy specimens diagnosed as alcoholic hepatitis (or one of its synonyms) in the surgical pathology files of the University of Utah Medical Center from 1968 through 1982. The clinical data and histologic sections of these patients were reviewed to identify cases of NASH. Of the 543 patients, 459 were excluded because of a definite or suspected history of alcohol use or uncertainty about its consumption. An additional seven patients with priorjejunoileal bypass procedures were also not included in the study. In 12 cases, the histologic material was insufficient for evaluation and 16 cases failed to meet the diagnostic criteria for alcoholic hepatitis as proposed by an international group of pathologists and described above.* The remaining 49 patients with primary NASH formed the basis of this study. Each patient was considered by attending physicians to be a nondrinker, a conclusion supported in the clinical record by at least three independent interviews with each patient, and, in most cases, by discussion with family members. In addition, 38 patients were practicing Mormons with a religious proscription of alcohol use. Although it is impossible to prove the nondrinking status of these 49 patients, they were conservatively selected and, by all available information, appeared not to consume alcohol. Pertinent clinical information, laboratory data, and follow-up details were obtained from hospital and clinic charts. Routine histologic sections from all liver specimens were stained with hematoxylin and eosin and trichrome or reticulin stains and reviewed without knowledge of clinical, biochemical, or outcome data. Sections were evaluated for the presence of hepatocyte damage and necrosis, neutrophilic inflammatory infiltration, mononuclear inflammatory infiltration, Mallory bodies, fibrosis, fat, and glycogenated nuclei. These features were semiquantitatively graded on an arbitrary 0 to 3+ scale with 0 denoting absence and 1+ , 2 + , and 3 + indicating mild, moderate, and severe degree. Biopsies showing cirrhosis or probable cirrhosis2’ were designated separately and their fibrosis not otherwise graded. Length of follow-up was determined from date of initial biopsy to date of death or last known

NONALCOHOLIC STEATOHEPATlllS (Lee)

TABLE 2.

TABLE 1. Conditions Associated With NASH

LaboratoryValues In Patients With NASH

Jejunoileal bypass’vs Obesit~5~g

Mean (Range)

Diabetes mellitusiO~i’ Extensive small bowel resection12~13 Weber-Christian

AST (IU) ALT(IU) Alkaline phosphatase (IU/L) Total bilirubin

diseasei

Gastroplasty for morbid obesity15 Therapeutic drugs Amiodarone’G~” Perhexilene

(mg/dL) Albumin (g/dL)

maleate’*

Glucocorticoidsig Synthetic estrogenGO

Abbreviations: AST, nine aminotransferase.

Percentage of Patients With Abnormal Values

89 104

(20 to 222) (35 to 198)

103

(44 to 211)

39

0.7 (0.2 to 1.4) 4.2 (3.3 to 5.0)

12 4

aspartate

aminotransferase;

ALT,

ala-

entity was represented, including cirrhosis in eight cases, and the details are summarized in Table 3. The severity of the different features varied from case to case. NASH commonly appeared as a process of mild to moderate activity, manifested by generally low grades of hepatocyte damage and inflammatory infiltration. The inflammatory component comprised neutrophils in all cases, but mononuclear cells were frequently admixed and predominated in 21 cases (Fig 2). Mallory bodies, recognized in 49% of the biopsies, were usually sparse. The extent of fibrosis varied considerably, ranging from delicate strands surrounding small groups of liver cells to sclerotic fi-

contact with the patient. Statistical analyses were performed using the paired and unpaired Student’s t test and the chisquare test as appropriate.

RESULTS The 49 patients averaged 53 years (range, 36 to 83) and included 11 men and 38 women. Their weights varied from 57 to 140 kg (mean 89) or, when standardized for height and medium frame in the Metropolitan Life Insurance Company tables, from 100% to 210% of ideal body weight with a mean of 144%. Thirty-four patients exceeded 130% of ideal body weight. A history of diabetes mellitus was noted in 25 patients and none had been treated with any drug associated with NASH. Although many patients had nonspecific complaints of weakness and fatigue, none presented primarily with symptoms of liver disease. NASH was generally discovered incidentally during the evaluation or follow-up of other medical conditions. These encompassed a wide variety of disorders: coronary artery disease/congestive heart failure (eight patients), cholelithiasis (six patients), peptic ulcer disease (five patients), hypertension (five patients), gynecologic conditions (three patients), psychiatric disorders (three patients), elective surgery (three patients), hyperlipoproteinemia associated with diabetes mellitus (two patients), peripheral vascular disease (two patients), breast carcinoma (two patients), endometrial carcinoma (one patient), Hodgkin’s disease (one patient), and thyroid nodule (one patient). Hepatomegaly was recorded in 30 patients, splenomegaly in three, and ascites in one. Liver function tests at the time of liver biopsy were abnormal in 42 patients (Table 2); mild to moderate elevations of transaminases were the most common alteration. Serologic markers were negative in all tested patients including hepatitis B surface antigen in 41 patients, antimitochondrial antibodies in 25, and antinuclear antibodies in 16. Biopsies were generally performed because of unexplained liver function test abnormalities or hepatomegaly. Four patients had incidental liver biopsies at cholecystectomy. The histologic appearances of the initial biopsies (46 needle and three surgical wedge specimens) were, by definition, indistinguishable from alcoholic hepatitis (Fig 1). The entire morphologic spectrum of that

FIQURE 1. NASH demonstrating features Identical to alcoholic hepatitis, including Mallory kdles (arrow). (Hemataxylin-easln stain; approximate magnification x 180.)

595

HUMAN PATHOLOGY

TABLE 3.

Volume 20, No. 6 [June 1989)

Histologic Features in NASH

Grade Hepatocyte damage and necrosis 1+ 2+ 3+ Neutrophil infiltration 1+ 2+ 3+ Mononuclear infiltration 1+ 2+ 3-k Mallory bodies 1+ 2+ 3+ Fibrosis 1+ 2+ 3+ Cirrhosis Fat 1+ 2+ 3+ Glycogenated nuclei 1+ 2+ 3+

tral valve replacement (one), and relapse of Hodgkin’s disease (one). Of the five cirrhotic patients with follow-up, three had died and two were alive after 1.3 and 1.8 years. Follow-up histologic material was present for 13 patients (eight repeat biopsy and five autopsy specimens). These specimens had been obtained after a mean follow-up length of 3.5 years (range, 1.2 to 6.9 years), as summarized in Table 4. Eight cases, including one with cirrhosis, showed little histologic difference between initial and follow-up specimens. There was no increase in the extent of fibrosis and none of the other histologic variables changed by more than one grade. However, five patients developed progression of fibrosis after follow-up ranging from 1.7 to 6.1 years. Cirrhosis was discovered in two of 12 noncirrhotic patients, one of whom initially had grade 2 fibrosis. No clinical or histologic features distinguished patients with and without progression.

No. of Cases (Total = 49) 49 30 17 2 49 23 23 3 41 27 14 0 24 20 4 0 49 19 13 9 8

DISCUSSION

49 11 18 20

NASH, defined as alcoholic hepatitis-type liver injury occurring in the absence of alcohol abuse, has been reported in several case reports and series. A common clinical setting emerges from these reports and is confirmed by this study of 49 patients. Al-

17 11 6 0

brous septa. Eight cases were associated with cirrhosis. Although not a diagnostic criterion in this study, fatty change was nonetheless present in all biopsies, often to a marked degree. Seventeen cases also showed glycogenated nuclei; 14 of these 17 had a history of diabetes mellitus. The grades of the histologic features were compared against both the degree and the frequency of liver function test abnormalities. No statistically significant correlation could be established and the degree of morphologic alterations could not be predicted by laboratory values. Follow-up information was available for 39 patients. The length of follow-up averaged 3.8 years (range, 1.0 to 16.8). During this period, only one patient developed clinical evidence of liver failure. This patient, a 57-year-old woman with diabetes mellitus who weighed 87 kg, presented with mild ascites and had cirrhosis proven on biopsy. Over the next 1.6 years, she developed increasing ascites, jaundice, and encephalopathy and died following esophageal variceal bleeding. No other patient had jaundice, ascites, encephalopathy or other stigmata of chronic liver disease noted at any follow-up examination. Statistical comparison of initial and last available liver function tests showed no significant change. A total of ten patients died. Cause of death was generally related to the patient’s primary medical problem, with only the one aforementioned patient dying of liver decompensation. Other causes included myocardial infarction (four patients), stroke (one), mesenteric arterial thrombosis (one), pneumonia following hip fracture (one), postoperative following mi-

FIQURE 2. Perivenular area typical of NASH with low-grade hepatocyte damage and predominant mononuclear infiltrate. (Hematoxylin-eosin stain; approximate magnificatton x 350.)

596

NONALCOHOLICSTEATOHEPATITIS(Lee) TABLE 4.

Histologic Follow-Up of NASH Follow-Up

Presenting Problem

Initial Biopsy MB

FC

GN

Fib

Length (vrs)

If 0 2+ 2+

1+ 0 0 1+

2+ If 1+ 3+

2+ 1c 0 0

If 1+ 1+ 2+

1.4 3.3 6.1 5.3

1+ 1+

0 1+

1+ 2+

2+ 3+

0 1+

2+ 2+

1.2 4.3

1+

2+

1+

0

3+

0

2+

6.9

Cholelithiasis Angina

2+ 2+ 1+ 2+

1+ 1+ 1+ 2+

2+ 2+ If 1+

2+ 1+ 0 1+

3+ 2+ 3+ 2+

2+ 1+ 0 0

2+ 2+ 3+ 3+

1.7 4.0 2.5 3.1

Venous thrombosis Thyroid adenoma

1+ 3+

1+ 2+

2+ 2+

0 2+

2+ 2+

1+ 0

3+ Cir

2.3 5.2

HD

N

M

Cholelithiasis Enlarged uterus Hypertension Cholelithiasis

1+ 1+ If 2+

If 2+ 1+ 3+

+/+I+

Schizophrenia Vaginal bleeding

1+ If

F

+I-

Duodenal ulcer

44 47 42 14

F M F M

+/+ +I+I-/+

Irrermlar menses

58 62

F F

iI+/+

Age

Sex

Obesitv/DM

63 54 58 48

F F F F

+/+ +I+ -/+I+

37 67

F F

66

Gas&culcer

status

HD

N

M

MB

FC

GN

Fib

Alive Alive Alive Dead (MI) Alive Dead (MD Dead ’ (stroke) Alive Alive Alive Dead (Ml) Alive Dead (pneumonia)

1+ 2+ 1+ 1+

1+ 2+ 2+ 2+

1+ 1+ 1+ 2f

1+ 1+ 0 If

2+ 2+ 2+ 2+

1+ 1+ 0 0

1+ 2+ 2+ 2+

1+ If

1+ 1+

1+ 1+

1+ 1+

2+ 2+

0 0

2+ 2+

1+

1+

0

0

2+

0

2+

2+ 1+ 1+ 2+

1+ 1+ 1+ 1+

2+ 2+ 2+ 1+

1+ 0 0 0

3+ 1+ 3+ 1+

1+ 0 0 0

3+ Cir 3+ 3+

1+ 2+

1+ 1+

2+ 2+

0 1+

1+ 2+

0 0

Cir Cir

Abbreivations: DM, diabetes mellitus; HD, hepatocyte damage; N, neutrophii infiltration; M, mononuclear change; GN, glycogenated nuclei: Fib, fibrosis; Cir, cirrhosis; Ml, myocardial infarction.

infiltration; MB, Mallory bodies; FC, fatty

The natural history of NASH has commended little prior attention, but progression to cirrhosis seems plausible for several reasons. Such an evolution is clearly established for NASH following jejunoileal bypass24*25 and for the histologically analogous alcoholic hepatitis. *K~ The concurrence in some reported cases of cirrhosis and NASH also suggests this outcome, although its frequency and time course are unclear. The follow-up data presented here help define the natural history. The patients of this study generally remained clinically well, even after more than 16 years of follow-up, and only one of the 39 whose status was known developed hepatic decompensation or died with liver failure or portal hypertension. However, neither clinical condition nor laboratory values correlate with the morphologic changes, as shown in this and other studies,28 so these features provide little information about possible progression to cirrhosis. Therefore, histologic follow-up is required, but few reported patients have had repeat liver biopsies. No progression of fibrosis was noted by Falchuk et allo in two patients rebiopsied after 1 year, whereas Thaler describes the development of micronodular cirrhosis in two cases.2g In this study no progression was seen in seven of the 12 noncirrhotic patients with repeat histologic examination. On the other hand, the other five patients showed increased fibrosis on follow-up and two of the five had developed cirrhosis. These data suggest that NASH is a static or slowly progressive process in many patients, but that a notable minority will develop increased fibrosis and ultimately cirrhosis. No features predictive of progression were identified. However, histologic follow-up was available for only 27% of the patients. Nevertheless, the continued clinical well-being of the others argues that severe liver disease did not generally supervene in this group. These conclusions are limited by the retrospective nature of the study and the possible bias introduced by ascertainment of cases on the basis of liver biopsy.

though it is possible that some reported patients successfully concealed their alcoholism, the similarity of features among various patients from differing institutions argues that NASH represents a clinicopathologic entity distinct from alcoholic hepatitis. NASH affects primarily middle-aged women who are obese and often diabetic.4*5*‘0~22 Over 75% of reported patients are female and the average age is 52 years with most patients between 45 and 65 years. Obesity is described in 50% to 90% of patients, depending on the definition of obesity used, and diabetes mellitus (type II, maturity onset) in 25% to 75%. However, because most studies are retrospective, the exact frequency of diabetes is uncertain, particularly as NASH in some instances has preceded the development of glucose intolerance.” Many patients are asymptomatic or mildly symptomatic and NASH is frequently an incidental finding occasioned by hepatomegaly or liver function test abnormalities. However, in some cases the clinical presentation is more dramatic with jaundice, ascites, encephalopathy, or esophageal varices, especially when cirrhosis is present.** NASH reveals the same histologic spectrum found in alcoholic hepatitis. In fact, the cases reviewed in this study were selected from those originally diagnosed as alcoholic hepatitis. NASH is usually only mildly active with low grade hepatocyte damage and inflammatory infiltration. Mononuclear inflammatory cells, rather than neutrophils, frequently predominate, an observation emphasized by Falchuk et al.rO Mallory bodies are present, but, as noted by Ludwig et a1,5 are often poorly formed and inconspicuous. Despite these qualitative differences, there is considerable morphologic overlap between NASH and alcoholic hepatitis and, in any given case, the distinction cannot be made on histologic grounds. 23 The practical importance is that this pattern of liver injury, although usually due to chronic alcoholism, entails a differential diagnosis (Table 1).

597

HUMAN PATHOLOGY

Volume 20, No. 6 (June 1989)

lesions of central pericellular fibrosis in morbid obesity, and after jejunoileal bypass. Am J Clin Path01 66:684-691, 1976 9. Nasrallah SM, Wills CE Jr, Galambos JT: Hepatic morphology in obesity. Dig Dis Sci 26:325-327, 1981 10. Falchuk KR, Fiske SC, Haggitt RC, et al: Pericentral hepatic fibrosis and intracellular hyalin in diabetes mellitus. Gastroenterology 78:535-541, 1980 11. Batman PA, Scheuer PJ: Diabetic hepatitis preceding the onset of glucose intolerance. Histopathology 9:237-243, 1985 12. Peura DA, Stromeyer FW, Johnson LF: Liver injury with alcoholic hyaline after intestinal resection. Gastroenterology 79:128-130, 1980 13. Craig RM, Neumann T, Jeejeebhoy KN, et al: Severe hepatocellular reaction resembling alcoholic hepatitis with cirrhosis after massive small bowel resection and prolonged total parenteral nutrition. Gastroenterology 79:131-137, 1980 14. Kimura H, Kako M, Yo K, et al: Alcoholic hyalins (Mallory bodies) in a case of Weber-Christian disease: Electron microscopic observations of liver involvement. Gastroenterology 78:807-8 12, 1980 15. Hamilton DL, Vest TK, Brown BS, et al: Liver injury with alcoholic like hyalin after gastroplasty for morbid obesity. Gastroenterology 85:722-726, 1983 16. Poucell S, Ireton J, Valencia-Mayoral P, et al: Amiodarone-associated phospholipidosis and fibrosis of the liver. Light, immunohistochemical, and electron microscopic studies. Gastroenterology 86:926-936, 1984 17. Simon JB, Manley PN, Brien JF, et al: Amiodarone hepatotoxicity simulating alcoholic liver disease. N Engl J Med 311:167-172, 1984 18. Pessayre D, Bichara M, Feldmann G, et al: Perhexiline maleate-induced cirrhosis. Gastroenterology 76:170-177, 1979 19. Itoh S, Igarashi M, Tsukada Y, et al: Nonalcoholic fatty liver with alcoholic hyalin after long-term glucocorticoid therapy. Acta Hepatogastroenterol 24:415-418, 1977 20. Seki K, Minami Y, Nishikawa M, et al: “Nonalcoholic steatohepatitis” induced by massive doses of synthetic estrogen. Gastroenterol Jpn 18: 197-203, 1983 21. Schlichting P, Fauerholdt L, Christensen E, et al: Clinical relevance of restrictive morphological criteria for the diagnosis of cirrhosis in liver biopsies. Liver 1:56-6 1, 198 1 22. Miller DJ, Ishimaru H, Klatskin G: Non-alcoholic liver disease mimicking alcoholic hepatitis and cirrhosis. Gastroenterology 77:A27, 1979 (abstr) 23. Itoh HS, Yougel T, Kawagoe K: Comparison between nonalcoholic steatohepatitis and alcoholic hepatitis. Am J Gastroenter01 82:650-654, 1987 24. Haines NW, Baker AL, Boyer JL, et al: Prognostic indicators of hepatic injury following jejunoileal bypass performed for refractory obesity: A prospective study. Hepatology 1:161-167, 1981 25. Vyberg M, Ravn V, Andersen B: Patterns of progression in liver injury followingjejunoileal bypass for morbid obesity. Liver 7:271-276, 1987 26. Galambos JT: Natural history of alcoholic hepatitis. III. Histological changes. Gastroenterology 63:1026-1035, 1972 27. Marbet UA, Bianchi L, Meury U, et al: Long-term histological evaluation of the natural history and prognostic factors of alcoholic liver disease. J Hepatol 4:364-372, 1987 28. Galambos JT, Wills CE: Relationship between 505 paired liver tests and biopsies in 242 obese patients. Gastroenterology 74:1191-1195, 1978 29. Thaler H: Relationship of steatosis to cirrhosis. Clin Gastroenterol4:273-280, 1975 30. Braillon A, Capron JP, Herve MA, et al: Liver in obesity. Gut 26:133-139, 1985 3 1. Andersen T, Gluud C: Liver morphology in morbid obesity: A literature study. Int J Obes 8:97-106, 1984 32. Capron J-P, Delamarre J, Dupus J-L, et al: Fasting in obesity. Another cause of liver injury with alcoholic hyaline? Dig Dis Sci 27:265-268, 1982 33. Diehl AM, Goodman 2, Ishak KG: Alcohol like liver disease in nonalcoholics. A clinical and histologic comparison with alcohol-induced liver injury. Gastroenterology 95: 1056-1062, 1988

The risk of cirrhosis in NASH can be contrasted with that of alcoholic hepatitis. Only two of 12 NASH patients developed cirrhosis, whereas this complication is noted in 38% to 50% of patients with alcoholic hepatitis after comparable follow-up.26~27 This difference further indicates that NASH is a distinct entity and that affected patients, despite the difficulty of excluding alcohol abuse in a given case, are not simply surreptitious alcoholics. The reason for the relative indolence of NASH is uncertain, but it suggests a less aggressive mechanism of liver cell injury than operates in alcoholic liver disease. These considerations lead to the concept that primary NASH is an intrinsically mild condition that can be potentiated by secondary aggravating factors to produce more severe fibrosis and cirrhosis. Such secondary factors may include minor alcohol consumption,30 deficient dietary protein, 3 l fasting,32 or possibly drugs that in normal hosts do not cause NASH. Jejunoileal bypass appears to be a similar aggravating factor, exacerbating the preexisting NASH found in massive obesity.24,25 In this study, unfortunately, no specific features distinguished those patients with and without progression so that this concept could not be tested. However, in a retrospective study of this kind, such aggravating factors might be difficult to establish. By selecting nondrinking patients without overt liver disease, this review may have yielded largely a subgroup of NASH patients with low-grade disease who lack prominent aggravating factors. Prospective analysis of patients with NASH, including histologic follow-up, will help answer these questions. A recent study by Diehl et al assessed the clinical and histologic features of NASH in comparison with alcoholic hepatitis and confirmed that NASH, although often clinically temperate, can progress and be potentially fata1.33 Acknowledgment. The helpful comments of Emmet B. Keeffe, MD, and secretarial skills of Phyllis Fritz are greatly appreciated.

REFERENCES 1. MacSween RNM, Burt AD: Histologic spectrum of alcoholic liver disease. Semin Liver Dis 6:221-232, 1986 2. Baptista A, Bianchi L, deGroote J, et al: Alcoholic liver disease: Morphologic manifestations. Lancet 1:707-711, 1981 3. Gerber MA, Popper H: Relation between central canals and portal tracts in alcoholic hepatitis. A contribution to the pathogenesis of cirrhosis in alcoholics. HUM PATHOL 3: 199-207, 1972 4. Adler M, Schaffner F: Fatty liver hepatitis obese patients. Am J Med 67:81 I-816, 1979

and cirrhosis

in

5. Ludwig J, Viggiano TR, McGill DB, et al: Nonalcoholic steatohepatitis. Mayo Clinic experiences with a hitherto unnamed disease. Mayo Clin Proc 55:434-438, 1980 Prog

6. Schaffner F, Thaler H: Nonalcoholic Liver Dis 8:283-298, 1986

fatty

liver

disease.

7. Peters RL, Gay T, Reynolds TB: Post-jejunoileal-bypass hepatic disease. Its similarity to alcoholic hepatic disease. Am J Clin Path01 63:318-331, 1975 8. Marubbio

AT Jr, Buchwald

H, Schwartz

MZ, et al: Hepatic

598