- Email: [email protected]
Clinical and histologic spectrum of nonalcoholic fatty liver disease associated with normal ALT values
Clinical and Histologic Spectrum of Nonalcoholic Fatty Liver Disease Associated With Normal ALT Values Pouneh Mofrad,1 Melissa J. Contos,2 Mahmadul Haque,1 Carol Sargeant,1 Robert A. Fisher,3 Velimir A. Luketic,1 Richard K. Sterling,1 Mitchell L. Shiffman,1 Richard T. Stravitz,1 and Arun J. Sanyal1 A retrospective study was performed to (1) characterize the clinical and histologic features of those with nonalcoholic fatty liver disease (NAFLD) and normal alanine aminotransferase (ALT) values, (2) compare the spectrum of NAFLD associated with normal versus elevated ALT levels, and (3) determine whether there were differences in the clinical or histologic spectrum of NAFLD between those with a low normal versus high normal ALT value. A total of 51 subjects with NAFLD and normal ALT were identified and compared with 50 consecutive subjects with NAFLD and elevated ALT. The major indications for liver biopsy in those with normal ALT were unexplained hepatomegaly (n ⴝ 21) and evaluation as a potential donor for living donor liver transplantation (n ⴝ 16). The 2 groups were comparable with respect to age, gender distribution, and ethnicity. Approximately 80% of cases in both groups had at least 1 feature of the metabolic syndrome, the major risk factor for NAFLD. The 2 groups were also comparable with respect to the grade of the individual histologic parameters of NAFLD. A total of 12 subjects with normal ALT levels had bridging fibrosis, whereas 6 had cirrhosis. Diabetes was the only factor independently associated with an increased risk of advanced fibrosis (bridging fibrosis or cirrhosis) by multivariate analysis (relative risk: 2.3, P < .01). The mean steatosis (1.6 vs. 2.16, P < .04) and perisinusoidal fibrosis scores (0.35 vs. 0.9, P < .049) were lower in those with low normal (<30 IU/L) ALT versus high normal ALT. However, the prevalence of advanced fibrosis was similar (5 of 15 vs. 13 of 36, respectively). In conclusion, (1) the entire histologic spectrum of NAFLD can be seen in individuals with normal ALT values, (2) the histologic spectrum in these individuals is not significantly different from those with elevated ALT levels, and (3) a low normal ALT value does not guarantee freedom from underlying steatohepatitis with advanced fibrosis. (HEPATOLOGY 2003;37:1286-1292.)
N
onalcoholic fatty liver disease (NAFLD) has recently been recognized to be one of the leading causes of chronic liver disease.1-3 This histology of this condition is characterized by either predominantly macrovesicular steatosis alone or steatohepatitis. The lat-
Abbreviations: NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; ALT, alanine aminotransferase. From the 1Division of Gastroenterology, Hepatology, and Nutrition, Department of Internal Medicine; 2Department of Pathology; and 3Division of Transplant Surgery, Virginia Commonwealth University, Richmond, VA. Received October 30, 2002; accepted March 9, 2003. Supported in part by a grant from the National Institutes of Health (K 24 DK 02755 to A.J.S.). This work was presented at an oral session at the annual meeting of the American College of Gastroenterology, Seattle, WA, October 2002. Address reprint requests to: Arun J. Sanyal, M.D., Professor of Medicine, MCV Box 980711, Richmond, VA 23298-0711. E-mail: [email protected]; fax: 804-828-2037. Copyright © 2003 by the American Association for the Study of Liver Diseases. 0270-9139/03/3706-0012$30.00/0 doi:10.1053/jhep.2003.50229 1286
ter condition includes steatosis along with varying combinations of findings, including cytologic ballooning, Mallory’s hyaline, pericellular fibrosis, and scattered inflammatory infiltrate.4 By definition, nonalcoholic fatty liver and nonalcoholic steatohepatitis (NASH) occur in individuals who do not consume alcohol in amounts generally considered to be harmful to the liver. Several published series have provided detailed information on the histologic and clinical spectrum of findings in patients with NAFLD.2,5-7 The majority of subjects in these studies were identified by the presence of elevated liver enzymes. It is, however, known that both fatty liver and NASH may exist without elevation of alanine aminotransferase (ALT).5,8 Although the full spectrum of NAFLD is known to occur in the presence of normal ALT levels, there have been no previous studies that have focused on those with NAFLD and normal ALT values or contrasted their clinical and histologic features with those having abnormal ALT levels. Based on an upper limit of
HEPATOLOGY, Vol. 37, No. 6, 2003
normal as 40 IU/L for ALT, a recent study has also suggested that some individuals with high “normal” ALT levels may have underlying NAFLD.9 We performed a retrospective study to (1) characterize the clinical and histologic features of those with NAFLD and normal ALT values, (2) compare the spectrum of NAFLD associated with normal versus elevated ALT levels, and (3) determine whether there were differences in the clinical or histologic spectrum of NAFLD between those with a low normal versus high normal ALT value.
Materials and Methods Case Identification. Based on an slide-based enzymatic method (Vitros Products; Johnson & Johnson Clinical Diagnostics, Rochester, NY), the range of normal ALT values at our institution is 9 to 52 IU/L for women and 21 to 75 IU/L for men. A retrospective analysis was performed on cases with NAFLD who had normal ALT values. Initially, liver biopsy records of all patients who had undergone liver biopsies between 1996 and 2002 were reviewed. Simultaneously, a clinical research database, which included all patients with NAFLD identified at the authors’ institution, was searched to identify those with normal ALT values. The 2 searches were matched to identify those with histologic evidence of fatty liver disease and normal ALT values. The clinical records of those identified by searching the biopsy records and the clinical research database were also reviewed. The study group was identified by a combination of normal ALT values, histologic evidence of fatty liver disease, and an attending physician’s assessment that the individual was not consuming alcohol in amounts considered to be harmful. For purposes of comparison, the study group was compared with a control group of 50 consecutive cases of NAFLD with elevated ALT values seen in the liver clinics at the Virginia Commonwealth University. All subjects had previously provided informed consent to have their data recorded in the database for the purpose of identification of the clinical spectrum and natural history of NAFLD, which was approved by the institutional review board of the Virginia Commonwealth University. Case Definitions. NAFLD was identified by both histologic and clinical criteria. The liver biopsy specimen was felt to represent fatty liver disease if either predominantly macrovesicular steatosis was present or if steatohepatitis was present. Steatohepatitis was defined by the minimal criteria of hepatic steatosis and scattered, mainly lobular inflammation with or without Mallory bodies, cytologic ballooning, and perisinusoidal fibrosis. Although it is arguable whether fat plus mild inflammation is a distinct category from steatosis alone, these minimal criteria were
MOFRAD ET AL.
1287
used to capture the full histologic spectrum of NAFLD. Being a retrospective study, it was not feasible to evaluate the amount of alcohol consumption using validated questionnaires. Thus, the degree of alcohol consumption was estimated from the clinic records. The fatty liver or steatohepatitis was considered to be nonalcoholic in origin if the clinical records indicated that the patient was felt to either be totally abstinent or to consume less than approximately 20 g of alcohol daily. In all of these cases, laboratory tests were reviewed to exclude the concomitant presence of viral hepatitis or other causes of chronic liver disease, e.g., hemochromatosis, Wilson’s disease, ␣1-antitrypsin deficiency, or others. Those with positive tests for these conditions were excluded. Histologic Analysis. The degree of steatosis was semiquantitatively graded from 0 to 4 as follows10: grade 1: up to 25% steatosis, grade 2: 26% to 50%, grade 3: 51% to 75%, and grade 4: more than 76%. The presence and severity of cytologic ballooning, Mallory’s hyaline, and pericellular fibrosis were evaluated as described by Brunt et al.11 Briefly, cytologic ballooning was graded as absent or present and, if present, as mild or marked. Mallory’s hyaline was graded from 0 to 2, with grade 0 ⫽ absent, grade 1 ⫽ occasional, and grade 2 ⫽ several. Glycogenated nuclei were similarly semiquantitatively graded from 0 to 3, depending on their prevalence. Perisinusoidal fibrosis was scored from 0 to 3 based on the percentage of zone 3 regions involved (0 ⫽ absent, 1 ⫽ up to 33%, 2 ⫽ 34% to 66%, and 3 ⫽ ⬎66%). The presence of bridging fibrosis or cirrhosis was also noted. Overall fibrosis stage was scored as follows: 0 ⫽ none, 1 ⫽ either pericentral/ pericellular or focal portal fibrosis, 3 ⫽ bridging fibrosis (central to central or central to portal), and 4 ⫽ cirrhosis. Clinical, Demographic, and Laboratory Data. Clinical information was obtained from patient records. Besides demographic data, information related to the indication for the liver biopsy and clinical features of NAFLD were all recorded. In the authors’ institution, all patients with documented NAFLD are seen in a dedicated multidisciplinary clinic for NAFLD, at which a trained nutritionist records parameters related to nutritional status. A full history and physical exam were available in all cases. Anthropometric data were obtained from the NAFLD clinical research database for these patients. The laboratory parameters that were identified included liver function tests (serum bilirubin, albumin, and prothrombin time) and the presence or absence of data indicative of alternate causes of liver disease, e.g., hepatitis C (HCV PCR), hepatitis B (hepatitis B surface antigen and core antibody), hemochromatosis (iron panel and Cys 282 Tyr mutation when available), Wilson’s disease (ceruloplasmin), -antitrypsin deficiency (␣1-antitrypsin
1288
MOFRAD ET AL.
HEPATOLOGY, June 2003
Table 1. Indications for Liver Biopsy Indication
Right upper quadrant discomfort ⫹ hepatomegaly Asymptomatic hepatomegaly Apparently healthy donors for liver transplantation Elevated serum ferritin Abnormal appearance of liver at cholecystectomy Premethotrexate liver biopsy Abnormal hepatic imaging study Clinical features of chronic liver disease Short-gut syndrome with elevated bilirubin
No.
11 10 16 2 3 2 2 4 1
level), and autoimmune hepatitis (antinuclear antibody). Finally, clinical and laboratory parameters related to the metabolic syndrome, as defined by Adult Treatment Panel III criteria,12 were also noted (hypertension, diabetes, obesity, hypertriglyceridemia). Data Analysis. For numerical parameters, e.g., triglyceride levels, the distribution as well as the mean and standard deviation were noted. Categorical data, e.g., presence or absence of diabetes, were recorded, and the proportion of the patient population having a given parameter was noted. Ordinal data, e.g., degree of cytologic ballooning, were graded, and the proportions of subjects with a given grade were noted. The factors predictive of advanced stage of NAFLD (bridging fibrosis or cirrhosis) were evaluated using multiple logistic regression analysis where the presence of advanced fibrosis was the dependent variable. A P value of .05 was considered to be significant.
Results A total of 386 cases of steatosis or steatohepatitis was identified from liver biopsies performed at the authors’ institution between 1996 and 2002. The clinical records of these individuals were retrieved, and 51 cases of nonalcoholic fatty liver or steatohepatitis with normal ALT values were identified. The principal indications for liver biopsy in these individuals were persistent hepatomegaly (n ⫽ 21) and donor evaluation for living donor liver transplantation (n ⫽ 16) (Table 1). The other indications included elevated ferritin levels, abnormal imaging characteristics of the liver suggestive of parenchymal liver disease, baseline biopsy prior to initiation of methotrexate therapy, and clinical features of portal hypertension without other evidence of liver disease. Demographic Data. The demographic features of the study group are shown in Table 2. The mean age of subjects with NAFLD and normal ALT values was 53 years, which was comparable with the general population of patients with NAFLD at the authors’ institution. The majority of the subjects were white (43 of 51), followed by
African Americans. Although this ethnic distribution matches that seen in the fatty liver disease clinic at Virginia Commonwealth University, the proportion of African Americans in both groups with NAFLD were substantially lower than those in the general liver clinics at this institution (8%-12% vs. 30%-40%, respectively). There were more women than men (35 vs. 15 or 16, respectively), which too was similar to the gender distribution of patients with NAFLD in the NAFLD clinic at the authors’ institution. Clinical Features and Comorbid Conditions. Fatigue and right upper quadrant discomfort were the most common symptoms seen in both those with normal ALT and the control group with elevated ALT values (Table 3). In the 3 patients in whom NAFLD was diagnosed by a liver biopsy performed at the time of cholecystectomy, there were no residual symptoms at the time of their evaluation for NAFLD in the liver clinic. Overall, a greater proportion of subjects with normal ALTs were completely asymptomatic compared with those with abnormal ALT values (20 of 51 vs. 10 of 50, respectively, P ⬍ .04, Fisher exact test). Approximately 60% (30 of 51 and 30 of 50; Table 3) of subjects in both groups had a relatively unremarkable physical examination, with the exception of their body weight and blood pressure. The majority of subjects in the normal ALT group as well as the control group were overweight (mean BMI 29 vs. 29.3, respectively); however, the 2 groups were comparable with respect to both the BMI as well as the waist-to-hip ratios (Table 2). Twentythree of 51 subjects with NAFLD and normal ALTs were obese, as defined by a BMI ⬎ 30,13 whereas 27 of 50 subjects with abnormal ALTs were obese. Twenty-nine of 51 patients with NAFLD and normal ALT levels were Table 2. Demographic and Clinical Features of Subjects With Normal Versus Abnormal ALT Parameter
Age (y), mean ⫾ SD Weight (lb), mean ⫾ SD BMI (kg/m2), mean ⫾ SD Waist-to-hip ratio, mean ⫾ SD Gender (M:F) Ethnicity White Hispanic African American Risk factors Obesity (BMI ⬎30) Diabetes mellitus Hypertension Hypertriglyceridemia Hypothyroidism Drugs known to cause NAFLD
Normal ALT (n ⴝ 51)
Abnormal ALT (n ⴝ 50)
P Value
53 ⫾ 7 210 ⫾ 32 29 ⫾ 3 0.89 ⫾ 0.07 16:35
52 ⫾ 6 220 ⫾ 40 29.3 ⫾ 3.8 0.91 ⫾ 0.08 15:35
NS NS NS NS NS
43 2 6
42 4 4
NS NS NS
23 29 24 12 1 0
27 24 20 18 1 0
NS NS NS NS NS –
HEPATOLOGY, Vol. 37, No. 6, 2003
MOFRAD ET AL.
1289
Table 3. Symptoms, Signs, and Laboratory Parameters in Those With NAFLD and Normal Versus Abnormal ALT Parameter
Symptoms Fatigue Right upper quadrant discomfort Pruritus Edema Encephalopathy Asymptomatic Signs Hepatomegaly Spider nevi Palmar erythema Jaundice Ascites Absence of findings of liver disease* Laboratory findings, mean ⫾ SD AST (IU/L) ULN: 75 IU/L ALT (IU/L) ULN: 75 IU/L Alk phos (IU/dL) ULN: 110 IU/dL Bilirubin (mg/dL) Albumin (gm/dL) Total cholesterol (mg/dL) Triglycerides (mg/dL) HDL/LDL cholesterol (mg/dL) Hemoglobin Alc (%)
No. Normal ALT (n ⴝ 51)
No. Abnormal ALT (n ⴝ 50)
31 13 2 1 0 20
35 10 1 0 0 10
NS NS NS NS – .04
21 4 1 0 0 30
19 2 1 0 0 31
NS NS NS – – NS
44 ⫾ 18 40 ⫾ 15 123 ⫾ 56 0.3 ⫾ 0.1 4 ⫾ 0.3 186 ⫾ 46 189 ⫾ 116 48 ⫾ 21/117 ⫾ 46 6.5 ⫾ 2.3
110 ⫾ 25 175 ⫾ 45 129 ⫾ 35 0.4 ⫾ 0.1 4.2 ⫾ 0.3 210 ⫾ 35 220 ⫾ 89 47 ⫾ 19/130 ⫾45 6.2 ⫾ 1.8
.0001 .0001 NS NS NS .004 NS NS NS
P Value
*There were no abnormal physical findings related to liver disease. Many of these subjects were, however, overweight or frankly obese.
diabetic, and 24 had hypertension. The proportion of subjects with features of metabolic syndrome was similar between those with or without normal ALT values. One patient each from the normal ALT group and the elevated ALT group had hypothyroidism. None of the subjects in either the study (normal ALT) or control (elevated ALT) group were taking medications known to be associated with the development of hepatic steatosis. As expected, the 2 groups differed significantly with respect to their AST and ALT levels. The alkaline phosphatase levels were comparable between the 2 groups. ␥-Glutamyltransferase levels are not part of the hepatic panel, and these data were not available. Also, the hepatic synthetic functions (serum albumin and prothrombin time) were normal in the majority of individuals and comparable across the 2 groups. Two patients with normal ALT values had mild hypoalbuminemia, whereas 1 patient with NASH and elevated ALT levels had hypoalbuminemia. The latter patient had diabetic nephropathy with proteinuria, whereas the 2 patients with normal ALT levels only had liver disease to account for their low serum albumin levels. Histologic Findings and Clinico-Histologic Correlations. Four broad histologic patterns were recognized as described by Matteoni et al.8 (Table 4): (1) fat alone, (2) fat ⫹ scattered inflammation, (3) fat ⫹ cytologic ballooning ⫾ inflammation, and (4) fat ⫹ cytologic balloon-
ing ⫾ Mallory’s hyaline ⫾ perisinusoidal fibrosis. The majority of subjects in both the normal ALT and abnormal ALT groups were in categories 3 and 4 above. The distribution of the histologic findings of NAFLD in those with normal ALT is shown in Table 5. The degree of steatosis was mild to moderate in the majority of individuals and similar to that in the control group with abnormal ALT levels (Fig. 1). Similarly, the mean grades of cytologic ballooning, pericellular fibrosis, Mallory’s hyaline, and inflammation were comparable across the 2 groups. Bridging fibrosis was noted in a total of 12 subjects; 4 of 12 had central to portal bridges, whereas 8 of 12 had central to central bridges. Five instances each of bridging fibrosis were seen in those being evaluated as
Table 4. Patterns of Histologic Findings in Those With NAFLD Pattern
Fat alone Fat ⫹ scattered inflammation Fat ⫹ ballooning ⫹ inflammation Fat ⫹ ballooning ⫾ Mallory hyaline ⫾ pericellular fibrosis
No. Normal ALT (n ⴝ 51)
No. Abnormal ALT (n ⴝ 50)
P Value
8
10
NS
8
10
NS
13
11
NS
22
19
NS
1290
MOFRAD ET AL.
HEPATOLOGY, June 2003
Table 5. Histologic Findings of NAFLD in Those With Normal ALT Parameter
Steatosis Cytologic ballooning Pericellular fibrosis Mallory’s hyaline Glycogen nuclei Lobular inflammation Portal inflammation Overall fibrosis
Grade 0
Grade 1
Grade 2
Grade 3
Grade 4
0 14 24 28 20 11 12 21
14 25 18 19 22 31 27 20
24 11 7 4 5 7 0 –
12 1 2 – 4 2 0 12*
1 – – – – – 0 6
*Four subjects had central to portal fibrosis. An additional 8 subjects had central to central bridging fibrosis.
donors for liver donor transplant and in those undergoing liver biopsy for hepatomegaly. The remaining 2 cases were seen in 2 subjects with hyperferritinemia. Of note, 6 patients with normal ALTs had biopsyproven cirrhosis. In 3 cases, the biopsy was performed for right upper quadrant discomfort and hepatomegaly. In the other 3 cases, there was clinical evidence of chronic liver disease. There was no relationship between ALT values and the presence of advanced fibrosis (bridging fibrosis or cirrhosis) (Fig. 2). Within the group with normal ALT levels, the mean BMI was somewhat higher in those with cirrhosis compared with those without (32 vs. 28.6, respectively). Also, 4 of the 6 subjects with cirrhosis had diabetes. To determine which clinical parameters were predictive of advanced fibrosis, multiple logistic regression analysis was performed when advanced fibrosis (bridging fibrosis or cirrhosis) was the dependent variable. Of the variables entered in the model (age, BMI, steatosis grade, diabetes, hepatomegaly alone, and hepatomegaly with the metabolic syndrome), only the presence of dia-
betes was independently associated with the presence of advanced fibrosis (odds ratio: 2.3 [1.5-3.1], P ⬍ .003). Subset Analysis. Twenty-one subjects with NAFLD and normal ALTs were initially evaluated for hepatomegaly, whereas 16 subjects had been initially evaluated as donors for liver transplantation. None of the individuals evaluated as transplantation donors had cirrhosis. However, 5 of 16 had bridging fibrosis. These 16 subjects were part of 105 subjects evaluated as donors. The liver sonograms were abnormal in 12 of these 16 cases and showed a heterogeneous echotexture in 10 of 16 and hyperechogenecity of the liver in 8 of 16 individuals. Overall, 4 of 5 instances of bridging fibrosis in this group were seen in those with an abnormal sonogram. On the other hand, 3 of 21 subjects undergoing liver biopsy for hepatomegaly had cirrhosis, and another 5 of 21 had bridging fibrosis. Thus, about 40% of patients with persistent hepatomegaly had evidence of advanced fibrosis. There was a high prevalence (6 of 14) of advanced fibrosis within the subset of subjects with hepatomegaly and diabetes.
Fig. 1. The mean (⫾SD) grade of the individual histologic findings of NAFLD in 51 subjects with NAFLD and normal ALT (■) and 50 subjects with NAFLD and elevated ALT levels ( ). There was no significant difference with respect to the grade of steatosis, cytologic ballooning, Mallory bodies, pericellular fibrosis, or portal fibrosis. These data indicate that the histologic spectrum of NAFLD in those with normal ALT are comparable with those with elevated ALT.
Fig. 2. The relationship between the fibrosis stage and the ALT value within the normal range is shown in 51 subjects with NAFLD and normal ALT levels. Fibrosis was scored as follows: 0 ⫽ none, 1 ⫽ either pericentral/pericellular or focal portal fibrosis, 3 ⫽ bridging fibrosis, 4 ⫽ cirrhosis. Five of 15 subjects with NAFLD and low normal ALT levels (⬍30 IU/L) had advanced fibrosis compared with 13 of 36 individuals with high normal ALT levels (31-75 IU/L) (P ⬍ 1, Fisher exact test).
HEPATOLOGY, Vol. 37, No. 6, 2003
MOFRAD ET AL.
Table 6. Comparison of Those With Low Normal Versus High Normal ALT Values
Parameter
Age (y), mean ⫾ SD Gender (M:F) Ethnicity (W:A:H) Weight (lb), mean ⫾ SD Diabetes (n) Hypertension (n) Triglycerides (mg/dL), mean ⫾ SD Histology, mean ⫾ SD Steatosis (grade) Cytologic ballooning (grade) Pericellular fibrosis (grade) Overall fibrosis (stage)
ALT (<30 IU/L) (n ⴝ 15)
ALT (31-75 IU/L) (n ⴝ 36)
P Value
51.8 ⫾ 5 3:12 12:3:0 206 ⫾ 34 11 8 175 ⫾ 177
54.1 ⫾ 7.5 13:23 31:3:2 212 ⫾ 31 18 16 181 ⫾ 96
NS NS NS NS NS NS NS
1.6 ⫾ 0.6 0.78 ⫾ 0.7 0.35 ⫾ 0.6 1.5 ⫾ 1.5
2.16 ⫾ 0.8 1.05 ⫾ 0.7 0.9 ⫾ 0.8 1.37 ⫾ 1.5
.04 NS .049 NS
NOTE. Numerical data were compared using unpaired t test. Proportions were compared using Fisher exact test and mean histologic grades compared by Mann Whitney test. Abbreviations: W, white; A, African American; H, Hispanic.
Fifteen of 51 subjects had ALT values less than 30 IU/L. Although the age and gender distribution in these subjects was comparable with those with ALT values between 30 and 75 IU/L (Table 6), those with low normal ALTs had a significantly lower grade of steatosis and pericellular fibrosis. Eleven of 15 subjects with low normal ALT levels were diabetic compared with 18 of 36 subjects with high normal ALT levels (P ⫽ .2 by Fisher exact test), whereas 5 of 15 subjects had advanced fibrosis compared with 13 of 36 subjects with high normal ALT values (P ⫽ 1, Fisher exact test).
Discussion The serum ALT value has long been used as a surrogate marker of liver injury.14,15 It is, however, well known that the ALT values do not correlate well with the severity of liver disease noted on liver biopsy in subjects with chronic liver disease.15 This is particularly true when one considers the stage of fibrosis present in an individual patient. The present study confirms this to be true also for NAFLD. The upper limit of normal for ALT for a given laboratory depends both on the methodology used as well as on the range of values established for apparently normal individuals in the region. The method used in our institutional laboratory is the commercially available Vitros ALT slide (Johnson & Johnson), which is a dry, multilayered analytical element coated on a polyester support. An 11-L drop of serum is placed on the slide and is evenly distributed from the spreading layer to the underlying layers. Alanine and ␣-ketoglutarate in the spreading layer are converted by the ALT in the serum to pyruvate and
1291
glutamate. Pyruvate is then converted to lactate with the conversion of NADH to NAD. The rate of oxidation of NADH is measured by reflectance spectrophotometry, and the rate of change in reflection density is converted to enzyme activity. When pyridoxal phosphate is used as a cofactor for the initial reaction, the resulting ALT values are about 15 to 20 units higher than in its absence, producing normal ALT levels up to 72 IU/L as noted in the manufacturers guide. The methods are rigorously tested for quality control in accordance with the manufacturer’s guidelines and the College of American Pathologists. These values were further modified to provide an upper limit of 75 IU/L based on the mean and standard deviations of values obtained from 2 cohorts of apparently normal individuals about 15 years ago. Thus, although the range of normal ALT values appears to be higher than those used by many laboratories, this is mostly due to the methods used in our laboratory. Another potential explanation of the finding of significant pathology in the face of normal ALT values is that the supposedly normal range of values is incorrect. The upper limits of normal for ALT values were established over 40 years ago and represent the mean ⫾ SD of the ALT values of a group of individuals felt to be free of liver disease.14 For a given laboratory, the upper limits of normal are also based on ALT values in apparently normal individuals in that area. This has recently been questioned, and, in a large, recent study,9 it was shown that many individuals with supposedly normal ALT levels had biochemical evidence of the metabolic syndrome, the major risk factor for NAFLD. Also, ALT values correlated directly with the body mass index, another risk factor for NAFLD. Based on these data, it was suggested that the upper limits of normal for ALT be decreased to 30 IU/L for men and 19 IU/L for women. In our study, 15 of 51 subjects had an ALT less than 30 IU/L. Although the degree of steatosis and pericellular fibrosis was less in this subset of subjects, there were no significant differences in the prevalence of advanced fibrosis between those with a low normal ALT (⬍30 IU/L) and high normal ALT (31-75 IU/L). Also, all stages of NAFLD were present in those with normal ALT values. Thus, although the public health and clinical utility of reducing the upper limits of normal for ALT continues to be debated,16 the present study provides additional data that, even at low ALT levels, significant liver disease can be present. It is interesting to note the relatively high prevalence of bridging fibrosis or cirrhosis in those with NAFLD and normal ALTs. This may be at least partly due to the bias inherent in a retrospective analysis. The differences in the rates of advanced fibrosis among those undergoing biopsy for hepatomegaly versus asymptomatic and apparently
1292
MOFRAD ET AL.
healthy individuals being evaluated as transplant donors indeed support this possibility. On the other hand, the data from the transplant donor group also indicate that the absence of any obvious symptoms, signs, or ALT values suggestive of liver disease does not guarantee the absence of clinically significant liver disease, i.e., advanced stage chronic liver disease. In this respect, these data corroborate and extend the existing literature regarding other chronic liver diseases, e.g., hepatitis C and normal ALT values.17 An important question is the following: How does one identify subjects with clinically significant NAFLD in the absence of symptoms, signs, or abnormal ALT values? It is clearly both unreasonable and impractical to perform routine liver biopsies for this purpose. Although much more work needs to be done to determine who should be screened for NAFLD and how such individuals should be evaluated, the current studies suggest that those with hepatomegaly, especially in the presence of diabetes, are more likely to have clinically significant underlying pathology. In summary, patients with NAFLD and normal ALT values may have steatohepatitis and advanced fibrosis. Diabetes is an independent predictor of advanced fibrosis. Although further work remains to be done to answer the clinical questions raised by these data, this study is an initial step toward the identification, characterization, and management of NAFLD in patients with normal ALTs.
References 1. Bellentani S, Saccoccio G, Masutti F, Croce LS, Brandi G, Sasso F, Cristanini G, et al. Prevalence of and risk factors for hepatic steatosis in Northern Italy. Ann Intern Med 2000;132:112-117.
HEPATOLOGY, June 2003
2. Wanless IR, Lentz JS. Fatty liver hepatitis (steatohepatitis) and obesity: an autopsy study with analysis of risk factors. HEPATOLOGY 1990;11:74-80. 3. Ground KEV. Prevalence of fatty liver in healthy adults accidentally killed. Aviat Space Environ Med 1984;55:59-61. 4. Ludwig J, Viggiano TR, McGill DB, Oh BJ. Nonalcoholic steatohepatitis: Mayo Clinic experiences with a hitherto unnamed disease. Mayo Clin Proc 1980;55:434-438. 5. Bacon BR, Farahvash MJ, Janney CG, Neuschwander Tetri BA. Nonalcoholic steatohepatitis: an expanded clinical entity. Gastroenterology 1994;107:1103-1109. 6. Diehl AM, Goodman Z, Ishak KG. Alcoholic liver disease in nonalcoholics. A clinical and histologic comparison with alcohol-induced liver injury. Gastroenterology 1988;95:1056-1062. 7. Powell EE, Cooksley WG, Hanson R, Searle J, Halliday JW, Powell LW. The natural history of nonalcoholic steatohepatitis: a follow-up study of forty-two patients for up to 21 years. HEPATOLOGY 1990;11:74-80. 8. Matteoni CA, Younossi ZM, Gramlich T, Boparai N, Liu YC, McCullough AJ. Nonalcoholic fatty liver disease: a spectrum of clinical and pathological severity. Gastroenterology 1999;116:1413-1419. 9. Prati D, Taioli E, Zanella A, Della TE, Butelli S, Del Vecchio E, Vianello L, et al. Updated definitions of healthy ranges for serum alanine aminotransferase levels. Ann Intern Med 2002;137:1-10. 10. Contos MJ, Cales W, Luketic VA, Sterling RK, Shiffman ML, Mills AS, Fisher RA, et al. Development of nonalcoholic fatty liver disease after liver transplant for cryptogenic cirrhosis. Liver Transpl 2001;7:363-373. 11. Brunt EM, Janney CG, Di Bisceglie AM, Neuschwander-Tetri BA, Bacon BR. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol 1999;94:2467-2474. 12. National Institutes of Health. Third report of the national cholesterol education program expert panel on detection, evaluation and treatment of high blood cholesterol in adults (Adult treatment panel III). NIH Publication 2001:1-3670. 13. Anonymous. Executive summary of the clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. Arch Intern Med 1998;158:1855-1867. 14. Karmen A, Wroblewski F, LaDue JS. Transaminase activity in human blood. J Clin Invest 1955;34:126-133. 15. Kallei L, Hahn A, Roder VZ. Correlation between histological findings and serum transaminase values in chronic diseases of the liver. Acta Med Scand 1964;175:49-56. 16. Kaplan MM. Alanine aminotransferase levels: what’s normal? Ann Intern Med 2002;137:49-51. 17. Alberti A, Morsica G, Chemello L. Hepatitis C viraemia and liver disease in symptom-free individuals with anti-HCV. Lancet 1992;340:697.
N
onalcoholic fatty liver disease (NAFLD) has recently been recognized to be one of the leading causes of chronic liver disease.1-3 This histology of this condition is characterized by either predominantly macrovesicular steatosis alone or steatohepatitis. The lat-
Abbreviations: NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; ALT, alanine aminotransferase. From the 1Division of Gastroenterology, Hepatology, and Nutrition, Department of Internal Medicine; 2Department of Pathology; and 3Division of Transplant Surgery, Virginia Commonwealth University, Richmond, VA. Received October 30, 2002; accepted March 9, 2003. Supported in part by a grant from the National Institutes of Health (K 24 DK 02755 to A.J.S.). This work was presented at an oral session at the annual meeting of the American College of Gastroenterology, Seattle, WA, October 2002. Address reprint requests to: Arun J. Sanyal, M.D., Professor of Medicine, MCV Box 980711, Richmond, VA 23298-0711. E-mail: [email protected]; fax: 804-828-2037. Copyright © 2003 by the American Association for the Study of Liver Diseases. 0270-9139/03/3706-0012$30.00/0 doi:10.1053/jhep.2003.50229 1286
ter condition includes steatosis along with varying combinations of findings, including cytologic ballooning, Mallory’s hyaline, pericellular fibrosis, and scattered inflammatory infiltrate.4 By definition, nonalcoholic fatty liver and nonalcoholic steatohepatitis (NASH) occur in individuals who do not consume alcohol in amounts generally considered to be harmful to the liver. Several published series have provided detailed information on the histologic and clinical spectrum of findings in patients with NAFLD.2,5-7 The majority of subjects in these studies were identified by the presence of elevated liver enzymes. It is, however, known that both fatty liver and NASH may exist without elevation of alanine aminotransferase (ALT).5,8 Although the full spectrum of NAFLD is known to occur in the presence of normal ALT levels, there have been no previous studies that have focused on those with NAFLD and normal ALT values or contrasted their clinical and histologic features with those having abnormal ALT levels. Based on an upper limit of
HEPATOLOGY, Vol. 37, No. 6, 2003
normal as 40 IU/L for ALT, a recent study has also suggested that some individuals with high “normal” ALT levels may have underlying NAFLD.9 We performed a retrospective study to (1) characterize the clinical and histologic features of those with NAFLD and normal ALT values, (2) compare the spectrum of NAFLD associated with normal versus elevated ALT levels, and (3) determine whether there were differences in the clinical or histologic spectrum of NAFLD between those with a low normal versus high normal ALT value.
Materials and Methods Case Identification. Based on an slide-based enzymatic method (Vitros Products; Johnson & Johnson Clinical Diagnostics, Rochester, NY), the range of normal ALT values at our institution is 9 to 52 IU/L for women and 21 to 75 IU/L for men. A retrospective analysis was performed on cases with NAFLD who had normal ALT values. Initially, liver biopsy records of all patients who had undergone liver biopsies between 1996 and 2002 were reviewed. Simultaneously, a clinical research database, which included all patients with NAFLD identified at the authors’ institution, was searched to identify those with normal ALT values. The 2 searches were matched to identify those with histologic evidence of fatty liver disease and normal ALT values. The clinical records of those identified by searching the biopsy records and the clinical research database were also reviewed. The study group was identified by a combination of normal ALT values, histologic evidence of fatty liver disease, and an attending physician’s assessment that the individual was not consuming alcohol in amounts considered to be harmful. For purposes of comparison, the study group was compared with a control group of 50 consecutive cases of NAFLD with elevated ALT values seen in the liver clinics at the Virginia Commonwealth University. All subjects had previously provided informed consent to have their data recorded in the database for the purpose of identification of the clinical spectrum and natural history of NAFLD, which was approved by the institutional review board of the Virginia Commonwealth University. Case Definitions. NAFLD was identified by both histologic and clinical criteria. The liver biopsy specimen was felt to represent fatty liver disease if either predominantly macrovesicular steatosis was present or if steatohepatitis was present. Steatohepatitis was defined by the minimal criteria of hepatic steatosis and scattered, mainly lobular inflammation with or without Mallory bodies, cytologic ballooning, and perisinusoidal fibrosis. Although it is arguable whether fat plus mild inflammation is a distinct category from steatosis alone, these minimal criteria were
MOFRAD ET AL.
1287
used to capture the full histologic spectrum of NAFLD. Being a retrospective study, it was not feasible to evaluate the amount of alcohol consumption using validated questionnaires. Thus, the degree of alcohol consumption was estimated from the clinic records. The fatty liver or steatohepatitis was considered to be nonalcoholic in origin if the clinical records indicated that the patient was felt to either be totally abstinent or to consume less than approximately 20 g of alcohol daily. In all of these cases, laboratory tests were reviewed to exclude the concomitant presence of viral hepatitis or other causes of chronic liver disease, e.g., hemochromatosis, Wilson’s disease, ␣1-antitrypsin deficiency, or others. Those with positive tests for these conditions were excluded. Histologic Analysis. The degree of steatosis was semiquantitatively graded from 0 to 4 as follows10: grade 1: up to 25% steatosis, grade 2: 26% to 50%, grade 3: 51% to 75%, and grade 4: more than 76%. The presence and severity of cytologic ballooning, Mallory’s hyaline, and pericellular fibrosis were evaluated as described by Brunt et al.11 Briefly, cytologic ballooning was graded as absent or present and, if present, as mild or marked. Mallory’s hyaline was graded from 0 to 2, with grade 0 ⫽ absent, grade 1 ⫽ occasional, and grade 2 ⫽ several. Glycogenated nuclei were similarly semiquantitatively graded from 0 to 3, depending on their prevalence. Perisinusoidal fibrosis was scored from 0 to 3 based on the percentage of zone 3 regions involved (0 ⫽ absent, 1 ⫽ up to 33%, 2 ⫽ 34% to 66%, and 3 ⫽ ⬎66%). The presence of bridging fibrosis or cirrhosis was also noted. Overall fibrosis stage was scored as follows: 0 ⫽ none, 1 ⫽ either pericentral/ pericellular or focal portal fibrosis, 3 ⫽ bridging fibrosis (central to central or central to portal), and 4 ⫽ cirrhosis. Clinical, Demographic, and Laboratory Data. Clinical information was obtained from patient records. Besides demographic data, information related to the indication for the liver biopsy and clinical features of NAFLD were all recorded. In the authors’ institution, all patients with documented NAFLD are seen in a dedicated multidisciplinary clinic for NAFLD, at which a trained nutritionist records parameters related to nutritional status. A full history and physical exam were available in all cases. Anthropometric data were obtained from the NAFLD clinical research database for these patients. The laboratory parameters that were identified included liver function tests (serum bilirubin, albumin, and prothrombin time) and the presence or absence of data indicative of alternate causes of liver disease, e.g., hepatitis C (HCV PCR), hepatitis B (hepatitis B surface antigen and core antibody), hemochromatosis (iron panel and Cys 282 Tyr mutation when available), Wilson’s disease (ceruloplasmin), -antitrypsin deficiency (␣1-antitrypsin
1288
MOFRAD ET AL.
HEPATOLOGY, June 2003
Table 1. Indications for Liver Biopsy Indication
Right upper quadrant discomfort ⫹ hepatomegaly Asymptomatic hepatomegaly Apparently healthy donors for liver transplantation Elevated serum ferritin Abnormal appearance of liver at cholecystectomy Premethotrexate liver biopsy Abnormal hepatic imaging study Clinical features of chronic liver disease Short-gut syndrome with elevated bilirubin
No.
11 10 16 2 3 2 2 4 1
level), and autoimmune hepatitis (antinuclear antibody). Finally, clinical and laboratory parameters related to the metabolic syndrome, as defined by Adult Treatment Panel III criteria,12 were also noted (hypertension, diabetes, obesity, hypertriglyceridemia). Data Analysis. For numerical parameters, e.g., triglyceride levels, the distribution as well as the mean and standard deviation were noted. Categorical data, e.g., presence or absence of diabetes, were recorded, and the proportion of the patient population having a given parameter was noted. Ordinal data, e.g., degree of cytologic ballooning, were graded, and the proportions of subjects with a given grade were noted. The factors predictive of advanced stage of NAFLD (bridging fibrosis or cirrhosis) were evaluated using multiple logistic regression analysis where the presence of advanced fibrosis was the dependent variable. A P value of .05 was considered to be significant.
Results A total of 386 cases of steatosis or steatohepatitis was identified from liver biopsies performed at the authors’ institution between 1996 and 2002. The clinical records of these individuals were retrieved, and 51 cases of nonalcoholic fatty liver or steatohepatitis with normal ALT values were identified. The principal indications for liver biopsy in these individuals were persistent hepatomegaly (n ⫽ 21) and donor evaluation for living donor liver transplantation (n ⫽ 16) (Table 1). The other indications included elevated ferritin levels, abnormal imaging characteristics of the liver suggestive of parenchymal liver disease, baseline biopsy prior to initiation of methotrexate therapy, and clinical features of portal hypertension without other evidence of liver disease. Demographic Data. The demographic features of the study group are shown in Table 2. The mean age of subjects with NAFLD and normal ALT values was 53 years, which was comparable with the general population of patients with NAFLD at the authors’ institution. The majority of the subjects were white (43 of 51), followed by
African Americans. Although this ethnic distribution matches that seen in the fatty liver disease clinic at Virginia Commonwealth University, the proportion of African Americans in both groups with NAFLD were substantially lower than those in the general liver clinics at this institution (8%-12% vs. 30%-40%, respectively). There were more women than men (35 vs. 15 or 16, respectively), which too was similar to the gender distribution of patients with NAFLD in the NAFLD clinic at the authors’ institution. Clinical Features and Comorbid Conditions. Fatigue and right upper quadrant discomfort were the most common symptoms seen in both those with normal ALT and the control group with elevated ALT values (Table 3). In the 3 patients in whom NAFLD was diagnosed by a liver biopsy performed at the time of cholecystectomy, there were no residual symptoms at the time of their evaluation for NAFLD in the liver clinic. Overall, a greater proportion of subjects with normal ALTs were completely asymptomatic compared with those with abnormal ALT values (20 of 51 vs. 10 of 50, respectively, P ⬍ .04, Fisher exact test). Approximately 60% (30 of 51 and 30 of 50; Table 3) of subjects in both groups had a relatively unremarkable physical examination, with the exception of their body weight and blood pressure. The majority of subjects in the normal ALT group as well as the control group were overweight (mean BMI 29 vs. 29.3, respectively); however, the 2 groups were comparable with respect to both the BMI as well as the waist-to-hip ratios (Table 2). Twentythree of 51 subjects with NAFLD and normal ALTs were obese, as defined by a BMI ⬎ 30,13 whereas 27 of 50 subjects with abnormal ALTs were obese. Twenty-nine of 51 patients with NAFLD and normal ALT levels were Table 2. Demographic and Clinical Features of Subjects With Normal Versus Abnormal ALT Parameter
Age (y), mean ⫾ SD Weight (lb), mean ⫾ SD BMI (kg/m2), mean ⫾ SD Waist-to-hip ratio, mean ⫾ SD Gender (M:F) Ethnicity White Hispanic African American Risk factors Obesity (BMI ⬎30) Diabetes mellitus Hypertension Hypertriglyceridemia Hypothyroidism Drugs known to cause NAFLD
Normal ALT (n ⴝ 51)
Abnormal ALT (n ⴝ 50)
P Value
53 ⫾ 7 210 ⫾ 32 29 ⫾ 3 0.89 ⫾ 0.07 16:35
52 ⫾ 6 220 ⫾ 40 29.3 ⫾ 3.8 0.91 ⫾ 0.08 15:35
NS NS NS NS NS
43 2 6
42 4 4
NS NS NS
23 29 24 12 1 0
27 24 20 18 1 0
NS NS NS NS NS –
HEPATOLOGY, Vol. 37, No. 6, 2003
MOFRAD ET AL.
1289
Table 3. Symptoms, Signs, and Laboratory Parameters in Those With NAFLD and Normal Versus Abnormal ALT Parameter
Symptoms Fatigue Right upper quadrant discomfort Pruritus Edema Encephalopathy Asymptomatic Signs Hepatomegaly Spider nevi Palmar erythema Jaundice Ascites Absence of findings of liver disease* Laboratory findings, mean ⫾ SD AST (IU/L) ULN: 75 IU/L ALT (IU/L) ULN: 75 IU/L Alk phos (IU/dL) ULN: 110 IU/dL Bilirubin (mg/dL) Albumin (gm/dL) Total cholesterol (mg/dL) Triglycerides (mg/dL) HDL/LDL cholesterol (mg/dL) Hemoglobin Alc (%)
No. Normal ALT (n ⴝ 51)
No. Abnormal ALT (n ⴝ 50)
31 13 2 1 0 20
35 10 1 0 0 10
NS NS NS NS – .04
21 4 1 0 0 30
19 2 1 0 0 31
NS NS NS – – NS
44 ⫾ 18 40 ⫾ 15 123 ⫾ 56 0.3 ⫾ 0.1 4 ⫾ 0.3 186 ⫾ 46 189 ⫾ 116 48 ⫾ 21/117 ⫾ 46 6.5 ⫾ 2.3
110 ⫾ 25 175 ⫾ 45 129 ⫾ 35 0.4 ⫾ 0.1 4.2 ⫾ 0.3 210 ⫾ 35 220 ⫾ 89 47 ⫾ 19/130 ⫾45 6.2 ⫾ 1.8
.0001 .0001 NS NS NS .004 NS NS NS
P Value
*There were no abnormal physical findings related to liver disease. Many of these subjects were, however, overweight or frankly obese.
diabetic, and 24 had hypertension. The proportion of subjects with features of metabolic syndrome was similar between those with or without normal ALT values. One patient each from the normal ALT group and the elevated ALT group had hypothyroidism. None of the subjects in either the study (normal ALT) or control (elevated ALT) group were taking medications known to be associated with the development of hepatic steatosis. As expected, the 2 groups differed significantly with respect to their AST and ALT levels. The alkaline phosphatase levels were comparable between the 2 groups. ␥-Glutamyltransferase levels are not part of the hepatic panel, and these data were not available. Also, the hepatic synthetic functions (serum albumin and prothrombin time) were normal in the majority of individuals and comparable across the 2 groups. Two patients with normal ALT values had mild hypoalbuminemia, whereas 1 patient with NASH and elevated ALT levels had hypoalbuminemia. The latter patient had diabetic nephropathy with proteinuria, whereas the 2 patients with normal ALT levels only had liver disease to account for their low serum albumin levels. Histologic Findings and Clinico-Histologic Correlations. Four broad histologic patterns were recognized as described by Matteoni et al.8 (Table 4): (1) fat alone, (2) fat ⫹ scattered inflammation, (3) fat ⫹ cytologic ballooning ⫾ inflammation, and (4) fat ⫹ cytologic balloon-
ing ⫾ Mallory’s hyaline ⫾ perisinusoidal fibrosis. The majority of subjects in both the normal ALT and abnormal ALT groups were in categories 3 and 4 above. The distribution of the histologic findings of NAFLD in those with normal ALT is shown in Table 5. The degree of steatosis was mild to moderate in the majority of individuals and similar to that in the control group with abnormal ALT levels (Fig. 1). Similarly, the mean grades of cytologic ballooning, pericellular fibrosis, Mallory’s hyaline, and inflammation were comparable across the 2 groups. Bridging fibrosis was noted in a total of 12 subjects; 4 of 12 had central to portal bridges, whereas 8 of 12 had central to central bridges. Five instances each of bridging fibrosis were seen in those being evaluated as
Table 4. Patterns of Histologic Findings in Those With NAFLD Pattern
Fat alone Fat ⫹ scattered inflammation Fat ⫹ ballooning ⫹ inflammation Fat ⫹ ballooning ⫾ Mallory hyaline ⫾ pericellular fibrosis
No. Normal ALT (n ⴝ 51)
No. Abnormal ALT (n ⴝ 50)
P Value
8
10
NS
8
10
NS
13
11
NS
22
19
NS
1290
MOFRAD ET AL.
HEPATOLOGY, June 2003
Table 5. Histologic Findings of NAFLD in Those With Normal ALT Parameter
Steatosis Cytologic ballooning Pericellular fibrosis Mallory’s hyaline Glycogen nuclei Lobular inflammation Portal inflammation Overall fibrosis
Grade 0
Grade 1
Grade 2
Grade 3
Grade 4
0 14 24 28 20 11 12 21
14 25 18 19 22 31 27 20
24 11 7 4 5 7 0 –
12 1 2 – 4 2 0 12*
1 – – – – – 0 6
*Four subjects had central to portal fibrosis. An additional 8 subjects had central to central bridging fibrosis.
donors for liver donor transplant and in those undergoing liver biopsy for hepatomegaly. The remaining 2 cases were seen in 2 subjects with hyperferritinemia. Of note, 6 patients with normal ALTs had biopsyproven cirrhosis. In 3 cases, the biopsy was performed for right upper quadrant discomfort and hepatomegaly. In the other 3 cases, there was clinical evidence of chronic liver disease. There was no relationship between ALT values and the presence of advanced fibrosis (bridging fibrosis or cirrhosis) (Fig. 2). Within the group with normal ALT levels, the mean BMI was somewhat higher in those with cirrhosis compared with those without (32 vs. 28.6, respectively). Also, 4 of the 6 subjects with cirrhosis had diabetes. To determine which clinical parameters were predictive of advanced fibrosis, multiple logistic regression analysis was performed when advanced fibrosis (bridging fibrosis or cirrhosis) was the dependent variable. Of the variables entered in the model (age, BMI, steatosis grade, diabetes, hepatomegaly alone, and hepatomegaly with the metabolic syndrome), only the presence of dia-
betes was independently associated with the presence of advanced fibrosis (odds ratio: 2.3 [1.5-3.1], P ⬍ .003). Subset Analysis. Twenty-one subjects with NAFLD and normal ALTs were initially evaluated for hepatomegaly, whereas 16 subjects had been initially evaluated as donors for liver transplantation. None of the individuals evaluated as transplantation donors had cirrhosis. However, 5 of 16 had bridging fibrosis. These 16 subjects were part of 105 subjects evaluated as donors. The liver sonograms were abnormal in 12 of these 16 cases and showed a heterogeneous echotexture in 10 of 16 and hyperechogenecity of the liver in 8 of 16 individuals. Overall, 4 of 5 instances of bridging fibrosis in this group were seen in those with an abnormal sonogram. On the other hand, 3 of 21 subjects undergoing liver biopsy for hepatomegaly had cirrhosis, and another 5 of 21 had bridging fibrosis. Thus, about 40% of patients with persistent hepatomegaly had evidence of advanced fibrosis. There was a high prevalence (6 of 14) of advanced fibrosis within the subset of subjects with hepatomegaly and diabetes.
Fig. 1. The mean (⫾SD) grade of the individual histologic findings of NAFLD in 51 subjects with NAFLD and normal ALT (■) and 50 subjects with NAFLD and elevated ALT levels ( ). There was no significant difference with respect to the grade of steatosis, cytologic ballooning, Mallory bodies, pericellular fibrosis, or portal fibrosis. These data indicate that the histologic spectrum of NAFLD in those with normal ALT are comparable with those with elevated ALT.
Fig. 2. The relationship between the fibrosis stage and the ALT value within the normal range is shown in 51 subjects with NAFLD and normal ALT levels. Fibrosis was scored as follows: 0 ⫽ none, 1 ⫽ either pericentral/pericellular or focal portal fibrosis, 3 ⫽ bridging fibrosis, 4 ⫽ cirrhosis. Five of 15 subjects with NAFLD and low normal ALT levels (⬍30 IU/L) had advanced fibrosis compared with 13 of 36 individuals with high normal ALT levels (31-75 IU/L) (P ⬍ 1, Fisher exact test).
HEPATOLOGY, Vol. 37, No. 6, 2003
MOFRAD ET AL.
Table 6. Comparison of Those With Low Normal Versus High Normal ALT Values
Parameter
Age (y), mean ⫾ SD Gender (M:F) Ethnicity (W:A:H) Weight (lb), mean ⫾ SD Diabetes (n) Hypertension (n) Triglycerides (mg/dL), mean ⫾ SD Histology, mean ⫾ SD Steatosis (grade) Cytologic ballooning (grade) Pericellular fibrosis (grade) Overall fibrosis (stage)
ALT (<30 IU/L) (n ⴝ 15)
ALT (31-75 IU/L) (n ⴝ 36)
P Value
51.8 ⫾ 5 3:12 12:3:0 206 ⫾ 34 11 8 175 ⫾ 177
54.1 ⫾ 7.5 13:23 31:3:2 212 ⫾ 31 18 16 181 ⫾ 96
NS NS NS NS NS NS NS
1.6 ⫾ 0.6 0.78 ⫾ 0.7 0.35 ⫾ 0.6 1.5 ⫾ 1.5
2.16 ⫾ 0.8 1.05 ⫾ 0.7 0.9 ⫾ 0.8 1.37 ⫾ 1.5
.04 NS .049 NS
NOTE. Numerical data were compared using unpaired t test. Proportions were compared using Fisher exact test and mean histologic grades compared by Mann Whitney test. Abbreviations: W, white; A, African American; H, Hispanic.
Fifteen of 51 subjects had ALT values less than 30 IU/L. Although the age and gender distribution in these subjects was comparable with those with ALT values between 30 and 75 IU/L (Table 6), those with low normal ALTs had a significantly lower grade of steatosis and pericellular fibrosis. Eleven of 15 subjects with low normal ALT levels were diabetic compared with 18 of 36 subjects with high normal ALT levels (P ⫽ .2 by Fisher exact test), whereas 5 of 15 subjects had advanced fibrosis compared with 13 of 36 subjects with high normal ALT values (P ⫽ 1, Fisher exact test).
Discussion The serum ALT value has long been used as a surrogate marker of liver injury.14,15 It is, however, well known that the ALT values do not correlate well with the severity of liver disease noted on liver biopsy in subjects with chronic liver disease.15 This is particularly true when one considers the stage of fibrosis present in an individual patient. The present study confirms this to be true also for NAFLD. The upper limit of normal for ALT for a given laboratory depends both on the methodology used as well as on the range of values established for apparently normal individuals in the region. The method used in our institutional laboratory is the commercially available Vitros ALT slide (Johnson & Johnson), which is a dry, multilayered analytical element coated on a polyester support. An 11-L drop of serum is placed on the slide and is evenly distributed from the spreading layer to the underlying layers. Alanine and ␣-ketoglutarate in the spreading layer are converted by the ALT in the serum to pyruvate and
1291
glutamate. Pyruvate is then converted to lactate with the conversion of NADH to NAD. The rate of oxidation of NADH is measured by reflectance spectrophotometry, and the rate of change in reflection density is converted to enzyme activity. When pyridoxal phosphate is used as a cofactor for the initial reaction, the resulting ALT values are about 15 to 20 units higher than in its absence, producing normal ALT levels up to 72 IU/L as noted in the manufacturers guide. The methods are rigorously tested for quality control in accordance with the manufacturer’s guidelines and the College of American Pathologists. These values were further modified to provide an upper limit of 75 IU/L based on the mean and standard deviations of values obtained from 2 cohorts of apparently normal individuals about 15 years ago. Thus, although the range of normal ALT values appears to be higher than those used by many laboratories, this is mostly due to the methods used in our laboratory. Another potential explanation of the finding of significant pathology in the face of normal ALT values is that the supposedly normal range of values is incorrect. The upper limits of normal for ALT values were established over 40 years ago and represent the mean ⫾ SD of the ALT values of a group of individuals felt to be free of liver disease.14 For a given laboratory, the upper limits of normal are also based on ALT values in apparently normal individuals in that area. This has recently been questioned, and, in a large, recent study,9 it was shown that many individuals with supposedly normal ALT levels had biochemical evidence of the metabolic syndrome, the major risk factor for NAFLD. Also, ALT values correlated directly with the body mass index, another risk factor for NAFLD. Based on these data, it was suggested that the upper limits of normal for ALT be decreased to 30 IU/L for men and 19 IU/L for women. In our study, 15 of 51 subjects had an ALT less than 30 IU/L. Although the degree of steatosis and pericellular fibrosis was less in this subset of subjects, there were no significant differences in the prevalence of advanced fibrosis between those with a low normal ALT (⬍30 IU/L) and high normal ALT (31-75 IU/L). Also, all stages of NAFLD were present in those with normal ALT values. Thus, although the public health and clinical utility of reducing the upper limits of normal for ALT continues to be debated,16 the present study provides additional data that, even at low ALT levels, significant liver disease can be present. It is interesting to note the relatively high prevalence of bridging fibrosis or cirrhosis in those with NAFLD and normal ALTs. This may be at least partly due to the bias inherent in a retrospective analysis. The differences in the rates of advanced fibrosis among those undergoing biopsy for hepatomegaly versus asymptomatic and apparently
1292
MOFRAD ET AL.
healthy individuals being evaluated as transplant donors indeed support this possibility. On the other hand, the data from the transplant donor group also indicate that the absence of any obvious symptoms, signs, or ALT values suggestive of liver disease does not guarantee the absence of clinically significant liver disease, i.e., advanced stage chronic liver disease. In this respect, these data corroborate and extend the existing literature regarding other chronic liver diseases, e.g., hepatitis C and normal ALT values.17 An important question is the following: How does one identify subjects with clinically significant NAFLD in the absence of symptoms, signs, or abnormal ALT values? It is clearly both unreasonable and impractical to perform routine liver biopsies for this purpose. Although much more work needs to be done to determine who should be screened for NAFLD and how such individuals should be evaluated, the current studies suggest that those with hepatomegaly, especially in the presence of diabetes, are more likely to have clinically significant underlying pathology. In summary, patients with NAFLD and normal ALT values may have steatohepatitis and advanced fibrosis. Diabetes is an independent predictor of advanced fibrosis. Although further work remains to be done to answer the clinical questions raised by these data, this study is an initial step toward the identification, characterization, and management of NAFLD in patients with normal ALTs.
References 1. Bellentani S, Saccoccio G, Masutti F, Croce LS, Brandi G, Sasso F, Cristanini G, et al. Prevalence of and risk factors for hepatic steatosis in Northern Italy. Ann Intern Med 2000;132:112-117.
HEPATOLOGY, June 2003
2. Wanless IR, Lentz JS. Fatty liver hepatitis (steatohepatitis) and obesity: an autopsy study with analysis of risk factors. HEPATOLOGY 1990;11:74-80. 3. Ground KEV. Prevalence of fatty liver in healthy adults accidentally killed. Aviat Space Environ Med 1984;55:59-61. 4. Ludwig J, Viggiano TR, McGill DB, Oh BJ. Nonalcoholic steatohepatitis: Mayo Clinic experiences with a hitherto unnamed disease. Mayo Clin Proc 1980;55:434-438. 5. Bacon BR, Farahvash MJ, Janney CG, Neuschwander Tetri BA. Nonalcoholic steatohepatitis: an expanded clinical entity. Gastroenterology 1994;107:1103-1109. 6. Diehl AM, Goodman Z, Ishak KG. Alcoholic liver disease in nonalcoholics. A clinical and histologic comparison with alcohol-induced liver injury. Gastroenterology 1988;95:1056-1062. 7. Powell EE, Cooksley WG, Hanson R, Searle J, Halliday JW, Powell LW. The natural history of nonalcoholic steatohepatitis: a follow-up study of forty-two patients for up to 21 years. HEPATOLOGY 1990;11:74-80. 8. Matteoni CA, Younossi ZM, Gramlich T, Boparai N, Liu YC, McCullough AJ. Nonalcoholic fatty liver disease: a spectrum of clinical and pathological severity. Gastroenterology 1999;116:1413-1419. 9. Prati D, Taioli E, Zanella A, Della TE, Butelli S, Del Vecchio E, Vianello L, et al. Updated definitions of healthy ranges for serum alanine aminotransferase levels. Ann Intern Med 2002;137:1-10. 10. Contos MJ, Cales W, Luketic VA, Sterling RK, Shiffman ML, Mills AS, Fisher RA, et al. Development of nonalcoholic fatty liver disease after liver transplant for cryptogenic cirrhosis. Liver Transpl 2001;7:363-373. 11. Brunt EM, Janney CG, Di Bisceglie AM, Neuschwander-Tetri BA, Bacon BR. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol 1999;94:2467-2474. 12. National Institutes of Health. Third report of the national cholesterol education program expert panel on detection, evaluation and treatment of high blood cholesterol in adults (Adult treatment panel III). NIH Publication 2001:1-3670. 13. Anonymous. Executive summary of the clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. Arch Intern Med 1998;158:1855-1867. 14. Karmen A, Wroblewski F, LaDue JS. Transaminase activity in human blood. J Clin Invest 1955;34:126-133. 15. Kallei L, Hahn A, Roder VZ. Correlation between histological findings and serum transaminase values in chronic diseases of the liver. Acta Med Scand 1964;175:49-56. 16. Kaplan MM. Alanine aminotransferase levels: what’s normal? Ann Intern Med 2002;137:49-51. 17. Alberti A, Morsica G, Chemello L. Hepatitis C viraemia and liver disease in symptom-free individuals with anti-HCV. Lancet 1992;340:697.