Digestive Diseases and Sciences

, Volume 52, Issue 10, pp 2512–2519

Orlistat Reverse Fatty Infiltration and Improves Hepatic Fibrosis in Obese Patients with Nonalcoholic Steatohepatitis (NASH)

Authors

  • Osamah Hussein
    • Department of Internal Medicine ASieff Government Hospital
  • Masha Grosovski
    • Liver UnitSieff Government Hospital
  • Sorina Schlesinger
    • Department of Internal Medicine ASieff Government Hospital
  • Sergio Szvalb
    • Department of PathologySieff Government Hospital
    • Department of Internal Medicine ASieff Government Hospital
    • Liver UnitSieff Government Hospital
    • Technion Institute of TechnologyFaculty of Medicine
    • Liver Clinic and Internal Medicine ASieff Government Hospital, P.O.H 1008
Original Article

DOI: 10.1007/s10620-006-9631-1

Cite this article as:
Hussein, O., Grosovski, M., Schlesinger, S. et al. Dig Dis Sci (2007) 52: 2512. doi:10.1007/s10620-006-9631-1

Abstract

Nonalcoholic steatohepatitis (NASH) may cause progressive hepatic fibrosis, cirrhosis, and hepatocellular carcinoma. Treatment, thus far, has been restricted to diet and weight loss, but without compelling results. In this study we aimed to evaluate the efficacy of orlistat therapy in obese patients with NASH. Fourteen obese patients with NASH underwent liver biopsy prior to and subsequent to 6 months treatment with orlistat (120 mg tid). Hepatic fat extension was graded as normal, mild, moderate, or severe. Hepatic fibrosis was scored on a scale from 0 to 4, with 0 denoting no fibrosis and 4, cirrhosis. Portal inflammation was scored as 0–3, with 0 = normal, 1 = mild, 2 = moderate, and 3 = severe inflammation. Fourteen patients had NASH associated with diabetes, hyperlipidemia, or obesity. Orlistat reduced fatty infiltration in 10 patients (70%; P<0.01), 3 of whom had normal liver fat content after treatment. Orlistat improved inflammatory activity by 2 grades in 28% and by 1 grade in 50% of patients and effected no change in 22% of patients. Five patients (35%) returned to normal inflammatory activity. Orlistat improved hepatic fibrosis by 2 grades in three patients (21%) and by 1 grade in seven patients (50%). There was no change in four patients (28%). Orlistat lowered aminotransferases levels, total cholesterol, triglycerides and low-density lipoprotein, respectively. Insulin resistance index and malonyl dialdehyde levels improved significantly after orlistat therapy, whereas HbAıc remained unchanged. In conclusion, in obese patients with NASH, liver fibrosis and inflammation improved after therapy with orlistat.

Keywords

OrlistatFatty liverNonalcoholic steatohepatitis (NASH)Liver biopsyMetabolic syndromeInsulin resistanceLipid peroxidationObesity

Background

Normally, less than 5% of the liver mass is fat by weight, but in patients with nonalcoholic steatohepatitis (NASH), as much as 50–80% of liver weight may be made up of fat, mostly in the form of triglycerides[1]. The clinical implications of NASH are derived mostly from its common occurrence in the general population (10–24%) and its potential to progress to fibrosis (40%), cirrhosis (30%), and hepatocellular carcinoma [2, 3]. NASH is the most common cause of cryptogenic cirrhosis and is an increasingly common indication for liver transplantation [3].

Obesity, diabetes mellitus type 2, and hyperlipidemia are conditions frequently associated with NASH [4]. The prevalence of NASH increases by a factor of 4.6 in obese people (body mass index [BMI], >30) [5]. Truncal obesity is an important risk factor for NASH, even in patients with a normal BMI [6].

Obesity is heterogeneous with regard to both etiology and associated metabolic risks. However, it is recognized that dietary fat is a major causative factor in the development of obesity [7]. Moreover, a high intake of dietary fat is associated with an adverse lipid profile and there is evidence to suggest that fat intake can promote insulin resistance [8, 9]. Thus, targeting dietary fat is a logical strategy for reducing obesity and lowering the risk of fatty infiltration in the liver.

Currently, standard practice is to advocate weight loss and exercise [10]. However, while dietary modification may enable patients to attain short-term weight loss, it is recognized that very few patients are able to achieve long-term maintenance of lower body weight and improvement in fatty liver risk factors through lifestyle modification alone [10]. Moreover, previous studies on the effect of weight loss in NASH has yielded mixed results; starvation or rapid weight loss may improve liver enzymes and steatosis but worsen inflammation and fibrosis [11, 12]. Steady and consistent weight loss within 6 months may improve fatty liver infiltration . Additionally, attempts to treat patients with NASH including ursodeoxycholic acid [13], clofibrate [14], and antioxidants like vitamin E have also failed [15] and definite pharmacological therapy is lacking.

Although the pathogenesis of NASH is unknown, it has been suggested that hepatic fatty infiltration may stem from continuous delivery of free fatty acids to the liver after ingestion of fatty foods and from increased splanchnic lipolysis of visceral fat, both of which increase hepatic insulin resistance [16]. Recently, two other major pathogenic mechanisms were implicated (second hit), i.e., oxidative stress/ lipid peroxidation, and tumor necrosis factor α (TNFα)/endotoxin-mediated injury [17].

Orlistat (Xenical) is a semisynthetic derivative of lipstatin which is a naturally occurring lipase inhibitor produced by streptomyces molds that reduces the absorption of dietary fat by 30% [18]. In clinical trials, orlistat-treated patients had significantly greater weight loss and reduction in total cholesterol, triglycerides, fasting insulin, and fasting glucose than patients treated with dietary modification alone. Moreover, orlistat had an additional direct cholesterol-lowering effect independent of weight reduction. Treatment with orlistat also resulted in a significant reduction in waist circumference (abdominal obesity) and improvement in triglycerides, high-density lipoprotein (HDL), and aoplipoprotein B among abdominally obese patients [19, 20]. Long-term treatment with orlistat is well tolerated by patients, and unlike other drugs used to treat fatty liver, it is not associated with adverse hepatic effect [18]. The current study shows that orlistat reverses fatty liver disease and improves fibrosis while decreasing body weight and reducing the absorption of dietary fat. To the best of our knowledge, this is the first study with paired liver biopsies prior to and after treatment.

Methods

Patient population

Twenty-two patients were recruited and 16 patients were enrolled in the study after undergoing clinical evaluation including complete medical history, physical examination, and baseline laboratory studies. The study was approved by the hospital internal review board, and all patients signed informed consent.

Inclusion criteria included liver ultrasound and liver histology compatible with the diagnosis of NASH based on steatosis and two of the following three histological features: (1) necroinflammatory foci with mononuclear cells and/or neutrophils; (2) ballooning degeneration of hepatocytes with or without Mallory's bodies; and (3) pericellular fibrosis [21], persistent elevation of aminotransferases to 1.5 times normal for >6 months, and exclusion of other liver diseases (i.e., alcoholic, autoimmune, cholestatic, viral, or metabolic liver diseases).

Exclusion criteria included uncontrolled diabetes mellitus and alcohol use of more than 30 g/day, as reported by the patient and confirmed by a family member. Treatment with ursodeoxycholic acid, thiazolidinediones, metformin, fibrate, estrogens, tamoxifen or calcium blockers, vitamin E, multivitamin complex including iron, or any medications known to cause steatosis, such as amiodarone, methotrexate, and glucocorticoids, within the previous 6 months was also an exclusion criterion. Pregnant or lactating women or those of childbearing potential with inadequate contraceptive measures were not included. Patients with psychiatric or neurologic disorders, a history of recurrent nephrolithiasis or symptomatic cholelithiasis, previous gastrointestinal tract surgery for weight reduction, a history or the presence of malignancy, a history of cardiovascular disease (e.g., stroke, ischemic heart disease, and congestive heart failure), or renal impairment with a plasma creatinine level >1.4 mg/dl were excluded as well.

Patients were provided a moderate calorie-restricted diet, with a median daily intake of 25 kcal/kg body weight/day, with 30% of the calories derived from fats, 50% from carbohydrates, and 20% from protein.

Standard biochemical tests were performed, as were viral hepatitis serology, autoimmune markers, ferritin, ceruloplasmin, α-1 antitrypsin, fasting glucose and insulin, triglycerides, total cholesterol, low-density lipoprotein (LDL), and HDL. An ultrasound was performed to assess the presence of parenchymal pathology, vascular patency, or ascites and to exclude hepatocellular carcinoma. All patients had normal serum creatinine and normal thyroid function test. Insulin resistance was estimated using the homeostasis model assessment (HOMA) index derived from the following equation:

HOMA index = [(fasting plasma glucose level, mg%, × 0.055) × (fasting plasma insulin level, μIU/ml)]/22.5 [22]. BMI was calculated as weight in kilograms divided by the square of height in meters. LDL was separated from plasma by discontinuous gradient ultracentrifugation at 4°C [23]. LDL oxidation was analyzed by the thiobarbituric acid reactive substance assay, which measures malondialdehyde (MDA) equivalents [24].

Drug therapy and follow-up

All patients received orlistat, 120 mg tid, and were followed up with liver enzyme tests at 2-month intervals during 6 months of treatment. At each visit, adverse events, concurrent illnesses, concomitant medication, and compliance with the study medication were assessed. Monitoring laboratory tests including alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase, bilirubin, γ-glutamyl transferase (GGT), prothrombin time (PT), and albumin were performed at 1, 3, and 6 months during the study. A complete medical history was repeated at the end of treatment. The patients underwent a second liver biopsy after 6 months of treatment.

Liver histology

A liver biopsy was performed at entry and repeated after 6 months of treatment with orlistat. Specimens were fixed in buffered formalin and embedded in paraffin. Hematoxylin and eosin (H&E) sections, periodic acid Schiff with diastase stain for inflammatory grading, and Mason's trichrome and reticulin stains were reviewed for fibrosis and architectural changes. The necroinflammatory grade and stage of fibrosis were assessed according to the method of Brunt et al. [25]. Steatosis was graded as follows: mild, 5–30% of hepatocytes affected; moderate, 30–60%; and severe, >60%. Fibrosis was scored on a scale from 0 to 4, with 0 denoting no fibrosis, 1 describing focal pericellular fibrosis in zone 3, 2 describing perivenular and pericellular fibrosis confined to zones 2 and 3, with or without portal/periportal fibrosis, 3 describing bridging fibrosis with architectural distortion but no cirrhosis, and 4 determining cirrhosis. Inflammatory grade was scored on a scale from 0 to 4, with grade 0 = no inflammation, grade 1 = sparse zone 3 inflammation, grade 2 = mild focal zone 3 hepatocyte inflammation, grade 3 = moderate zone 3 hepatocyte inflammation, and grade 4 = severe zone 3 hepatocyte inflammation [25]. The specimens were also examined for histological features of Mallory's bodies, ballooning degeneration, acidophilic necrosis, sinusoidal fibrosis, and polymorphonuclear infiltrates. Each biopsy was analyzed and graded by the same pathologist, blinded to the patients’ clinical and laboratory findings. All patients had a dietary consult during the study period. Compliance with the study medication was monitored by counting the number of the capsules returned at each visit.

Statistical analysis

Demographic, clinical, and pathological features are categorized as continuous or categorical variables. Comparisons between variables at baseline and at the end of the study were done by chi-square test for categorical variables and by Wilcoxon signed-rank test for continuous variables. The extent of fat expressed as grade (mild, moderate, or severe or as scored) was assessed by Kendall's τ correlation of ordinal variables. A P value <0.05 was considered significant.

Results

Sixteen patients (10 women and 6 men), with a mean age of 42±3 years (median, 44 years; range, 26–56 years) and a mean BMI of 31.5 kg·m2 (median, 31; range, 29–37), were enrolled in the study. A total of 32 biopsies were performed. Four of the 32 (2 patients) contained insufficient material. The mean size of 28 liver biopsy specimens (14 patients) was 0.9±0.6 cm. Six patients experienced adverse events during treatment with orlistat including nausea, abdominal cramps, loose stools, and body odor. There were no adverse events noted in hematology or blood chemistry results.

Table 1 summarizes the demographic features and metabolic findings of patient population. All enrolled patients were obese but none considered to be morbidly obese (BMI, >40 kg/m2). Four patients had mixed hyperlipidemia, five patients had diabetes mellitus type 2, and five patients had obesity alone (BMI, 30–37 kg/m2). Five of nine obese and hyperlipidemic patients had at least three components of metabolic syndrome (triglycerides, >150 mg/dl; HDL-C, <40 mg/dl for men and <50 mg/dl for women; blood pressure, >130/85 mm Hg; fasting glucose, >110 mg/dl; waist circumference, >102 cm for men and >88 cm for women).
Table 1

Clinical and biochemical characteristics of 14 enrolled obese patients with abnormal liver enzyme (alanine aminotransferase, ALT)

Age

Gender

Weight (kg) before treatment

Weight (kg) after treatment

Preexisting condition

ALT (IU/L)

56

F

68

66

Hyperlipidemia

41

42

F

89

85

Obesity

55

50

F

83

76

Obesity

172

50

F

126

123

Hyperlipidemia

83

33

M

110

101

Hyperlipidemia

50

47

F

87

85

Diabetes

40

29

M

108

102

Diabetes

124

32

F

100

103

Diabetes

100

26

M

104

87

Obesity

110

43

F

91

88

Obesity

107

44

F

82

76

Diabetes

107

36

M

96

90

Obesity

79

53

F

73

70

Hyperlipidemia

92

52

F

101

94

Diabetes

46

Table 2 and Fig. 1 summarize baseline histological features and changes after 6 months of treatment with orlistat. Sixty-four percent (9/14) of patients had severe steatosis. Regression of fatty infiltration occurred in 10 of 14 patients (71%; P<0.01). Four female patients with severe steatosis showed no change in the extent of steatosis but experienced improvement in activity grade and stage of fibrosis. Three of the four were diabetic. None lost more than 10% of body weight during the study period. Seven of 14 patients (50%) improved their fibrosis extension by 1 grade, 3 of 14 patients improved fibrosis by 2 grades (21%), and in 3 of 14 patients the extent of fibrosis did not change. Twenty-eight percent (n=14) showed improved inflammatory activity by 2 grades, 50% (n=7) by 1 grade, and 22% (n=3) demonstrated no change. Compared to baseline, none of the patients had worsening of liver histology at the end of treatment period with the exception of one obese nondiabetic patient who progressed from stage 2 fibrosis to stage 3 fibrosis despite a weight loss of 7 kg.
Table 2

Baseline histological features and changes after 6 months of treatment with orlistat

Patient no.

Fat extension before

Fat extension after

Inflammation before

Inflammation after

Fibrosis before

Fibrosis after

Weight change (%)

1

Severe

Severe

A2

A0

F2

F0

−2.9

2

Severe

Mild

A2

A1

F2

F1

−4.5

3

Mild

Normal

A4

A2

F2

F3

−8.4

4

Mild

Normal

A2

A2

F1

F1

−2.4

5

Severe

Mild

A2

A0

F2

F0

−8.2

6

Severe

Severe

A2

A1

F2

F1

−2.3

7

Moderate

Mild

A1

A1

F1

F0

−5.5

8

Severe

Severe

A2

A1

F2

F1

+3

9

Severe

Mild

A2

A1

F1

F0

−16.3

10

Severe

Moderate

A2

A2

F2

F1

−3.3

11

Severe

Severe

A3

A2

F3

F1

−7.3

12

Mild

Normal

A1

A0

F1

F1

−6.3

13

Moderate

Mild

A1

A0

F1

F0

−4.1

14

Severe

Mild

A2

A0

F2

F2

−6.9

Note. Fat extension: normal, <10%; mild, <30%; moderate, 30–60%; severe, >60%. Fibrosis scale: 0 = none, F1= portal, F2 = portal +septa, F3 = bridging, F4 = cirrhosis. Portal and periportal inflammatory activity (A): 0=normal, 1=mild, 2=moderate, 3=severe.

The mean weight decreased from 94.1±15.5 kg (median, 93.5 kg) pretreatment to 89±15.1 kg (median, 87.0 kg) posttreatment (−5.3% change). One patient did not lose weight. Of the patients who lost weight, the percentage body weight lost ranged from −2.3% to −16.3% (median, −5%; mean, 5.4%±4.3%; Table 2). Seven patients lost at least 6% of their body weight. The weight loss correlated only with the extension of fat (r=0.5), and not with the changes in fibrosis or in inflammation.

A significant improvement in serum levels of ALT, AST, alkaline phosphatase, and GGT occurred at the end of treatment with orlistat (ALT, 84±10 vs 43±5 IU/L, P<0.0004; AST, 72±11 vs 32±4 IU/L, P<0.0004; alkaline phosphatase, 115±10 vs 102±7 IU/L, P<0.02; GGT, 100±21 vs 55±6 IU/L, P <0.01; Fig. 2). Similarly, a significant improvement in serum lipid profile was also noted at the end of treatment (total cholesterol, 229±12 vs 194±13 mg/dl, P<0.001; triglycerides, 238±23 vs 163±10 mg/dl, P<0.001; and LDL, 143±11 vs 120±5 mg/dl, P<0.003). Serum levels of HDL, bilirubin, albumin, and prothrombin remained unchanged in all patients. Insulin resistance index and MDA levels improved significantly after orlistat therapy while HbAıc remained unchanged respectively (HOMA index: normal, 0.8–5.4; 6.5±2.5 vs 3.3±1.2, P<0.05; MDA: normal, <0.3 nmol/ml; 0.42±0.03 vs 0.37±0.02 nmol/ml, P<0.01; and HbAıc: normal, 3.8–6.4%; 7.1%±3.1% vs 7.5%±2.6%, P=NS) .
https://static-content.springer.com/image/art%3A10.1007%2Fs10620-006-9631-1/MediaObjects/10620_2006_9631_Fig1_HTML.jpg
Fig. 1

Histologic features of liver biopsy specimens obtained before and after orlistat therapy in four patients with nonalcoholic steatohepatitis. (A) A specimen obtained from patient 1, with severe fatty infiltration. (B) A specimen obtained from the same patient 6 months later, with a significant decrease in the extent of fat. (C) A specimen obtained from patient 5; the portal tract is enlarged by lymphocytic infiltrate and mild zone 3 inflammation. (D) A specimen obtained from the same patient 6 months later, with no inflammation. (E) A specimen obtained from patient 10; the portal tract is enlarged by fibrous expansion with bridging fibrosis and pericellular fibrosis. (F) A specimen from the same patient 6 months later, with only minimal pericellular fibrosis. (G) A specimen obtained from patient 11; the extent of fat is severe and the fibrosis is moderate, with perivenular and pericellular fibrosis in zones 2 and 3. (H) A specimen from the same patient 6 months later, with significant improvement of fat extension and regression of fibrosis

https://static-content.springer.com/image/art%3A10.1007%2Fs10620-006-9631-1/MediaObjects/10620_2006_9631_Fig2_HTML.gif
Fig. 2

Changes in aminotransferase levels before and after treatment with orlistat

Discussion

The results of this study clearly indicate that 6 months of orlistat therapy reduced hepatic fatty infiltration, inflammation, and liver fibrosis in 71%, 71%, and 78% of NASH patients, respectively. The results also demonstrate improved liver enzymes, body weight, lipid profile, insulin resistance index, and oxidative stress in the majority of patients.

NASH is a common condition that progresses to cirrhosis in up to 30% of patients [3]. The major pathogenic mechanism of NASH is the increase in delivery of free fatty acids to the liver from dietary sources and from splanchnic lipolysis [7, 8]. Presently, gradual and sustained weight loss is the only known therapy that improves steatosis [26].

Advanced hepatic fibrosis is usually considered to be an irreversible process, even when the underlying cause is treated or removed [27]. However, the fibrosis can be ameliorated, e.g., in patients treated with ursodiol or methotraxate for primary biliary cirrhosis [28, 29]. Regression of hepatic fibrosis in patients with chronic hepatitis C has been observed after treatment with interferon-α [30]. We cannot rule out a synergestic effect of weight loss on the regression of liver fibrosis. The presence of steatosis may contribute to necroinflammatory injury with the development of fibrosis in NASH patients. Thus, regression of steatosis may improve or arrest inflammation and the degree of fibrosis resulting from long-term treatment with orlistat [31] regardless of weight loss.

The mechanism underlying the effect of orlistat remains unknown. Orlistat therapy successfully lowers body weight, but there may be additional mechanisms by which orlistat reverses fatty infiltration and improves hepatic fibrosis. In the present study, four patients with severe steatosis showed no change in the extent of hepatic steatosis but, nevertheless, demonstrated an improvement in activity grade and fibrosis stage, suggesting that orlistat may mediate its effect by improving insulin resistance and lipid peroxidation regardless of the decreased fat content in the liver.

In addition to the role of total dietary fat in the current obesity epidemic, there is the distinct possibility that different dietary fats may have different effects on cardiovascular risk factors, body overweight, fatty infiltration of liver, and insulin resistance [8].

Orlistat reduces the absorption of dietary fat up to a level of 30% [19] and may modulate insulin action by changing not only the amount of fat delivered to the liver, but also the type of fat. Saturated fatty acids increase insulin resistance, while unsaturated fat, particularly monounsaturated fat, improves insulin sensitivity [9]. In this study, the concentrations and types of free fatty acids before and after orlistat therapy in the serum or in the liver were not measured.

Patients with hypertriglyceridemia have elevated TNF-α, which induces systemic inflammation as represented by elevated C-reactive protein (CRP) and fibrinogen [32]. Whether orlistat-decreased levels of postprandial triglycerides lower the inflammatory component associated with NASH remains to be determined.

Various treatment options have been attempted in NASH patients. Antioxidants have been tested in pilot studies, without compelling results [15]. In a trial of clofibrate, a fibrate, no effect was demonstrated on aminotransferases or histology [13]. Gemfibrozil improved serum aminotransferases as well as serum lipids after 1 month of therapy as demonstrated by blood tests, however, no liver biopsy was performed [14]. Troglitazone has been shown to normalize liver enzymes in a small number of patients, but with only mild histological improvement [33]. Ursodeoxycholic acid leads to cosmetic improvement in liver enzymes, but no documented effect on histology has been shown [13]. Recently, treatment of seven patients with betaine for 1 year showed a marked improvement in the degree of steatosis, necroinflammatory activity, and stage of fibrosis [34].

Orlistat is not absorbed and therefore has no systemic effects associated with other anti-obesity treatments [19]. Adverse reactions are minimal and include nausea, gastrointestinal distress, and fecal incontinence.

Without a placebo group, it is uncertain whether orlistat has a direct effect on steatohepatitis independent of weight reduction. The improvement in insulin resistance by orlistat was mainly due to a reduction in fasting plasma glucose levels in diabetic patients and in fasting insulin levels in nondiabetic subjects [35]. These results suggest that the attenuation in insulin resistance after weight reduction may have different mechanisms in obese individuals with or without diabetes. Plasma values of lipid peroxidation markers (MDA) may reflect the process of lipid peroxidation occurring in the liver [36]. However, it could be an expression of an enhanced production of free radicals in the circulation as documented in obese and diabetic subjects [37]. In our patients, MDA levels were reduced with orlistat therapy, which may be due to the lowering of plasma LDL levels, which may increase cellular LDL receptor synthesis. This will shorten the residence time of LDL in the plasma, and thus reduce its exposure to free radicals, preserving LDL antioxidant content, which makes it more resistant to oxidation.

One limitation of our clinical study is the absence of a placebo group. Thus, all the changes observed could be nonspecifc and do not necessarily reflect any unique property of orlistat, and trials comparing orlistat-induced weight loss against the effect of similar weight loss induced by dieting or exercise are needed before any claim that orlistat has a specific benefit can be substantiated.

In conclusion, the weight loss induced by orlistat reverses the fatty infiltration and fibrosis of the liver in a small group of obese patients with NASH. These findings hold promise for a better therapy in the future. In light of the worldwide obesity epidemic, it is our belief that a larger multicenter study would be valuable.

Copyright information

© Springer Science+Business Media, LLC 2006