the natural history of non-alcoholic fatty liver …...the natural history of non-alcoholic fatty...
TRANSCRIPT
The Natural History of Non-Alcoholic Fatty Liver Disease
Leon A. Adams M.B.B.S, F.RA.C.P.
This thesis is presented for the degree of Doctor of Philosophy
School of Medicine and Pharmacology
The University of Western Australia
2006
ii
DECLARATION
This thesis is the work ofthe author unless otherwise specified.
iii
ACKNOWLEDGEMENTS
I would like to express my gratitude to my supervisors for mentoring me through
this thesis. Particularly I would like to express my thanks to A/Prof. Paul Angulo
for the tremendous opportunity to work with him, during which I profited from
his perceptive advice, patience and friendship. I would also like to thank Prof.
John Olynyk for his support and generosity which have been unqualified, as well
as A/Prof. Gary Jeffrey for his encouragement and help in providing me with the
persistence and focus to complete this.
There have been many other people to thank who have provided support, advice
and friendship; Jenny St. Sauvier, Jim Lymp, Sky Sanderson, Keith Lindor,
Jayant Talwalkar, Steve Brown, Scott Harmsen, Phunchai Charatcharoenwitthay,
Felicity Enders, Jill Keach, Dawn Schultz, Jessica Donlinger and Gwen Boe.
Thanks also to Mum and Dad for giving me an inquiring mind, inspiring me and
spending many a long hour proof-reading and thinking about this thesis. Lastly,
thankyou to Louisa, whom without her, none of this would ever have been done.
iv
ABSTRACT
Nonalcoholic fatty liver disease (NAFLD) is a potentially serious condition
present in 20-30% ofthe general population. Insulin resistance is an important
pathogenic factor for NAFLD. Subsequently, as the prevalence of conditions
associated with insulin resistance such as diabetes and obesity increase in the
community, the prevalence of NAFLD is also likely to be increasing. Thus
NAFLD is becoming a frequently encountered clinical scenario for physicians and
general practitioners. Knowledge ofthe natural history of NAFLD is vital to
guide clinical management decisions regarding investigation, monitoring and
treatment as well as for patient counselling. At a population level, knowledge of
the natural history guides resource allocation and public health policy on
prevention and treatment.
It is recognized that some patients with NAFLD progress to cirrhosis which
may be complicated by hepatocellular carcinoma, liver failure and death.
However, previous studies examining the natural history of NAFLD have had
significant methodological problems including selection biases, limited numbers
and short follow-up periods. Therefore the natural history is not clearly defined,
with the magnitude and rate of disease morbidity and mortality unclear and
prognostic factors related to these outcomes unknown.
This thesis examined the natural history of NAFLD in a number of cohorts
generated from the Rochester Epidemiology Project and the diagnostic index of
the Mayo Clinic. Firstly, the clinical evolution to cirrhosis, liver related death and
V
overall death was examined in a community based cohort. Secondly, the
histological evolution of subclinical NAFLD was detailed in a cohort of patients
with serial liver biopsies. The third and fourth experimental chapters examined
the clinical outcomes of subjects with NAFLD and diseases associated with
insulin resistance, namely hypopituitarism/hypothalamic disease and type 2
diabetes mellitus.
To evaluate the clinical evolution of NAFLD, 420 community based subjects
were followed for a median duration of 7.1 years (range 0.1-23.5). Death
occurred in 12.6% ofthe cohort and 1.7% died fro'm liver related causes. NAFLD
was associated with an increased risk of death (standardized mortality ratio 1.34
(95% CI 1.003-1.76], p=0.03) in comparison to the general population. Liver
disease accounted for 13% of all deaths among NAFLD patients in contrast to less
than one percent of all deaths in the general population. Impaired fasting
glycemia and high body mass index (BMI) were risk factors for overall and liver-
related death. The occurrence of cirrhosis and its complications was relatively
low, at 5% and 3.1% respectively.
The histological evolution of NAFLD was examined in a cohort of 103 subjects
who had undergone serial liver biopsies a mean of 3.2 years (range 0.7-21) apart.
Progressive fibrosis was observed in 37% of subjects with nine percent becoming
cirrhotic. Fibrosis progressed slowly at an average rate 0.02 + 0.66 stages/year or
at 0.09 + 0.67 stages/year when cirrhotics were excluded. Diabetes mellitus and
BMI were significantly associated with fibrosis rate on multivariate analysis. In
contrast to fibrosis, the histological features of steatosis and necro-inflammation
and serum aminotransaminase levels improved over time.
Following the identification of diabetes mellitus and obesity as significant
adverse prognostic factors, the natural history of NAFLD was examined among
other conditions associated with insulin resistance. Firstly, 21 subjects with
hypopituitarism/hypothalamic disease were identified and followed. From time of
diagnosis, patients were observed to develop obesity, glucose intolerance and
dyslipidemia and were subsequently diagnosed with NAFLD. The course ofthe
liver disease in these patients was severe. Cirrhosis developed in 29% ofthe
cohort and liver related death or transplantation occurred in 14% after a median
duration of follow-up of only 6.0 years (range 1.0 to 10 years).
To further examine the natural history of NAFLD among conditions associated
with insulin resistance, a cohort of 337 community-based subjects with type 2
diabetes mellitus were followed for a median of 9.9 years (range 0.1-25.0).
During follow-up, a diagnosis of NAFLD was found in 116 patients and was an
independent risk factor for death (hazard ratio 1.63, 95% confidence interval 1.04-
2.56). Liver related deaths were more frequent among patients with NAFLD
occurring in 19% compared to 0% among those without NAFLD.
In summary, NAFLD is a slowly progressive disease that leads to cirrhosis and
liver related death in a minority of individuals. These sequelae contribute to the
increased risk of overall death present among NAFLD patients in the community
and among NAFLD patients with diabetes mellitus. Therefore, because NAFLD
is common in the general community, the potential disease burden is
considerable. Clinicians should be aware that patients with clinical phenotypes of
insulin resistance, namely obesity, diabetes mellitus and
hypopituitarism/hypothalamic disease are at increased risk of progressive liver
vii
disease and death. These patients should be appropriately counselled, considered
for liver biopsy and targeted for interventions aimed at reducing liver related
morbidity and mortality.
vi ii
PUBLICATIONS
Original papers
1. Adams LA, Keach J, Lindor KD, Angulo P. Nonalcoholic Fatty Liver
Disease Among Patients with Pituitary and Hypothalamic Dysfunction.
Hepatology 2004; 39: 909-914.
2. Adams LA, Sanderson S, Lindor KD, Angulo P. The Histological Course of
Nonalcoholic Fatty Liver Disease: A Longitudinal Study of 103 Patients with
Sequential Liver Biopsies. J Hepatol. 2005; 42: 132-138.
3. Adams LA, Lymp JF, St.Sauver J, Sanderson SO, Lindor KD, Brown LS,
Angulo P. The Natural History of Nonalcoholic Fatty Liver Disease: A
population based cohort study. Gastroenterology 2005; 129: 113-21.
Abstracts
1. Adams LA, Keach J, Lindor KD, Angulo P. Time Course of Fibrosis
Progression in Patients with Nonalcoholic Fatty Liver Disease. Hepatology 2003:
38 (Supp 1) 206A.
2. Adams LA, Feldstein A, Lindor KD, Angulo P. The Spectrum of
Nonalcoholic Fatty Liver Disease in Patients with Hypothalamic and Pituitary
Dysfunction. Hepatology 2003: 38 (Supp 1) 504A.
3. Adams LA, Lymp J, St.Sauver J, Feldstein A, Lindor KD, Brown LS, Angulo
P. The Natural History of Nonalcoholic Fatty Liver Disease: A Population Based
Study. Hepatology 2004:40 (Supp.l); 582A.
IX
4. Adams LA, Harmsen W S , St.Sauver JL, Charatcharoenwitthay P, Enders FT,
Lindor KD, Therneau TM, Angulo P. Nonalcoholic Fatty Liver Disease is a Risk
Factor for Mortality Among Patients with Diabetes: A Population Based Cohort
Study. Hepatology 2005; 42 (Suppl. 1): 614A.
Oral Presentations
1. Adams LA, Keach J, Lindor KD, Angulo P. Time Course of Fibrosis
Progression in Patients with Nonalcoholic Fatty Liver Disease. Presidential
Plenary Session, American Association for the Study of Liver Disease , Boston
2003.
2. Adams LA, Lymp J, St.Sauver J, Feldstein A, Lindor KD, Brown LS, Angulo
P. The Natural History of Nonalcoholic Fatty Liver Disease: A Population Based
Study. Australian Liver Association Workshop. Victor Harbour, South Australia,
April 2005.
3. Adams LA. NAFL or NAFLD? The Natural History of Nonalcoholic Fatty
Liver Disease. Division of Gastroenterology and Hepatology Research
Conference. Mayo Clinic, Minnesota, January 2005.
%
LIST OF ABBREVIATIONS
ALT
AST
BMI
CT
CI
CVD
FFA
GH
HCC
HDL
HICDA
HOMA
HR
IFG
IHD
LRD
MLIS
MRI
MRS
NAFLD
NASH
OR
OC
OSA
PCOS
QUICKI
REP
SMR
SREBP
TNFa
ULN
UNOS
US
WHO
Alanine aminotransferase
Aspartate aminotransferase
Body mass index
Computed tomography
Confidence interval
Cerebrovascular disease
Free fatty acids
Growth hormone
Hepatocellular carcinoma
High density lipoprotein
Hospital adaptation ofthe international classification of disease
Homeostasis model of assessment
Hazard ratio
Impaired fasting glycemia
Ischemic heart disease
Liver related death
M a y o clinic laboratory information system
Magnetic resonance imaging
Magnetic resonance spectroscopy
Nonalcoholic fatty liver disease
Nonalcoholic steatohepatitis
Odds ratio
Olmsted county
Obstructive sleep apnoea
Polycystic ovarian syndrome
Quantitative insulin sensitivity check index
Rochester Epidemiology Project
Standardized mortality ratio
Sterol regulatory element binding protein
Tumour necrosis factor alpha
Upper limit of normal
United network for organ sharing
Ultrasound
World health organization
TABLE OF CONTENTS
Declaration ii
Acknowledgements iii
Abstract iv
Publications viii
List of Abbreviations x
Table of Contents xi
List of Tables xv
List of Figures xvii
CHAPTER ONE - INTRODUCTION AND LITERATURE REVIEW
1.1 Introduction 1.
1.2 Objectives and Aims 3.
1.3 Thesis Outline 3.
1.4 Literature Review 5.
1.4.1 NAFLD Definition and Terminology 5.
1.4.2 Epidemiology 8.
1.4.3 Pathophysiology 12.
1.4.4 Histology 15.
1.4.5 Radiology 19.
1.4.6 Natural History 22.
1.4.6.1 Clinical Progression 22.
1.4.6.2 Histological Progression 26.
1.4.7 NAFLD and Cryptogenic Cirrhosis 30.
1.4.8 NAFLD and Liver Transplantation 30.
1.4.9 NAFLD and Hepatocellular Carcinoma 32.
1.4.10 NAFLD and Metabolic Disease 33.
1.4.10.1 NAFLD and The Metabolic Syndrome 34.
1.4.10.2 NAFLD and other Metabolic Diseases 37.
1.4.10.3 NAFLD and Diabetes Mellitus 38.
1.4.11 Summary 39.
CHAPTER TWO - CLINICAL EVOLUTION OF NAFLD
xii
2.1 Introduction 41.
2.2 Aims 42.
2.3 Methods 43.
2.3.1 Setting 43.
2.3.2 Rochester Epidemiology Project 43.
2.3.3 Case Ascertainment 44.
2.3.4 Patient Information 46.
2.3.5 Statistical Analysis 47.
2.4 Results 49.
2.4.1 Patient Population 49.
2.4.2 Incidence of Diagnosis of N A F L D 52.
2.4.3 Overall Mortality 52.
2.4.4 Predictors of Overall Mortality 57.
2.4.5 Liver-related Morbidity and Mortality 57.
2.4.6 Predictors of Liver-related Mortality 60.
2.4.7 Metabolic Complications 60.
2.4.8 Liver Histology 62.
2.5 Discussion 66.
CHAPTER THREE - HISTOLOGICAL EVOLUTION OF NAFLD
3.1 Introduction 72.
3.2 Aims 73.
3.3 Methods 73.
3.3.1 Case Ascertainment 73.
3.3.2 Patients 75.
3.3.3 Liver Histology 79.
3.3.3 Statistical Analysis 80.
3.4 Results 81.
3.4.1 Patient Characteristics 81.
3.4.2 Fibrosis Progression 85.
xiii
3.4.3 Predictors of Fibrosis Progression 88.
3.4.4 Rate of Fibrosis Progression 90.
3.4.5 Predictors of Rate of Fibrosis Progression 90.
3.4.6 Change in Aminotransaminase Levels 92.
3.4.7 Change in Other Histological Parameters 94.
3.5 Discussion 96.
CHAPTER FOUR - EVOLUTION OF NAFLD IN INSULIN RESISTANCE
RELATED CONDITIONS: HYPOPITUITARISM/ HYPOTHALAMIC
DISEASE
4.1 Introduction 102.
4.2 Aims 104.
4.3 Methods 104.
4.2.1 Case Ascertainment 104.
4.3.2 Patients 106.
4.3.3 Statistical Analysis 107.
4.4 Results 107.
4.4.1 Patient Characteristics 107.
4.4.2 Pituitary /Hypothalamic Disease 108.
4.4.3 Development ofthe Metabolic Syndrome 108.
4.4.4 Association with N A F L D 112.
4.4.5 Liver Related Morbidity and Mortality 116.
4.4 Discussion 117.
CHAPTER FIVE - EVOLUTION OF NAFLD IN INSULIN RESISTANCE
RELATED CONDITIONS: DIABETES MELLITUS
5.1 Introduction 122.
5.2 Aims 124.
5.3 Methods 124.
5.3.1 Study Setting 124.
5.3.2 Case Ascertainment 124.
5.3.3 Case Definitions 125.
5.3.4 Data Abstraction 127.
5.3.5 Statistical Analysis 128.
5.4 Results 128.
5.4.1 Patient Characteristics 130.
5.4.2 Follow-up 131.
5.4.3 Mortality 131.
5.4.4 Effect of N A F L D on Mortality 131.
5.5 Discussion 134.
CHAPTER SIX - GENERAL DISCUSSION
6.1 Findings, Significance and Implications 138.
6.2 Future Directions 145.
6.3 Conclusions 149.
APPENDICES
Appendix One 151.
Appendix T w o 152.
Appendix Three 156.
Appendix Four 157.
Appendix F ive 158.
Appendix Six 159.
REFERENCES 16.2.
LIST OF TABLES
Table 1.1 Primary and secondary types of N A F L D 7
Table 1.2. Histological Grading and staging of NAFLD 18
Table 1.3 Natural history clinical studies of NAFLD 24
Table 1.4 Risk factors for advanced fibrosis in NAFLD 27
Table 1.5 Natural history histological studies of NAFLD 29
Table 1.6 Definitions ofthe metabolic syndrome 35
Table 2.1 Baseline clinical and laboratory features of patients
diagnosed with NAFLD in Olmsted County during
1980 to 2000 (n= 435) 50
Table 2.2 Causes of death among Olmsted County residents with
NAFLD (n=53) and the Minnesotan general population 54
Table 2.3 Predictors of overall mortality by univariate proportional
hazard modelling 58
Table 2.4 Predictors of overall mortality by multivariate cox
proportional hazard modelling 59
Table 2.5 Predictors of liver-related mortality by univariate
proportional hazard modelling 61
Table 2.6 Histological features of patients with NAFLD who
underwent liver biopsy 63
Table 2.7 Baseline clinical and laboratory features of NAFLD
patients who underwent biopsy compared to those
who did not 64
XVI
Table 3.1 Change in liver histology was not different between
patients on drug treatment compared to untreated patients 77
Table 3.2 Change in liver histology was not different between
patients enrolled in clinical trials compared to patients
biopsied for clinical indications 78
Table 3.3 Clinical features at time of initial liver biopsy 82
Table 3.4 Laboratory features at time of initial liver biopsy 83
Table 3.5 Histological features at initial liver biopsy 84
Table 3.6 Change in fibrosis stage between first and last biopsy 86
Table 3.7 Progression of fibrosis stage according to initial fibrosis
stage and time interval between biopsies 87
Table 3.8 Patients with progressive, stable or regressive fibrosis 89
Table 3.9 Predictors of fibrosis rate by univariate regression analysis 91
Table 3.10 Predictors of fibrosis rate by multivariate linear regression
analysis 93
Table 4.1 Liver biochemistry and histology at the time of diagnosis
of NAFLD 114-5
Table 5.1 Clinical characteristics at time of diagnosis of diabetes 129
Table 5.2 Causes of death among patients with diabetes mellitus
with or without N A F L D 132
Table 5.3 Effect of N A F L D on overall survival among subjects
with diabetes mellitus 133
•
LIST OF FIGURES
Figure 1.1 Histological features of non-alcoholic steatohepatitis 16
Figure 1.2 Longitudinal ultrasonographic view demonstrating
hyper-echoic liver compared to the cortex ofthe
right kidney 20
Figure 1.3 Non-contrast abdominal CT scan demonstrating decreased
attenuation ofthe hepatic parenchyma compared to the spleen
with relative enhancement ofthe hepatic vasculature 21
Figure 2.1 Methods of diagnosis of patients with NAFLD in Olmsted
County (1980-1999) 51
Figure 2.2 Overall survival of patients diagnosed with NAFLD in
Olmsted County, Minnesota between 1980-1999 55
Figure 2.3 Overall survival of patients diagnosed with NAFLD in
Olmsted County, Minnesota between 1980-1993 56
Figure 2.4 Survival of patients diagnosed with NAFLD between
1980-1999 in Olmsted County who underwent liver
biopsy compared to those who did not 65
Figure 3.1 Histological features at initial and final biopsy 95
Figure 4.1 Mean levels of triglyceride in the first 48 months
after diagnosis of pituitary /hypothalamic disease 110
Figure 4.2 Mean levels of cholesterol in the first 48 months
after diagnosis of pituitary/hypothalamic disease 111
1
CHAPTER ONE
GENERAL INTRODUCTION AND LITERATURE REVIEW
1.1 INTRODUCTION
Nonalcoholic fatty liver disease (NAFLD) is a common and potentially fatal
disease. Recent epidemiological studies have demonstrated that NAFLD is now
the commonest chronic liver condition in the western world. This is directly
related to the emerging epidemic of obesity and diabetes which are clinical
conditions associated with insulin resistance; a key pathogenic factor in the
development of NAFLD. Thus as the prevalence of these metabolic conditions
increase, it is likely that NAFLD will also become more prevalent.
The significance of NAFLD is determined by its natural history. The
natural history of NAFLD is defined as its prognosis and course without medical
intervention. [1] Examining the natural history of NAFLD involves assessing the
clinical, histological and metabolic evolution of disease. The clinical endpoints of
most consequence are liver related mortality and all-cause mortality. Other
relevant clinical endpoints include morbidity caused by cirrhosis due to the
complications of liver failure, ascites, hepatic encephalopathy, variceal
haemorrhage and hepatocellular carcinoma (HCC). Evaluation ofthe histological
course of NAFLD, particularly progression of fibrosis, allows insight into disease
course prior to the clinical manifestations of cirrhosis. Finally, as NAFLD is part
ofthe systemic metabolic disturbance involving insulin resistance, its impact upon
the development of metabolic disease and morbidity and mortality associated with
metabolic disease is also an important feature of its natural history. Thus,
2
evaluation ofthe natural history of NAFLD involves accurately quantifying the
magnitude of these outcomes and determining the rates at which they are occur.
In addition, as the natural history may be variable, it is important to determine risk
factors for these outcomes as well as to examine the natural history in different
populations.
Knowledge ofthe natural history of NAFLD is critical; firstly from a public
health perspective it provides insight regarding the magnitude and significance of
this emerging public health problem; and secondly from an individual patient
perspective, knowledge ofthe natural history provides important information
regarding prognosis which in turn has implications regarding patient counselling
and monitoring. Furthermore, identification of specific prognostic factors
provides insight into the disease pathogenesis and guides patient management,
particularly the need for further investigation and therapy.
It appears that the natural history of NAFLD is one of progressive liver
injury occurring in a minority of subjects which may lead to cirrhosis and its
complications. It is also emerging that NAFLD is likely to be responsible for a
significant proportion of cases of cryptogenic cirrhosis. [2] However, studies
examining the natural history of NAFLD have had a number of limitations. The
number of patients examined has generally been few and the follow-up period
often relatively short, thus limiting the ability to determine disease related
morbidity and mortality of this chronic condition. Furthermore, studies to date
originate from tertiary referral centres limiting the ability to generalize these
studies to the general population. In addition, important prognostic variables have
not been identified, leaving the clinician unable to stratify patients who are most
at risk of developing complications. Furthermore, the interplay and significance
of insulin resistance on natural history is unclear. Although insulin resistance or
related clinical conditions such as obesity and diabetes are clearly pathogenically
related to the development of NAFLD, it is less clear whether they play significant
roles in the progression of NAFLD. Thus it is unknown whether subjects with
insulin resistance related conditions such as diabetes or
hypopituitarism/hypothalamic disease are at risk of an accelerated disease course.
Therefore it is apparent there are currently significant deficits in our knowledge of
the natural history of NAFLD.
1.2 OBJECTIVES AND AIMS
The overall objective of this thesis is to define the natural history of NAFLD. The
broad hypothesis is that NAFLD is associated with significant morbidity and
mortality among subjects with insulin resistance related conditions. The specific
aims are to examine the clinical and histological evolution of NAFLD among the
general community and among subgroups who have clinical features of insulin
resistance (diabetes mellitus, hypopituitarism/hypothalamic disease) who may be
at risk of an accelerated natural history.
1.3 THESIS OUTLINE
Chapter One, the literature review, will examine the epidemiology of NAFLD to
place into perspective the potential magnitude of this disease. The current
knowledge regarding the pathophysiology of NAFLD will be examined in relation
to understanding potential mechanisms leading to disease progression and thus
patient morbidity and mortality. In addition, the histological features of NAFLD
4
will be reviewed and the current state of knowledge ofthe natural history of
NAFLD will be detailed. In particular, the associations of NAFLD with
cryptogenic cirrhosis, HCC and liver transplantation will be examined. Finally
the review will examine the relationship between NAFLD and clinical conditions
associated with insulin resistance such as the metabolic syndrome,
hypopituitarism/hypothalamic disease and diabetes mellitus.
Chapter Two examines the clinical evolution of NAFLD in a community-
based cohort of subjects to the endpoints of cirrhosis, HCC, liver related death or
transplantation and overall death. Chapter Three studies the histological evolution
of NAFLD, particularly the rates of fibrosis progression in a large cohort of
subjects who have undergone serial liver biopsies. These chapters also examine
potential metabolic, clinical and histological factors that may influence the disease
course of NAFLD. Chapters Four and Five examine the natural history of
NAFLD in two subgroups of individuals with insulin resistance related
conditions, namely patients with hypopituitarism/hypothalamic dysfunction and
patients with diabetes mellitus. Specifically, Chapter Four examines the impact of
NAFLD on development of liver-related morbidity and mortality among patients
with hypopituitarism/hypothalamic dysfunction. Chapter Five examines the
impact of NAFLD on liver-related mortality and overall mortality among a
community-based cohort of individuals with diabetes mellitus.
5
1.4 LITERATURE REVIEW
1.4.1 NAFLD Definition and Terminology
NAFLD is defined by the presence of hepatic steatosis in the absence of excessive
alcohol intake. The minimum amount of hepatic steatosis required for the
diagnosis has been defined by consensus as 5 %.[3] This cut-off is supported by
the findings of a large cross-sectional population-based study which determined
hepatic fat content using the sensitive technique of magnetic resonance
spectroscopy (MRS). This study found that the 95 percentile for hepatic
steatosis content among subjects without risk factors for NAFLD was 5.5%.[4]
The cut-off defining 'excessive' alcohol intake is not known. It is well
established that alcoholic excess/binge may result in hepatic steatosis [5], however
the amount of alcohol which transforms a normal liver to a fatty liver is less well
established. Twenty grams of alcohol per day is commonly accepted as the
appropriate cut-off for NAFLD,[3,6] as population based studies have
demonstrated that the risk of chronic liver disease and cirrhosis begins to increase
after this level. [7,8]
Simple steatosis and non-alcoholic steatohepatitis (NASH) represent the
histological spectrum of NAFLD. Simple steatosis describes the presence of
hepatocellular steatosis in the absence of evidence of hepatocellular damage,
whereas NASH refers to hepatic steatosis in association with lobular inflammation
and ballooned hepatocytes.[3] Fibrosis may be present but is not necessary for the
definition of NASH. It is also recognised that a significant proportion of subjects
with cryptogenic cirrhosis have previously unrecognised NASH. [2]
6
NAFLD may be categorized as primary or secondary depending on the underlying
pathogenic factors (table 1.1). Primary NAFLD is the commonest form and is
associated with insulin resistant states such as diabetes mellitus and obesity.
Multiple mechanisms are implicated in secondary forms of NAFLD [9] including;
disturbances of hepatic lipid transport as in Wolmans disease [10] or choline
deficiency[11]; defective mitochondrial oxidation as seen in mushroom toxin
exposure and acute fatty liver of pregnancy[12,13]; or induction of metabolic risk
factors as observed with corticosteroid administration. The distinction from
secondary types is important as these have differing treatment and prognoses.[14]
In particular, amiodarone and perhexilene induced steatohepatitis may progress to
cirrhosis and liver failure if not recognized early. [9]
7
Table 1.1 Primary and secondary types of NAFLD
Primary N A F L D
Metabolic Obesity, glucose intolerance, hypertension,
features hypertriglyceridemia, low H D L cholesterol[15,16]
Secondary N A F L D
Drugs Corticosteroids, [17] tamoxifen,[18] tormifene,[19]
synthetic oestrogens,[20] diltiazem, nifedipine, [21]
verapamil, methyldopa,[22] choloroquine,[9]
zidovudine,[23] tetracycline,[24] didanosine,
stavudine,[23] aspirin,[14] valproate,[25] cocaine,[14]
Amiodarone,[26] perhexilene,[27]
methotrexate,[28]irinotecan, oxaliplatin[29]
Infections Hepatitis C, [30] human immunodeficiency virus,[31 ]
small bowel diverticulosis with bacterial overgrowth, [3 2]
gram negative sepsis [33]
Hypobetalipoproteinemia,[34] lipodystrophy, [35] Weber-
Christian syndrome,[36] acute fatty liver of
pregnancy, [12] Reyes syndrome, [3 7] Cholesterol ester
storage disease, [10] Wolmans disease, [10] Wilsons
disease[38], Dorfman Chanarin syndrome,[39] Adult
onset type 2 citrullinemia,[40]
Toxins Organic solvents,[41]mushroom toxins (Aminanta
phalloides, Lepiota),[13] phosphorus poisoning,
petrochemical exposure,[42] Bacillus cereus toxin[43]
Nutritional Rapid weight loss, intestinal bypass surgery, [44]
starvation, protein calorie malnutrition, coeliac
disease,[45] inflammatory bowel disease,[46] total
parenteral nutrition,[47] choline deficiency[48]
Metabolic
conditions
8
1.4.2 Epidemiology
Knowledge ofthe prevalence and incidence of NAFLD is principal to
understanding its significance. At an individual patient level, estimates ofthe
disease prevalence provide a likelihood of diagnosis in any one person. From a
public health perspective, it provides an estimate ofthe magnitude ofthe problem
in the general population.
It is clear NAFLD is highly prevalent across a range of ages and ethnic
groups. The prevalence rates reported across different studies vary to some
extent, according to the sensitivity ofthe instrument used to detect NAFLD and
the cut-points used to exclude alcohol use. MRS and liver biopsy are very
sensitive at detecting hepatic steatosis whereas the sensitivity of ultrasound varies
between 49-94%, being less sensitive in the presence of lesser degrees of hepatic
steatosis and higher body mass index. [49,50]
Raised serum aminotransaminase levels not associated with excess alcohol,
viral hepatitis or iron overload, are present in 2.8-5.5% ofthe general population
and are associated with metabolic risk factors for NAFLD such as obesity and
glucose intolerance. [51,52] However, it is apparent that aminotransaminase levels
are insensitive for the detection of NAFLD, with up to 80% of patients having
normal levels at any one time. [4]
A large population based study of 2287 adult subjects aged 30 to 65 years
from Dallas County in the United States, found the adjusted prevalence of hepatic
steatosis as detected by MRS to be 34%. [4] The majority of cases could be
attributed to NAFLD as less than 10% ofthe population drank excessively (>30
gms/day men, >20 gm/day women). Study participants were part ofthe Dallas
Heart Study which consisted of a probability sample of subjects to include 50%
9
black and 50% non-black subjects. Fifty-eight individuals with extreme obesity
(>145 kg) were excluded because ofthe limitations ofthe MRS table. Overall,
hepatic steatosis was slightly more common among males than females with a
1.1:1 ratio, although this increased to 1.8:1 among Caucasians. The prevalence
was higher among Hispanics (45%) compared to Caucasians or African
Americans (33% and 31%, respectively). This difference was only partly
explained by disparity in metabolic risk factors such as obesity, across the ethnic
groups. Overrepresentation of Hispanics has also been observed in newly
diagnosed patients with NAFLD attending health care clinics as well as subjects
diagnosed with NASH related/crypto genie cirrhosis.[53,54]
The prevalence of NAFLD as detected by liver biopsy among potential live
liver donors is comparable to that detected by MRS in the above population based
study. Analysis ofthe liver biopsies of 100 potential donors from the United
States found NAFLD in 33%.[55] It should be remembered that subjects
undergoing health work-up for potential live liver donation are a select group of
individuals who lack major health co-morbidities. Autopsy studies of plane crash
and automobile accident victims reveal hepatic steatosis in 15.6-24% of
individuals, however these studies are confounded by co-existing alcohol
consumption. [56,57]
The prevalence of hepatic steatosis as detected by the less sensitive
technique of ultrasound, is between 13-50% in population based studies of adults
from Japan, China, Korea and Italy. [58-62] Nomura and colleagues found hepatic
steatosis in 14% of 2574 individuals attending free health clinics in Japan. [58]
Alcohol excess was a likely significant contributor among males with 76/168
(45%>) of those with fatty liver drinking >60 grams per day. Another Japanese
10
study of 1680 male clerical workers who underwent mandatory health checkups,
determined the prevalence of fatty liver to be 21.8%, although the contribution
from alcohol was not quantified. [63]
More recently, Fan and colleagues found the adjusted prevalence of fatty
liver among 3175 randomly selected adults from Shanghai, China to be
17.3%). [61] The prevalence of fatty liver among those with an alcohol
consumption of <40 grams per week for more than one year, was 15.4%.
Similarly, a health survey of 4009 Chinese administrative officers attending an
annual health examination revealed a similar prevalence of hepatic steatosis of
12.9%o. [59] Although little information is provided on the degree of alcohol
intake, this is likely to be closely representative ofthe prevalence of NAFLD as
all participants denied regular alcohol ingestion.
In contrast to these reasonably consistent prevalence figures of fatty liver of
between 13-22%, a recent study from Korea found 49.6% of individuals attending
voluntary health screening visits had fatty liver on ultrasound. [62] This
convenience sample may have been biased towards individuals with health
problems who were more concerned about their health and thus more likely to
attend a voluntary screening program. Concordantly, the prevalence of metabolic
risk factors such as central obesity and glucose intolerance was higher in this
Korean study compared to the Chinese study by Fan (43% vs. 36% and 12% vs.
8%> respectively). In addition, alcohol may have accounted for a significant
proportion of hepatic steatosis in the Korean study, as no information on alcohol
consumption was provided.
NAFLD appears to become more prevalent with increasing age, with a peak
prevalence as determined by ultrasound of 25.6% in 40-49 year olds recorded in
11
Japan, whereas a peak of 21.9% was found among adults 65 years and older in
China. [58,59] To date, the only population based study involving children
documented a prevalence of 2.6% (as determined by ultrasound) among 4-12 year
olds, which increased to 22% in the presence of obesity. [64] There are no
population based epidemiological studies examining the elderly, although a
prevalence of NAFLD of 46% has been documented among hospitalized
octogenarians.[65] Interestingly NAFLD in this age group was not associated
with metabolic risk factors suggesting that other pathogenic mechanisms may be
important in the elderly.
The population-based prevalence ofthe different histological subtypes of
NAFLD is difficult to ascertain, as a liver biopsy is required to distinguish
between them. An autopsy study of 351 non-alcoholic patients, revealed a
prevalence of steatohepatitis of 2.7% among normal weight individuals which
increased to 18.5% among massively obese individuals (defined as > 40% above
ideal body weight or fat pad >3cm thick).[66] Diabetes mellitus was also a risk
factor for steatohepatitis, increasing the prevalence from 4.7% among non-
diabetics to 12.2% among those with diabetes. Cirrhosis was found in 9% of
patients with steatohepatitis. Among morbidly obese patients undergoing bariatric
surgery, the prevalence of simple steatosis is 30-38%, the prevalence of NASH is
25-36%) and cirrhosis occurs in 1.4-4.8%.[67-69] The prevalence of cirrhosis
among morbidly obese individuals may be higher as subjects with overt cirrhosis
may be precluded from undergoing surgery and thus would have been excluded
from these series.
Only one study to date has examined the prevalence of fatty liver over
time. [70] Among nearly 40,000 individuals attending health check-ups at a
Japanese University Hospital, the prevalence of fatty liver as detected by
ultrasound increased from 12.6% in 1989 to 30.3% by the year 2000. Throughout
the follow-up period, the prevalence of risk factors for the development of
NAFLD such as obesity, hypertriglyceridemia and hyperglycemia did not change
suggesting other factors such as more sensitive ultrasound technology or increased
operator awareness may have contributed to the rising incidence. However, data
supporting a real increase in the prevalence of NAFLD has recently emerged from
a study comparing the prevalence of elevated aminotransaminase levels in the
U.S. population.[71] Elevated aminotransaminase levels not due to hepatitis C or
excessive alcohol consumption and thus likely secondary to NAFLD, increased
from 4.3% in the period 1988-1994 to 8.1% in 1999-2002.
From the range of epidemiological studies published to date, it is clear that
NAFLD is highly prevalent in the general population. It effects all racial groups
and ages and is now recognized to be the commonest worldwide liver condition,
thus potentially having a tremendous impact on community health and well-being.
It remains particularly crucial to determine the natural history of this common
disease.
1.4.3 Pathophysiology
The pathogenic mechanisms which lead to hepatic triglyceride accumulation and
subsequent hepatocellular damage are complex and incompletely understood.
Accumulation of hepatic triglyceride results when lipid influx and de novo
synthesis in the liver exceeds lipid export and oxidation. Insulin resistance is
clearly important, acting to promote lipolysis of peripheral adipose tissue which
increases free fatty acid (FFA) influx into the liver, subsequently driving hepatic
13
triglyceride production. [72] In addition, hyperinsulinemia and hyperglycemia
promote de novo lipogenesis by up-regulating lipogenic transcription factors such
as sterol regulatory element binding protein-lc (SREBP-lc) and carbohydrate
response element binding protein.[73,74] Furthermore, insulin mediated activation
of SREBP-lc increases malonyl-CoA levels which inhibits FFA oxidation,
thereby favouring hepatic triglyceride accumulation. [73] Lipid export from the
liver may be impaired because of defective incorporation of triglyceride into
apolipoprotein B or reduced apolipoprotein B synthesis or excretion.[75,76]
Hepatic triglyceride accumulation subsequently leads to hepatic insulin
resistance by interfering with tyrosine phosphorylation of insulin receptor
substrates 1 and 2.[77,78] This may exacerbate systemic insulin resistance
creating an escalating cycle of insulin resistance leading to NAFLD which
worsens insulin resistance providing further stimulus for hepatic fat accumulation.
Supporting this idea is evidence that patients with NASH are more insulin
resistant than age, gender and BMI matched controls. [79] In addition,
hypertransaminasemia due to NAFLD chronologically follows weight gain but
precedes glucose intolerance. [80] However, whether the presence of NAFLD
exacerbates insulin resistance to a degree which leads to clinical complications
such as vascular disease is unknown. Particularly, it is unknown whether NAFLD
worsens insulin resistance among patients with diabetes mellitus leading to a
higher incidence of diabetes related morbidity and mortality. This hypothesis is
explored in the fourth experimental chapter or fifth chapter overall.
Hepatic lipid accumulation does not universally result in hepatocellular
injury, indicating that additional secondary insults are important.[81] Substrates
derived from adipose tissue such as FFA, tumour necrosis factor a (TNF-a),
leptin and adiponectin have been implicated in contributing to liver damage in
NAFLD. Increased hepatic FFA oxidation can generate oxygen radicals with
subsequent lipid peroxidation, cytokine induction and mitochondrial
dysfunction. [82] FFA may also lead to hepatocyte apoptosis which is a prominent
mechanism of cellular injury among NAFLD patients.[83,84] Genetic
polymorphisms of inflammatory and fibrogenic cytokines such as TNF-a, tumour
growth factor p\ angiotensinogen have been implicated in the tendency to progress
to NASH, as has polymorphisms of manganese superoxide dimutase which
protects against reactive oxygen species.[85-87] Pro-inflammatory TNF-a levels
may also be increased secondary to gut derived bacterial lipopolysaccharide and
increased fat mass.[84,85,88] Interestingly, a family of proteins stimulated by
cytokine production called 'suppressors of cytokine signalling' exacerbate insulin
resistance and up-regulate SREBP-lc potentially leading to a vicious cycle of
cytokine induction, suppression, insulin resistance and hepatic steatosis. [89]
Insulin sensitizing and potentially hepatoprotective cytokines such as
adiponectin may be inappropriately low among NASH patients.[90] Leptin
appears to be required for hepatic fibrogenesis after experimental liver injury. [91]
In addition, leptin deficient ob/ob mice develop a phenotype of hyperphagia,
obesity, insulin resistance and fatty liver,[92] which is similar to that seen after
experimental hypothalamic injury.[93] Recently disturbances in the
hypothalamic-pituitary-adrenal axis have been described among patients with
NAFLD. Patients with type 2 diabetes mellitus and NAFLD compared to those
without NAFLD, have a subtle hypothalmo-pituitary-adrenal axis disturbance
manifested by increased urinary Cortisol excretion and reduced Cortisol
suppression by dexamethasone.[94] This may be due to the presence of increased
15
amounts of visceral adipose tissue which when compared to subcutaneous fat has
higher activity of 11 (3-hydroxysteroid dehydrogenase resulting in increased
conversion of inactive cortisone to active Cortisol.[95]
Recently, microvascular insufficiency due to architectural distortion from
swollen lipid-laden hepatocytes and fibrosis has been implicated to impair
hepatocyte oxygen and nutrient exchange leading to an inflammatory
response. [96] The inflammatory response has been hypothesised to lead to further
venous obstruction and eventual development of fibrosis. [97]
1.4.4 Histology
The histological changes of NAFLD are indistinguishable from that produced by
alcohol. [98] Thus the diagnosis of NAFLD cannot be made by histological means
alone and requires the clinical exclusion of excessive alcohol intake. The
histological hallmark of NAFLD is hepatocellular triglyceride accumulation,
which is predominantly macrovescicular, although it is often mixed with
microvescicular fat which implies defective mitochondrial FFA oxidation. A
mixed mononuclear/neutrophilic lobular infiltrate may accompany steatosis and is
frequently associated with hepatocyte ballooning and less commonly Mallory's
hyaline. Hepatocellular ballooning, disarray and fibrosis are typically
predominant in zone three ofthe hepatic lobule. Fibrosis is typically pericellular
and perisinusoidal giving a "chickenwire" appearance (Figure 1.1). Eventually
fibrotic septae form between the hepatic vein and portal tract and nodules may
form with cirrhosis. Interestingly, a pattern of fibrosis which is portal based rather
than zone three, has been described among paediatric (predominantly obese)
16
^1. *>
• \ #<f
»i*Pe- ^ . - ^ .. * . - « - % •
» » *
Figure 1.1 Histological features of non-alcoholic steatohepatitis (trichrome stain x
200).[72] Macrovescicular steatosis and pericellular/perisinusoidal fibrosis
located predominantly in zone 3 around the central vein (CV). (Source Dr S.
Sanderson)
17
subjects with NASH as well as morbidly obese adult individuals.[68,99,100]
Whether this reflects different underlying pathogenic insults is not known.
The most widely adopted histological scoring system for NASH is the Brunt
classification (Table 1.2).[101] The histological distinction between simple
steatosis and NASH varies widely in the literature. A recent single topic
conference defined NASH by the presence of steatosis in combination with
lobular inflammation and hepatocyte ballooning, although steatosis with fibrosis
in the absence of significant ballooning or inflammation is also often used.[3] It
appears however, that many pathologists view these features necessary but not
sufficient for a histopathological diagnosis of NASH. One study in which nine
pathologists serially interpreted 50 biopsies from patients with NAFLD, found
that only 68% of cases with the combination of steatosis, lobular inflammation
and hepatocyte ballooning where labelled as having NASH. [102] When fibrosis
was added as a diagnostic feature, the proportion increased to 82%. The authors
subsequently concluded that a composite score of steatosis, lobular inflammation
and ballooning could be used to diagnose NASH; a score of three or less
excluding NASH, a score of four a borderline diagnosis and a score 5 or greater
definite for NASH. Further complicating the histological diagnosis of NASH is
inter-observer and intra-observer discrepancy between histopathologists with
agreement scores (Kappa statistics) being moderate at 0.66 and 0.61-0.62,
respectively. [102,103] In addition, sampling error may also result in under-
diagnosis of NASH with discrepancy in the diagnosis noted in 30-41%) of cases
where paired liver biopsies were taken.[103,104]
18
Table 1.2. Grading and staging of N A F L D according to Brunt and
colleagues.fi 01]
Steatosis Grade
1 <3 3%> of hepatocytes involved
2 33-66%o of hepatocytes involved
3 >66% of hepatocytes involved
Necroinflammatory Activity Grade
Occasional ballooned zone 3 hepatocytes, scattered acinar
PMN's ± lymphocytes, no or mild chronic portal
inflammation
Zone 3 predominant hepatocyte ballooning, obvious intra-
acinar polymorphs, ± zone 3 pericellular fibrosis, mild to
moderate chronic portal and intra-acinar inflammation
Zone 3 predominant obvious ballooning and disarray, intra-
acinar polymorphs ± mild chronic inflammation, mild or
moderate portal inflammation
Fibrosis Stage
0 N o fibrosis
1 Zone 3 predominant perisinusoidal/pericellular fibrosis
2 Stage 1 + focal or extensive periportal fibrosis
3 Stage 2 + focal or extensive bridging fibrosis
4 Cirrhosis
19
1.4.5 Radiology
Radiological techniques such as ultrasound (US), computed tomography (CT) and
magnetic resonance imaging (MRI) are useful to confirm the presence of hepatic
steatosis. Hepatic steatosis on US appears as fine hyperechoic shadows (see
Figure 1.2). Additional characteristic features are; hyperechoic liver parenchyma
in comparison to the renal cortex; attenuation ofthe diaphragm and blurring of
portal vessels.[105] Studies at least a decade old have demonstrated the sensitivity
and specificity of US for hepatic steatosis to be between 60-94% and 84-95%
respectively. [50] Positive predictive values are estimated to be 87-96%>.[50,106]
Accuracy may be increasing further with newer and improved ultrasound
technology. As mentioned previously, the sensitivity of US varies according to
the degree of hepatic steatosis present, as well as with BMI. For example, the
sensitivity of ultrasound at detecting 10-19%, 20-29% and > 30% hepatic steatosis
was 55%), 71% and 80% respectively in one study of 100 subjects evaluated for
living liver donation.[55] Among morbidly obese patients (BMI 35-40 kg/m ),
the sensitivity is as low as 49%>.[49] Thus a positive US is confirmatory for the
diagnosis of hepatic steatosis, whereas a negative US does not exclude disease.
With non-contrast images by CT scan, hepatic steatosis has a low
attenuation and appears darker than the spleen (Figure 1.3). By examining the
differential attenuation between the liver and spleen, a reasonable quantitative
estimate of hepatic steatosis can be made using CT.[107,108] The sensitivity of
CT at detecting greater than 33% hepatic steatosis is up to 93%), with a positive
predictive value of 76%.[109] Both magnetic resonance phase contrast imaging
techniques and MRS are reliable at detecting steatosis and offer good correlation
20
RT KID LONG
£ a
?- "
Figure 1.2 Longitudinal ultrasonographic view demonstrating hyper-echoic liver
compared to the cortex ofthe right kidney (Source Dr J. Talwalkar).
21
Figure 1.3 Non-contrast abdominal C T scan demonstrating decreased attenuation
ofthe hepatic parenchyma compared to the spleen with relative enhancement of
the hepatic vasculature (Source Dr J. Talwalkar).
with hepatic fat volume. [110] However the routine application of M R I is limited
by cost and lack of availability.
Although radiological imaging is useful for confirming the presence of
hepatic steatosis, it is not useful for distinguishing between the different subtypes
of NAFLD (ie. simple steatosis vs. NASH). [109,111] Nor can radiological
imaging stage the degree of liver fibrosis among patients with NAFLD.
1.4.6 Natural History
The clinical evolution of NAFLD is related to the emergence of cirrhosis and its
complications of variceal bleeding, hepatic encephalopathy, ascites, hepatocellular
carcinoma and liver failure. The sub-clinical course of NAFLD can be examined
by studying the histological evolution of NAFLD over time. In particular,
examining hepatic fibrosis, which is predictive of cirrhosis, liver related morbidity
and liver related mortality, provides insight into disease course.
1.4.6.1 Clinical Progression
Progressive NAFLD may lead to cirrhosis with complications of HCC, liver
failure and liver related death. Subfulminant liver failure is a rare presentation of
NAFLD but has been reported in patients with presumed unrecognized cirrhosis
who decompensate due to an unknown insult.[112,113] The association of
NAFLD with the specific outcomes of cryptogenic cirrhosis, HCC and liver
transplantation will be examined in detail later in the review.
Several studies have attempted to examine the incidence of cirrhosis, HCC and
liver related death among patients with NAFLD (summarized in Table 1.3).
23
Table 1.3 Natural history clinical studies of NAFLD.
Author
Matteoni
[115]
Teli
[117]
Dam-Larsen
[114]
Lee
[118]
Powell
[116]
Cortez-Pinto
[119]
Year
1999
1995
2004
1989
1990
2003
n
98
40
109
39
42
32
Diagnosis
NAFLD
Steatosis
Steatosis
NASH
NASH
NASH
Follow-up
(yrs)
8.3 ±5.4
9.6
(0.1-16.0)
16.7
(0.2-21.9)
3.8
(1.0-16.8)
4.5
(1.5-21.5)
5.9 ±4.7
Cirrhosis
20/98
(20%)
0/40
(0%)
1/109
(1%)
3/39
(8%)
2/42
(5%)
3/32
(8%)
HCC
1/98
(1%)
0/40
(0%)
0/109
(0%)
0/39
(0%)
1/42
(2%)
0/32
(0%)
LRD
9/98
(9%)
0/40
(0%)
1/109
(1%)
39
(3%)
1/42
(2%)
3/32
(8%)
Footnote: Follow-up provided as median (range) or mean (± standard deviation),
HCC = hepatocellular carcinoma, LRD = liver related death.
[114-119] The occurrence of cirrhosis ranged between 0 % to 2 0 % , the incidence
of HCC was between 0% to 2% and the incidence of liver related death was 0% to
9%. The differences reported between studies are due to variable follow-up
periods and different patient populations. The number of patients in these studies
varied from 32 to 109 and the mean/median follow-up varied from 3.8 to 16.7
years. Importantly, these studies have demonstrated that the different histological
subtypes of NAFLD have different prognosis. Dam-Larsen and colleagues
followed 109 patients with biopsy proven hepatic steatosis without inflammation
or fibrosis, for a median 16.7 years; only one patient developed cirrhosis and
subsequently died from their liver disease.[114] Similarly, follow-up of 40
patients in the United Kingdom with simple steatosis for a median of 9.6 years
revealed no progression to cirrhosis or liver related deaths.[117] Matteoni
followed-up 98 patients from the Cleveland Clinic who had the full histological
spectrum of NAFLD. [115] Patients were divided into four histological subtypes;
type 1, fatty liver alone; type 2, fat accumulation and lobular inflammation; type
3, fat accumulation and ballooning degeneration; type 4, fat accumulation,
ballooning degeneration, and either Mallory hyaline or fibrosis. It is not clear
how types 2 and 3 were differentiated as lobular inflammation and hepatocyte
ballooning commonly co-exist in patients with NASH. [101,102] Follow-up of 49
type 1 patients with simple steatosis revealed progression to cirrhosis in 2 (4%)
with one patient (2%) dying from their liver disease. No patients with type 2
NAFLD developed cirrhosis or liver related death. In contrast, cirrhosis was more
common in type 3 and 4 NAFLD occurring in 4/19 (21%) and 14/54 (26%)
respectively. Similarly, liver related death occurred in 1/19 (5%) and 7/54 (13%)
respectively. The rates of occurrence of cirrhosis in types 1 and 2 versus types 3
and 4 were significantly different (p<0.001), however the analysis was not limited
to incident cases of cirrhosis introducing the possibility of incidence/prevalence
bias. There was a significant trend for liver related death to be higher in types 2, 3
and 4 compared to type 1 (p=0.08) and types 3 and 4 compared to types 1 and 2
(p=0.06). Interpretation ofthe significance ofthe various histological subtypes in
this study is made difficult by not knowing how many patients had cirrhosis at the
study inclusion, as this may have significantly skewed patient survival. Despite
this, it appears this study is consistent with other studies showing a relatively
benign outcome for patients with simple steatosis compared to those with NASH.
Three studies have examined the clinical outcomes of patients with a
histological diagnosis of NASH. [116,118,119] Patient numbers in these studies
have been relatively small (32 to 42) as has the duration of follow-up (3.8 to 5.9
years). Despite relatively short periods of follow-up, a rather alarming occurrence
of cirrhosis (5-8%>) and liver related death (2-8%) were recorded. However,
hepatitis C which has evidence of hepatic steatosis in up to 70% of cases[120] was
not excluded in any of these studies.
Jepsen and colleagues retrospectively evaluated the outcomes of 1,804
patients who were discharged from hospital with a diagnosis of non-alcoholic or
unspecified fatty liver over a mean of 6.5 years of foliow-up.[121] Compared to
the general population, patients with non-alcoholic/unspecified fatty liver had a
significantly increased risk of death (standardized mortality ratio [SMR] 2.6, 95%
confidence interval 2.4-2.9), which was principally due to an increased risk of
death from hepatobiliary disease (SMR 19.7, 95% C.I. 15.3-25.0), infections
(SMR 6.7, 95%) C.I. 2.9-13.1), other gastrointestinal disease (SMR 4.8, 95% C.I.
2.9-7.4), suicide (SMR 3.1 95% C.I. 1.8-5.2) and respiratory disease (SMR 3.0
95%) C.I. 2.2-4.0). Using the same cohort, the authors published a similar study
demonstrating that patients with non-alcoholic/unspecified fatty liver also had an
increased risk of HCC, colon cancer and pancreatic cancer. [122] These studies
suffer from several methodological issues including potential misclassification
bias from incorrect coding, uncertainty about criteria required for diagnosis, lack
of exclusion of other hepatic diseases such as hepatitis C and selection bias as all
patients were hospitalized.
A major limitation ofthe above studies is that they have limited numbers of
subjects with generally a short duration of follow-up. This has prevented
determination of prognostic factors that influence outcome. In addition, all ofthe
studies originated from major tertiary referral centres in which patients underwent
liver biopsy, often due to clinical indications such as a suspicion of cirrhosis.
Thus the patient selection of these studies limits generalization of their findings to
the general population. Furthermore, only one study has examined the full
spectrum of NAFLD with the remaining studies limiting inclusion to patients with
either steatosis or NASH. [115] Chapter Two examines the clinical evolution of
NAFLD among a large number of community-based individuals who have the full
spectrum of NAFLD.
1.4.6.2 Histological Progression
Since the first descriptions of NAFLD, it has been recognized that hepatic fibrosis
may be progressive and result in cirrhosis.[123] Clinical factors associated with
advanced fibrosis (bridging fibrosis and cirrhosis) among patients with NAFLD
include insulin resistance and its clinical correlates of obesity, diabetes mellitus
and hypertriglyceridemia (Table 1.4). The relationship between insulin resistance
Table 1.4 Risk factors for advanced fibrosis in N A F L D
Author Year n Patient Risk Factors
Population
Angulo[128] 1999 144 NASH Age
Obesity (BMI >30)
Diabetes
AST:ALT>1
Ratziu[129]* 2000 93 Overweight, Age
raised liver BMI >28 kg/m2
enzymes Triglyceride >1.7 mmol/1
ALT > 2x ULN
Dixon[67] 2001 105 Bariatric
surgery
patients
Hypertension
Elevated ALT
Elevated c-peptide
Marchesini[16] 2003 163 NAFLD The Metabolic Syndrome
Hui[90] 2004 109 NAFLD HOMA-IR
Bugianesi[130] 2004 263 NAFLD Insulin sensitivity
Ferritin
Footnote: * outcome was rare septal fibrosis/ numerous septae / cirrhosis. U L N =
Upper limit of normal; AST= aspartate aminotransaminase; ALT= alanine
aminotransaminase; HOMA-IR= homeostasis model of assessment of insulin
resistance.
and advanced fibrosis may be confounded by the fact that cirrhosis leads to
hyperinsulinemia and impaired insulin sensitivity.[124] Therefore factors
associated with advanced fibrosis in these cross-sectional studies are not
necessarily causative.
A number of studies have prospectively examined patients with serial liver
biopsies to determine the progression of fibrosis in patients with NAFLD (Table
1.5). [116-118,125-127] The proportion of patients with histological features of
NASH with progressive fibrosis ranges between 31-57%. The only study
examining patients with simple steatosis revealed progressive fibrosis in 8%> over
a median 11 years follow-up. This implies fibrosis progression is less common
among different histological subtypes of NAFLD. However, the evolution
between simple steatosis to NASH is unknown. In addition, these studies have
been hampered by small numbers, which has limited the ability to define risk
factors for fibrosis progression. A variety of different histological scoring
systems have also been used making direct comparisons difficult. Furthermore,
the majority of patients in these studies underwent repeat liver biopsy for clinical
reasons, potentially biasing the results towards patients with more severe or
atypical disease. Knowledge ofthe rate of fibrosis progression provides an
important insight into the natural history of liver disease and is examined in detail
among a large group of subjects in Chapter Three.
Table 1.5 Natural history histological studies of N A F L D
Author
Hui[131]
Fassio [125]
Harrison [126]
Evans [127]
Lee [118]
Powell [116]
Teli [117]
Year
2005
2004
2003
2002
1989
1990
1995
n
17
22
22
7
13
13
12
Diagnosis
NAFLD
NASH
NASH
NASH
NASH
NASH
Steatosis
Median f/u
(range) years
6.1 (3.8-8.0)
4.3 (3.0-14.3)
5.7(1.4-15.7)*
8.2 (5.5-14.0)*
3.2(1.2-6.9)
4.3(1.0-9.0)
11 (7.6-16.0)
Progressive
fibrosis
9/17(53%)
7/22 (32%)
7/22 (32%)
4/7 (57%)
5/13 (38%)
4/13(31%)
1/12 (8%)
Footnote: * mean
1.4.7 N A F L D and Cryptogenic Cirrhosis
Cryptogenic cirrhosis accounts for approximately 10% of cases of cirrhosis [132]
and 6% of transplants performed in the United States (United Network of Organ
Sharing - UNOS, personal communication). It is becoming increasingly
recognized that NAFLD is responsible for a substantial proportion of cases of
cryptogenic cirrhosis. This is based on an important observation by Powell and
others, that patients with NAFLD may lose histological evidence of hepatic
steatosis as the disease progresses to cirrhosis.[l 16,133] Thus the histological
picture may resemble one of cryptogenic cirrhosis. Subsequently, examination of
metabolic risk factors among patients with cryptogenic cirrhosis has revealed a
similar prevalence of obesity, diabetes and dyslipidemia compared to patients with
NASH, but a higher prevalence compared to patients with cirrhosis from other
aetiologies.[132,134-136] Furthermore, 38-52% patients transplanted for
cryptogenic cirrhosis develop NAFLD post-transplantation compared to 18% of
patients transplanted due to other etiologies.[137,138] Therefore it is important
when examining the natural history of NAFLD, to consider inclusion of subjects
with cryptogenic cirrhosis who have histological or clinical features suggestive of
NAFLD as exclusion of these subjects may falsely under-estimate disease related
morbidity and mortality.
1.4.8 NAFLD and Liver Transplantation
The number of liver transplants performed in the United States for NASH has
increased over the past decade, from 0.1% of transplants in 1994 to 2.1% in 2004
(UNOS, personal communication). This may partly be due to increased
awareness of NASH as a cause of cryptogenic cirrhosis, because the proportion of
31
subjects transplanted for cryptogenic cirrhosis decreased from 9.1% to 6.4%> over
the same time period. In total, NAFLD and cryptogenic cirrhosis account for
approximately one in twelve liver transplants in the United States. A significant
proportion of patients with NAFLD related cirrhosis are unlikely to be listed for
transplantation due to co-morbidities associated with insulin resistance such as
ischemic heart disease and other vascular disease. [ 139-141] Thus reliance on rates
of liver transplantation as an end-point for NAFLD morbidity may underestimate
the true impact ofthe disease.
The presence of steatosis related to metabolic conditions of obesity and
diabetes, is also a risk factor for progressive fibrosis among other chronic liver
diseases such as hepatitis C and alcoholic liver disease. [120] Thus the impact of
NAFLD is also likely to contribute to the likelihood of disease progression and
need for transplantation among patients with chronic hepatitis C infection and
alcoholic liver disease.
Recurrent steatosis post-transplantation is significantly higher among
patients transplanted for NASH cirrhosis (60-100%) compared to patients
transplanted for cholestatic liver disease (5%), alcoholic liver disease (15%), or
hepatitis C (15%>).[142,143] Over a two to three year period post transplantation,
10-33% redevelop NASH, with progression to cirrhosis and need for re-
transplantation also documented. Development of fatty liver is associated with
cumulative steroid dose, although small numbers in these studies may have
prevented identification of other risk factors. [142]
Given the high prevalence of NAFLD, it appears that relatively few
individuals deteriorate to the stage where they require liver transplantation and are
fit enough to undergo the procedure. Those with an accelerated disease course
requiring transplantation are at risk of disease recurrence, which may also have an
accelerated course. It is clearly important for natural history studies to be
performed to identify these at-risk individuals as early as possible so that the
disease course can be modified.
1.4.9 NAFLD and Hepatocellular Carcinoma
The occurrence of HCC among patients with NAFLD appears to be relatively low,
between 0-2%> (see Table 1.3). However, as NAFLD may effect up to one in three
people,[4] the potential disease burden may be considerable. Although HCC has
been documented to rarely occur in NAFLD in the absence of cirrhosis,[144] the
risk appears predominantly in the presence of cirrhosis.
The risk of HCC in patients with NASH related cirrhosis is difficult to
quantify due to the relatively low numbers of patients, inclusion of prevalent
(rather than incident) cases of HCC, lead-time bias and patient populations
including cryptogenic cirrhosis rather than NASH cirrhosis. One Japanese referral
centre found 6/16 (38%>) of patients referred with NASH cirrhosis had a
complicating HCC. [145] In contrast, Hui et al found no cases of HCC in 23
NASH patients with cirrhosis after a median follow-up of five years.[146] The
Japanese study included patients presenting with HCC at time of diagnosis of
NASH related cirrhosis and also routinely screened their cirrhotic patients for
HCC, potentially accounting for the difference between the two studies.
Small studies have demonstrated that cryptogenic cirrhosis is also
associated with an increased risk of HCC. Specifically, a retrospective analysis of
obese patients with cryptogenic cirrhosis revealed 8/27 (30%) developed HCC
over a median follow-up of only nine months which was similar to the incidence
of H C C among matched patients with hepatitis C. [ 147] A m o n g cirrhosis-
related cases of HCC from a United States referral centre, cryptogenic cirrhosis
was the second commonest aetiology of liver disease, accounting for 29%
(30/105) of all cases. [148] Approximately half of the cases of cryptogenic
cirrhosis had histological or clinical features of NAFLD suggesting it was
contributing substantially to the risk of HCC. In contrast, an Italian centre found
cryptogenic cirrhosis to account for only 7% (23/641) of HCC which may reflect
the higher prevalence of viral hepatitis in Italy which accounted for 71%» of all
HCC cases. [149] The prevalence of obesity and diabetes mellitus was
significantly higher in patients with HCC and cryptogenic cirrhosis compared to
those with HCC and viral or alcohol related cirrhosis, suggesting that NAFLD is
the underlying aetiology in many patients with cryptogenic cirrhosis complicated
by HCC.
The above studies confirm the hepatocarcinogenic potential of NAFLD,
however the magnitude of risk remains unclear due to a lack of well characterised
cohort studies of sufficient sample size and duration of follow-up. This is
discussed more fully in Chapter Two ofthe thesis.
1.4.10 NAFLD and Metabolic Disease
It is well established that insulin resistance is a key factor involved in the
pathogenesis of NAFLD. [3] Therefore conditions associated with insulin
resistance such as the metabolic syndrome and diabetes mellitus are commonly
associated with NAFLD and may have prognostic significance. [16,150] An
important part ofthe current study is to gain insight into the impact and
significance of metabolic factors on the natural history of N A F L D . The complex
associations between NAFLD and metabolic disease are outlined below.
1.4.10.1 NAFLD and The Metabolic Syndrome
The metabolic syndrome represents a constellation of risk factors which are
closely related to insulin resistance and associated with increased cardiovascular
risk. [151] Definitions ofthe metabolic syndrome revolve around the presence of
obesity, hyperglycemia and dyslipidemia (see Table 1.6). Insulin resistance is
present in up to 66-83%) of subjects with NAFLD and is a strong risk factor for
NAFLD among normoglycemic non-obese individuals. [152-154] Concordantly,
metabolic risk factors are common among subjects with NAFLD, being present in
the majority (85%) of subjects with NAFLD. The majority (56-79%) of patients
with NAFLD are overweight (BMI >25 kg/m2) however, 69% of lean patients
(BMI<25 kg/m ) have at least one metabolic risk factor. [16] One third of
individuals with NAFLD have the complete metabolic syndrome.[4,16,130]
Metabolic risk factors significantly increase the risk of NAFLD as follows; central
obesity (odds ratio 5.9, [95% confidence interval 3.7-9.1]), hypertriglyceridemia
(OR 3.4, [2.1-5.6]), low HDL cholesterol (OR 6.2, [3.7-11.1]), systolic
hypertension (OR 2.0, [1.3-3.1]) and fasting hyperglycemia (OR 9.1, [4.3-
20.0]).[157] Among non-obese (BMI 18.5-29.9 kg/m2), non-diabetic (fasting
glucose <7.0 mmol/1) individuals, metabolic factors increase the risk of NAFLD
as demonstrated in a study of 768 individuals where the following features were
associated with a diagnosis of NAFLD; impaired fasting glycemia (odds ratio 2.8,
[95% confidence interval 1.5-5.20]), hypertriglyceridemia (OR 2.8, [2.0-4.0]),
hyperuricemia (OR 2.6, [1.6-4.1]), central obesity (OR 2.4, [1.7-3.4]),
35
Table 1.6 Definitions ofthe metabolic syndrome
Adult Treatment Panel IH[151]
Diagnosis requires three of the five criteria below;
Obesity Waist circumference > 102 cm men; >88 cm women
Hyperglycemia > 110 mg/dl or > 6.1 mmol/1
Hypertriglyceridemia > 150 mg/dl or 1.7 mmol/1
Low H D L cholesterol <40mg/dl (<1.0 mmol/1) men; <50mg/dl (<1.3 mmol/1)
women
Hypertension Blood pressure >135/85 m m H g
World Health Organization^55]
Diagnosis requires hyperinsulinemia or hyperglycemia plus any 2 other criteria;
Hyperinsulinema Upper quartile of fasting insulin in non-diabetic population
Hyperglycemia > 110 mg/dl or > 6.1 mmol/1
Obesity Waisthip ratio >0.9 men; >0.85 women or B M I >30 kg/m2
Hypertension Blood pressure >160/90 m m H g
Hypertriglyceridemia Triglycerides > 150 mg/dl or 1.7 mmol/1; H D L cholesterol
and/or low H D L <35mg/dl (<0.9 mmol/1) men; <39mg/dl (<1.0 mmol/1)
cholesterol women
Microalbuminuria Urinary albumin >20 ugm/min or albxreat ratio >20 mg/g.
International Diabetes Federation[156]
Diagnosis requires central obesity plus two ofthe other four criteria below;
Obesity Waist Circumference
Europoids: > 94 cm men; > 80 cm women
South Asians/Chinese: > 90 cm men; > 80 cm women
Japanese: > 85 cm men; > 90 cm women
Hyperglycemia > 100 mg/dl or > 5.6 mmol/1
Hypertriglyceridemia > 150 mg/dl or 1.7 mmol/1
Low H D L cholesterol <40mg/dl (<1.0 mmol/1) men; <50mg/dl (<1.3 mmol/1)
women
Hypertension Blood pressure > 130/85 m m H g
hypertension (OR 1.7, [1.2-2.4]) and low H D L cholesterol levels (OR 1.4, [11.0-
2.0]).[158]
Central or upper body obesity (apple shape) appears to be a greater
metabolic risk factor for NAFLD than lower body obesity (pear shape). Visceral
fat has greater rates of lipolysis and cortisone enzymatic activity resulting in
enhanced glucocorticoid expression and increase free fatty acid (FFA) release into
the portal vein which flows directly to the liver.[95,159] Consequently, visceral
adipose mass correlates more closely with degree of hepatic steatosis than
BMI.[160] Concordantly, a study of 2,704 residents of New York State
demonstrated a greater correlation between serum hepatic aminotransferase levels
and abdominal height (a measure of central obesity) than BMI.[161] Abdominal
obesity as determined by elevated waist circumference also increases the risk of
developing NASH as opposed to simple steatosis. [16]
Not surprisingly, as metabolic risk factors are common among individuals
diagnosed with NAFLD, individuals with metabolic risk factors such as obesity,
hyperglycemia and hyperlipidemia commonly have NAFLD. A random
population-based sample from Northern Italy demonstrated that the prevalence of
NAFLD increased dramatically from 16% among non-obese (BMI<25 kg/m ),
non-diabetic (fasting glucose <7.0 mmol/1) individuals to 76% in obese non-
diabetic subjects.[162] Among morbidly obese patients undergoing bariatric
surgery, the prevalence of fatty liver is 78-96%>. [67,163,164] The prevalence of
NAFLD also increases with severity of glucose intolerance. Among Japanese
adults undergoing routine health examinations, NAFLD was present in 27%> of
subjects with normal fasting glucose (<6.1 mmol/1), 43% among those with
impaired fasting glycemia (6.1-6.9 mmol/1) and in 6 2 % among patients with
newly diagnosed diabetes (>7.0 mmol/1).[60]
Hypertriglyceridemia and low high density lipoprotein (HDL) cholesterol
are the characteristic lipid abnormalities that accompany insulin resistance and
obesity. [165] Among patients attending an urban based hospital lipid clinic, the
prevalence of NAFLD was 70%> among those with hypertriglyceridemia compared
to 33%) among patients with isolated hypercholesterolemia.[166]
1.4.10.2 NAFLD and other Metabolic Diseases
The clinical features ofthe metabolic syndrome are recognized to be closely
associated with insulin resistance. Other clinical syndromes also closely
associated with insulin resistance include polycystic ovarian syndrome (PCOS),
lipodystrophy, obstructive sleep apnoea (OSA), hypopituitarism and hypothalamic
disease. [167-170] Polycystic ovarian syndrome is characterised by ovulatory
disturbances, hyperandrogenemia and polycystic ovaries. [167] Women with
PCOS are typically hyperinsulinemic and often centrally obese and frequently
have features of metabolic syndrome. [171] The prevalence of raised alanine
aminotransferase levels is 30%> among this group of patients, most likely as a
result of concomitant NAFLD. [172] The significance of NAFLD in PCOS has
not been examined. The lipodystrophies are a group of rare disorders signified by
abnormal adipose distribution, insulin resistance and hypoleptinemia.[35] The
course of NAFLD among some patients with lipodystrophy is aggressive with
cirrhosis reported in the severe generalized forms.[35] Interestingly, the
histological changes of NAFLD in these patients may respond to leptin
supplementation. [ 173]
Obstructive sleep apnoea (OSA) refers to sleep-disordered breathing
associated with collapse ofthe upper airways. OSA is associated with insulin
resistance independent of body weight, putatively because of increased
sympathetic drive, disordered leptin metabolism and sleep deprivation. [174]
Patients with NAFLD and severe OSA are more likely to have elevated liver
enzymes and hepatic necroinflammation and fibrosis, perhaps due to hypoxic
injury.[175] Furthermore, liver enzyme abnormalities (but not insulin resistance)
in patients with OSA improve with continuous positive airway pressure
treatment. [176]
Hypopituitarism and hypothalamic dysfunction are metabolic conditions
which are also recognized to associated with insulin resistance and abnormal
leptin levels. [168,177-180] The relationship between these metabolic conditions
and NAFLD is unknown. It is hypothesized in Chapter Five, that NAFLD may
develop subsequent to hypopituitarism/hypothalamic disease as a result of
metabolic abnormalities. In addition, the natural history may be severe as
observed in other clinical conditions associated with severe insulin resistance such
as lipodystrophy.
1.4.10.3 NAFLD and Diabetes Mellitus
The incidence of diabetes mellitus is increasing dramatically in the Western
world.[181] Recent prevalence figures from the United States demonstrate that
7.9%) ofthe population has diabetes mellitus and the life-time risk of developing
diabetes among people born in the year 2000 is between 33-38%.[181,182]
Diabetes mellitus and NAFLD are both part of a spectrum of diseases that revolve
around insulin resistance. It is therefore not unexpected that these diseases
frequently co-exist. A m o n g populations of N A F L D patients from tertiary
hospitals, the prevalence of diabetes mellitus varies from 27-62%. [16,60,90]
Among patients with diabetes mellitus attending hospital clinics, the prevalence of
NAFLD is 49-55%.[183,184] In studies originating from the general population,
the prevalence of impaired fasting glycemia (fasting glucose > 110 mg/dl or 6.0
mmol/1) among patients with NAFLD is only 18%, illustrating the substantial
referral bias involved in studies from referral centres. [4]
Diabetes mellitus is a risk factor for advanced liver fibrosis on liver biopsy
among NASH patients. [128] The mechanism may be related to hyperinsulinemia
and hyperglycemia promoting fibrosis via up-regulation of connective tissue
growth factor.[185] In addition, diabetes is a risk factor for liver related death as
well as overall death among patients with NAFLD. [150] Whether NAFLD is a
significant risk factor for mortality among patients with diabetes mellitus is
explored in Chapter Five.
1.4.11 SUMMARY
NAFLD is a common condition affecting every one in three to one in five
individuals. The pathogenesis of NAFLD is intimately related to insulin
resistance and thus NAFLD frequently occurs in the presence of metabolic
conditions such as diabetes mellitus and obesity. Diagnosis requires confirmation
of hepatic steatosis by imaging or liver biopsy and clinical exclusion of excessive
alcohol intake. NAFLD may cause progressive hepatic fibrosis that may lead to
cirrhosis with subsequent serious complications including liver failure and
hepatocellular carcinoma. It is clear that only a proportion of subjects will
develop these complications, however at present the magnitude of this risk is
uncertain. Furthermore, the ability to predict patients most at risk of disease
morbidity or mortality is lacking. Therefore there is a critical need to define the
natural history of this common disease.
CHAPTER TWO
CLINICAL EVOLUTION OF NAFLD
2.1 INTRODUCTION
Nonalcoholic fatty liver disease is a common cause of chronic liver disease that
may progress to cirrhosis and its complications. NAFLD is pathogenically related
to insulin resistance and thus closely associated with the metabolic syndrome.
With the increasing prevalence of obesity, diabetes and the metabolic syndrome
[181,186,187], NAFLD is increasingly recognized and is now known to be
present in up to 31% of American adults, which equates to over 60 million
people. [73] Histologically and clinically, NAFLD may manifest as a wide
spectrum of disease, from asymptomatic bland steatosis to nonalcoholic
steatohepatitis with advanced fibrosis and complications of liver failure and portal
hypertension. [14] In addition, NAFLD is now recognized to account for many
cases of cryptogenic cirrhosis. [2]
Despite being common and potentially serious, the natural history of
NAFLD remains poorly defined. The few studies reported to date have limited
numbers and originate from specialist centres with highly selected
populations. [114-118] To date, the largest study including patients with the full
histological spectrum of NAFLD found the occurrence of cirrhosis to be 20% with
a liver related death rate of 9% over a mean follow-up of 8.3 years.[115]
Extrapolation of these morbidity and mortality rates to the 60 million people in
the United States general population with NAFLD, would equate to an enormous
disease burden over the next decade. However, as population-based studies to
determine the long-term prognosis of N A F L D have not been conducted, it
remains uncertain whether morbidity and mortality rates currently reported can be
generalized to community based practice where patients may have a milder
disease. Knowledge ofthe natural history of NAFLD is needed to guide patient
prognosis with implications for management decisions regarding further invasive
investigation, need for treatment and ongoing monitoring. In addition,
quantification ofthe disease burden in the community has important public health
ramifications.
2.2 AIMS
The broad aim of this chapter is to examine the natural history of NAFLD in a
general population setting. The specific aims are to determine among subjects
diagnosed with NAFLD in Olmsted County, Minnesota;
1) Rate of overall mortality in comparison with the Minnesota general
population of same age and gender.
2) Rate of liver related morbidity and mortality.
3) Predictors of overall and liver-related mortality
This study has been published in part (Adams et al, Gastroenterology, 2005; 129:
113-21).
2.3 M E T H O D S
2.3.1 Setting
The study population consisted of residents of Olmsted County, which has a
geographical size of 1695 km2 (654 miles2) and is located in south-eastern
Minnesota ofthe United States of America. According to census data, the total
population in the year 2000 was 124,000 people, with 81% of these living in
urban areas and the remainder in rural farming areas.[188] The proportion of
Caucasians in the population was 90.3%, as compared to the general US
population of 75.1%. The proportion of people with college education was 34.7%
as compared to the general US population proportion of 24.4%.
2.3.2 Rochester Epidemiology Project
To examine the natural history of NAFLD in Olmsted County, the resources of
the Rochester Epidemiology Project (REP) were utilized. The REP is a unique
data linkage system funded by the National Institutes of Health to investigate
disease in a population setting. The REP indexes all medical diagnoses made by
health care providers in Olmsted County Minnesota. [189] The two main
providers of health care in Olmsted County (Mayo Clinic and Olmsted Medical
Centre) as well as local private physicians are indexed, allowing coverage of
primary, secondary and tertiary health care. Consequently all diagnoses and
medical interventions made in the outpatient office or clinic visits,
hospitalizations, emergency room visits, nursing home care, surgical procedures,
autopsies and death certificates are routinely abstracted and recorded. Diagnoses
are coded according to the Hospital Adaptation ofthe International Classification
of Diseases (HICDA). Linkage of these medical records allows assessment of
disease epidemiology and outcomes. Each year, 87%. ofthe population of
Olmsted County will visit a health care provider linked to the REP on at least one
occasion, and in any given four-year period, at least one health encounter is
recorded in over 94%> ofthe OC residents. This allows evaluation ofthe health
status of what is effectively the whole county. [189]
2.3.3 Case Ascertainment
Patients residing in Olmsted County who had been diagnosed with fatty liver,
hepatic steatosis, steatohepatitis or cryptogenic cirrhosis over a 20 year period
between January 1st 1980 and January 1st 2000, were ascertained from the REP
master diagnostic index. Residents with other liver disease including viral
hepatitis, cholestatic liver disease, hemochromatosis, Wilson's disease, alpha one
anti-trypsin deficiency or auto-immune hepatitis were excluded. In addition,
subjects with alcohol related conditions were also excluded. The specific HICDA
codes utilized for the search are listed in Appendix One.
Although fatty liver was recognized prior to 1980, this liver condition was
better characterized in 1980, hence we chose to identify patients after this
date. [123] The date 01/01/2000 was chosen to allow a 20-year ascertainment
period. Follow-up was extended up to 12/31/2003.
All medical records were reviewed in detail, and patients were included
only if fatty infiltration ofthe liver was confirmed on imaging studies (ultrasound,
computed tomography or magnetic resonance imaging) or liver biopsy. In
addition, because a sizable proportion of cryptogenic cirrhosis is due to
NAFLD[2], patients diagnosed with cryptogenic cirrhosis who also had the
metabolic syndrome prior to diagnosis were also included in the cohort. The
metabolic syndrome was defined using the criteria proposed by the National
Cholesterol Education Program (ATP III), (see Table 1.6).[151] As waist
circumference to determine central obesity was not measured for most of our
patients, we used BMI >30 kg/m to define obesity in concordance with the WHO
definition of obesity.[190] Patients were excluded if there was evidence of other
liver disease on clinical history or examination, laboratory studies, imaging or
biopsy. Patients at risk of viral hepatitis due to injecting drug use or blood
product transfusion prior to 1992, were excluded if they had not had hepatitis B or
C serology performed after their exposure. In addition, patients with secondary
causes of fatty liver (listed in Table 1.1 - Chapter One) were excluded. Patients
with an average weekly ethanol consumption >140 grams were excluded. [3]
During the study period, information on alcohol consumption (type, amount,
frequency, duration as well as alcohol abuse screening questions) was
prospectively collected as part ofthe medical record, by a patient history form
filled out by the patient and reviewed by a nurse and physician at each clinical
encounter.
Initial search of the REP diagnostic index identified 620 patients. By chart
review, 86 had evidence or risk factors for other liver disease; 46 had a history of
excessive alcohol ingestion; 41 had secondary causes of fatty liver, five were non
residents of Olmsted County; four were initially diagnosed pre-1980; two patients
denied research authorization for use of their medical records for research, and
one patient had cryptogenic cirrhosis but only two features ofthe metabolic
syndrome. These patients were excluded from the study. The remaining 435
patients formed the study population. Fifteen of these patients were first
diagnosed with fatty liver at the time of post-mortem and were therefore excluded
from the survival analysis.
2.3.4 Patient Information
The following baseline information was recorded on a data abstraction sheet (see
Appendix Two) from the medical record at time of either radiographic or
histologic confirmation of NAFLD diagnosis; age, gender, ethnicity, BMI,
average weekly alcohol intake, history of diabetes (fasting glucose > 7.0 mmol/1
[> 126 mg/dl] or requiring dietary or pharmacological therapy), impaired fasting
glycemia (IFG, fasting glucose > 6.1 mmol/1 [> 110mg/dl]), hypertriglyceridemia
(fasting triglyceride > 1.7 mmol/1 [> 150 mg/dl]), low fasting HDL cholesterol (<
1.0 mmol/1 [< 40mg/dl] for men and <1.3 mmol/1 [< 50mg/dl] for women) or
hypertension (BP > 130/>85 mmHg or under treatment). As waist circumference
was not routinely available, obesity was classified according to the WHO
definition of BMI >30 kg/m . A prior medical history of conditions which may
potentially affect overall survival were also abstracted including ischemic heart
disease (defined as prior myocardial infarct or angina pectoris), cerebrovascular
disease (defined as prior cerebrovascular accident or transient ischemic attack)
and history of previous malignancy (type and date of diagnosis). In addition
symptoms of fatigue, abdominal pain or discomfort were noted. The following
laboratory values within one month of diagnosis were obtained; aspartate
aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase,
bilirubin, albumin, platelet count, prothrombin time, fasting glucose and lipids
(triglyceride, HDL cholesterol, total cholesterol) in addition to date of first
documented abnormal aminotransferase level. The following information was
recorded if performed within 6 months of diagnosis (number of patients with
results available in parentheses); hepatitis B (332) and C serology (307), anti-
nuclear antibody (190), gamma globulin (170), anti-smooth muscle antibody (95),
anti-mitochondrial antibody (120), ferritin and transferrin saturation (257),
ceruloplasmin (135), alpha-1 antitrypsin level and phenotype (124) as well as
radiological findings (415). Liver histology was scored according to the schema
developed by Brunt and colleagues[101] by a single liver pathologist (Dr Schyler
Sanderson, Division of Anatomical Pathology, Mayo Clinc) who was unaware of
the patient details. NASH was defined as the presence of steatosis plus lobular
inflammation plus hepatocellular ballooning in accordance with recent conference
guidelines[3] or steatosis plus any stage of fibrosis.
Follow-up was obtained from medical charts and death certificates.
Subsequent diagnosis of cirrhosis (clinical or histological), ascites, hepatorenal
syndrome, variceal haemorrhage, jaundice, encephalopathy, hepatocellular
carcinoma, liver transplantation, diabetes, hypertension, hypertriglyceridemia or
low HDL were recorded. In addition, date and cause of death were determined.
2.3.5 Statistical Analysis
Data review and statistical analysis was performed by L.Steven Brown
(Department of Biostatistics, Mayo Clinic) and myself with consultation from
James F. Lymp (Dept. of Biostatistics) and Jenny St.Sauver (Division of
Epidemiology, Mayo Clinic). The results are displayed in tables with categorical
variables presented as number and percent and continuous variables presented as
mean and standard deviation. The five yearly incidence of NAFLD diagnoses was
calculated using U.S. census data regarding the total population of Olmsted
County. [188]
Trends in incidence rates over time were analysed using Poisson regression
modelling with the response variable being the number of NAFLD cases and the
predictor variable being five year intervals from 1980-1999.
Observed survival ofthe NAFLD cohort and the expected survival ofthe
general Minnesota population ofthe same age and gender are described using
Kaplein-Meier curves. The expected survival ofthe Minnesota population was
calculated using the method of Ederer, based on age and gender specific
conditional probabilities of death in the subsequent year from published census
tables.[191] One-sample log rank tests were used for comparison with the
Minnesota population and risk was quantified using standardized mortality ratios
(SMR) (calculated as the observed deaths/expected deaths after adjustment for age
and gender).
Cox proportional hazards modelling was used to evaluate overall mortality
in the NAFLD cohort. Variables that were significantly associated with overall
mortality at the 0.05 level in separate Cox proportional hazard models were
analysed in a multivariable model. Risk associated with significant variables were
quantified using hazard ratios (HR). Analysis was performed using SAS Release
8.2 (SAS Institute Inc., Cary, NC).
2.4 R E S U L T S
2.4.1 Patient Population
The study cohort of 435 patients were predominately middle aged Caucasians
with an equal gender distribution (Table 2.1). Ofthe non-Caucasian patients, 20
(5%) were Asian, 6 (1%) Arabic, 5 (1%) Hispanic and 2 (0.5%) African-
American. Features of the metabolic syndrome were common. In particular,
obesity and dyslipidemia was present in over two thirds ofthe population; IFG
and hypertension were present in one third and one quarter ofthe cohort had
diabetes mellitus. Two thirds ofthe cohort had elevated aminotransaminases at
diagnosis. A history of ischemic heart disease was present in 39 (9%) patients
and cerebrovascular disease in 14 (3%). Two hundred and sixty three (60%)
patients were lifetime non-smokers, whereas 171 (40%) were current or previous
smokers (minimum of one cigarette per day). A history of prior malignancy was
found in 29 (7%) patients.
The diagnosis was confirmed by imaging in the majority of cases (Figure
2.1), either by abdominal ultrasound (n=350), computed tomography (n=62) or
magnetic resonance imaging (n=l). Ofthe patients undergoing liver biopsy, the
commonest indication was for investigation of abnormal liver tests (n=43),
followed by post-mortem biopsies (n=20), suspicion of cirrhosis (n=6),
unexplained hepatomegaly (n=4) and staging for malignancy (n=4). Seven
patients had 'other' indications including pre-methotrexate therapy and
investigation of suspected masses. Five patients who underwent post-mortem
biopsies were diagnosed with NAFLD prior to death by imaging studies. Only
two patients in the cohort were diagnosed with cryptogenic cirrhosis and the
Table 2.1 Baseline clinical and laboratory features of patients diagnosed with
NAFLD in Olmsted County during 1980 to 2000 (n= 435).
Variable
Age (years)
Male
Caucasian
Abdo. pain/fullness
Fatigue
BMI (kg/m2)
Obesity
Glucose (mg/dl)
Diabetes
Hyperglycemia
Hypertension
Triglyceride (mg/dl)
Hypertriglyceridemia
49+15
213/435
(49%)
402/435
(92%)
115/420
(27%)
54/419
(13%)
33.5 ±6.5
299 / 420
(71%)
119 + 49
112/435
(26%)
159/435
(37%)
155/431
(36%)
227 ±161
263/389
(68%)
Variable
HDL cholesterol
(mg/dl)
Low HDL
cholesterol
A S T (U/l)
Elevated A S T
A L T (U/L)
Elevated A L T
Bilirubin (mg/dl)
Elevated Bilirubin
Alk. Phosphatase
(mg/dl)
Elevated Alk.
Phos.
Prothrombin (sec.)
Elevated
Prothrombin
Platelet count
(xl09/l)
Low Platelet count
41 ±12
117/359
(33%)
51 ±52
276/418
(66%)
77±48
198/254
(78%)
0.8 ±1.1
48 / 393
(12%)
199+115
344 / 404
(85%)
10.6+1.2
20 / 208
(10%)
245 ± 66
14/407(3%)
Footnote: Data provided as mean ± standard deviation or number (percentage) with patients with missing data excluded from denominator. Elevated A S T as >31 U/L; elevated A L T >40 U/L; elevated bilirubin >1.2 mg/dl; elevated alkaline phosphatase >115 U/L in males and >108 in females; elevated prothrombin time >
12.0 sec; low platelet count <150 x 109/1.
51
15 Patients initially diagnosed at post-mortem excluded from survival analysis
435 Patients 420
Patients
Imaging Alone (n=348)
Imaging and Biopsy (n=65)
Biopsy Alone (n=5)
Cryptogenic Cirrhosis + Metabolic Syndrome (n=2)
Figure 2.1 Methods of diagnosis of patients with N A F L D in Olmsted County
(1980-1999).
metabolic syndrome. One of these was an obese (BMI 52.2 kg/m2) 44 year old
male with a history of diabetes, hypertension, hypertriglyceridemia and low HDL
cholesterol. The other patient was a 71 year old obese (BMI 30.8 kg/m2)
hypertensive female with a medical history of low HDL cholesterol and glucose
intolerance.
2.4.2 Incidence of Diagnosis of NAFLD
The age and sex adjusted incidence rate for the diagnosis of NAFLD was 21.7
/l00,000 person years [95% confidence interval, 19.6-23.7] for all persons and
was 29.7/100,000 person years [95% CI 26.9-32.5] for adults 20 years or older.
There was a significant increase in the incidence rate of diagnosis for all persons
over the study period; 4.3/100,000 [2.4-6.2] in 1980-84; 9.1/100,000 [6.5-11.8] in
1985-89; 26.4/100,000 [221.0-30.7] in 1990-94; 38.4/100,000 [33.4-43.4] in
1995-99 (pO.001). The increase in incidence of diagnosis over the 20-year study
period among adults aged 20 or older was also significant (p<0.001). More
patients were cirrhotic at presentation during the first decade compared to the
second (3/58 [5.2%] vs. 5/362 [1.4%], p<0.05).
2.4.3 Overall Mortality
Survival analysis was performed on 420 patients after excluding 15 patients
whose initial diagnosis of NAFLD was at post-mortem. From time of diagnosis
of NAFLD, mean (+ standard deviation) follow-up was 7.6 ± 4.0 years (median
7.1 years, range 0.1-23.5), culminating in a total of 3192 person-years. Follow-up
to within 6 months ofthe study end (12/31/03) was available in a further 281
(67%) patients, equalling a total of 334 (80%) patients who were either censored
due to death or had acceptable follow-up. Overall 53 (12.6%) patients had died
by the end of follow-up. The causes of death ofthe study cohort are shown in
Table 2.2 and compared to the commonest causes of death in the adult
Minnesotan general population. [192] The commonest reasons for death were
malignancy (28% in total, 9% bowel, 8% pancreas, 4% breast, 9% other),
ischemic heart disease (25%) and liver disease (13%). One patient died from
HCC on a background of cirrhosis and was classified as a liver related death.
Survival among NAFLD patients was significantly less than the expected survival
ofthe general Minnesota population of similar age and gender, with a SMR of
1.34 [(95%> CI 1.003-1.76), p=0.03] as shown in Figure 2.2. To examine patients
with at least 10 years of follow-up, a subset of 161 subjects diagnosed with
NAFLD between 1980-1993 were analysed (Figure 2.3). As follow-up
lengthened, survival worsened compared to the general population (SMR 1.55
[95% CI 1.11-2.11], p=0.005). At 10 years, the survival in this subgroup of
NAFLD patients was significantly lower compared to the Minnesota population
(77% vs. 87%, p<0.005, log rank test).
Table 2.2 Causes of death among Olmsted County residents with NAFLD (n=53)
and the Minnesotan general population.
1.
2.
3.
4.
5.
6.
7,
8.
9.
10.
11.
12.
Olmsted County
Residents with NAFLD
Malignancy
Ischemic heart disease
Liver Disease
Infection
Obstructive lung disease
Congestive cardiac failure
Cerebrovascular accident
Gastrointestinal bleed
Pulmonary embolus
Aortic aneurysm
dissection
Smoke inhalation
Retroperitoneal
N (%)
15
(28%)
13
(25%)
7 (13%)
6(11%)
2 (4%)
2 (4%)
1 (2%)
1 (2%)
1 (2%)
1 (2%)
1 (2%)
1 (2%)
Minnesota General
Population
Malignancy
Heart disease
Cerebrovascular
disease
Chronic respiratory
disease
Accidents
Diabetes mellitus
Alzheimer's disease
Influenza/Pneumonia
Renal disease
Suicide
Parkinson's disease
Hypertension
N (%)
24.5%
23.5%
7.4%
5.0%
4.6%
3.2%
3.1%
2.2%
1.8%
1.2%
1.1%
1.1%
13.
haemorrhage
Unknown 2 (4%o) Liver Disease 0.9%
^ 0.6 -1
| 0.5 ^
I 0.4 (0
0.3
0.2
0.1
0.0
Expected
p = 0.03
i i i i i i i i i i i i i i i >
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Time (years)
No. at risk 420 399 389 382 361 306 254 217 176143 109 71 54 40 31 23 14
Figure 2.2 : Overall survival of patients diagnosed with N A F L D in Olmsted
County, Minnesota between 1980-1999. Survival is compared to the general
population of Minnesota ofthe same age and gender.
1
0
0,
^ 0
5" - o ra | 0 OT 0
0
0
0
0.
No. at risk
9
8 •
7 •
6
5
4 •
3 •
2
1
0
p = 0.005
Expected
'"•••**
Observed
— i 1 • • • • i i i i i • i i i >
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Time (years) 161 155 151148 141138 133 128 119 114 106 71 54 40 31 23 14
Figure 2.3 Survival of patients diagnosed with N A F L D in Olmsted County,
Minnesota between 1980-1993. Survival is compared to the general population of
Minnesota ofthe same age and gender.
Predictors of Overall Mortality
By univariate Cox regression analysis, an increased risk of death was associated
with age, baseline diagnosis of cirrhosis, history of ischemic heart disease,
previous or current smoking history, hypertension and IFG/diabetes (Table 2.3).
Risk of death was not significantly associated with the presence of dyslipidemia,
BMI or the metabolic syndrome. In a multivariable Cox regression model, only
IFG/diabetes, cirrhosis and age remained significantly associated with death
(Table 2.4).
Liver-Related Morbidity and Mortality
Cirrhosis was diagnosed in 21 (5%) patients with 8 (2%) diagnosed at
initial presentation and 13 (3%) during follow-up. Thirteen (3.1%) patients
developed cirrhosis related complications; including ascites in 10 (2%); jaundice
in 8 (2%); portosystemic encephalopathy in 7 (2%); variceal bleeding in 5 (1%);
and HCC in 2 (0.5%) patients. One patient underwent liver transplantation.
Median (50%) survival of patients diagnosed with cirrhosis was 6.8 years.
Liver-related death occurred in 7 (1.7%) patients, and was the third leading
cause of death after malignancy and ischemic heart disease (Table 2.2). Four
(1%) patients died of liver failure, 2 (0.5%) of variceal haemorrhage and one
(0.25%o) of HCC. If patients with cirrhosis on presentation were excluded from
the analysis, then only two liver-related deaths occurred in the remaining 412
patients, although 13 developed cirrhosis. Only one ofthe liver-related deaths
occurred in a patient previously diagnosed with cryptogenic cirrhosis plus the
metabolic syndrome.
58
Table 2.3 Predictors of overall mortality by univariate proportional hazard
modelling
Variable
Age (per year)
Gender (male)
Cirrhosis at baseline
Ischemic heart disease
Smoking
Hypertension
IFG/Diabetes
BMI
Hypertriglyceridemia
Low H D L cholesterol
Metabolic Syndrome
Parameter
Estimate
0.09
0.12
3.32
1.64
0.59
0.60
1.79
-0.01
0.48
0.10
-0.08
Standard
Error
0.01
0.28
' 0.37
0.31
0.28
0.28
0.32
0.02
0.30
0.37
0.28
Hazard Ratio
(95% CI)
1.09(1.01-1.12)
1.13(0.65-1.96
27.6(13.3-57.3)
5.15 (2.83-9.37)
1.81 (1.05-3.12)
1.82(1.06-3.12)
5.97(3.19-11.2)
0.99(0.95-1.04)
1.61 (0.89-2.90)
1.10(0.54-2.27)
0.92(0.53-1.59)
P value
<0.001
0.7
<0.001
O.001
0.03
0.03
O.001
0.8
0.1
0.8
0.8
Table 2.4 Predictors of overall mortality by multivariate Cox proportional hazard
modelling
Variable
Age (per decade)
IFG/Diabetes
Cirrhosis at baseline
Smoking
Hypertension
Ischemic heart disease
Parameter
Estimate
0.08
0.97
1.13
0.35
-0.32
0.08
Standard
Error
0.01
0.34
0.48
0.29
0.29
0.37
Hazard Ratio
(95% CI)
2.2(1.7-2.7)
2.6(1.3-5.2)
3.1 (1.2-7.8)
1.4(0.8-2.5)
0.7(0.4-1.3)
1.1 (0.5-2.3)
P value
O.0001
0.005
0.02
0.2
0.3
0.8
Predictors of Liver Related Mortality
Age, gender, presence of cirrhosis at diagnosis and metabolic risk factors were
examined separately in Cox proportional hazard models for their association with
liver-related death (Table 2.5). Age, cirrhosis and BMI were all significantly
predictive of liver-related death on univariate analysis. As the total number of
liver-related deaths was relatively small, the magnitude of risk associated with
each variable needs to be interpreted with caution, as reflected by the wide
confidence intervals. As the number of events (liver-related deaths) was only
seven, multi-variate analysis was unable to be performed.
Metabolic Complications
During follow-up, 91 (22%) people were diagnosed with diabetes after a mean of
4.6 ± 3.2 years from their diagnosis of NAFLD. Ninety-six (23%) people were
diagnosed with either hypertriglyceridemia and/or low HDL cholesterol 4.7 ± 3.8
years after initial NAFLD diagnosis and 94 (22%) patients were diagnosed as
having hypertension after a mean of 5.2 ± 3.1 years. Thus at the end of follow-up,
203 (47%) ofthe cohort were diabetic, 249 (57%) had hypertension and 359
(83%) had dyslipidemia. These conditions were not routinely screened for in all
patients, and thus the proportion of patients developing metabolic complications
after the diagnosis of NAFLD may be higher.
61
Table 2.5 Predictors of liver-related mortality by univariate proportional hazard modelling
Variable
Age (per year)
Gender (male)
Cirrhosis at baseline
B M I
IFG/Diabetes
Low H D L
Hypertriglyceridemia
Metabolic syndrome
Parameter
Estimate
0.06
0.79
5.38
0.10
6.14
0.63
0.11
1.51
Standard
Error
0.03
0.84
0.87
0.05
4.67
1.00
0.84
1.08
Hazard Ratio
(95% CI)
1.07(0.01-1.12)
2.21 (0.48-15.4)
217(39-1203)
1.10(1.01-1.21)
463 (0.05-4395203)
1.88(0.23-15.7)
1.11 (0.24-7.75)
4.52 (0.54-37.5)
P value
0.01
0.3
O.001
0.03
0.2
0.5
0.9
0.2
Liver Histology
The histological features ofthe 65 patients who underwent liver biopsy (excluding
the 20 post-mortem biopsies) are shown in Table 2.6. Patients with NAFLD who
underwent liver biopsy had a significantly shorter survival compared to those who
did not (Figure 2.4), although no patient died from a liver biopsy complication.
Patients undergoing liver biopsy were more likely to be symptomatic, have
IFG/diabetes or more advanced liver disease as evidenced by laboratory markers
of portal hypertension and impaired hepatic synthetic function (Table 2.7).
Forty-nine patients had NASH, and 17 (35%) of these died, including four
(8%) from liver related causes. Ten patients had bland steatosis without
inflammation, ballooning or fibrosis. None of these patients died from liver
related causes or developed cirrhosis. Overall mortality was lower among patients
with bland steatosis compared to the remaining patients (20% vs. 35%) although
this did not reach statistical significance (p=0.27, log rank test).
Table 2.6 Histological features of patients with NAFLD who underwent liver
biopsy (n=61*).
Fibrosis Stage Inflammation Grade Steatosis Grade
0 22 (36%)
11 (18%)
9 (15%)
11 (18%)
8 (13%)
0 12 (20%)
37 (61%)
12(19%)
1 28 (46%)
20 (33%)
13 (21%)
Ballooning Mallory's Hyaline
0 16 (26%)
35 (58%)
10 (16%)
0 34 (56%)
22 (36%)
5 (8%)
Footnote: Biopsies scored according to schema by Brunt and colleagues
*Biopsies in four patients confirmed hepatocellular steatosis but were of
inadequate size to accurately stage fibrosis.
17
Table 2.7 Baseline clinical and laboratory features of NAFLD patients who
underwent biopsy compared to those who did not (n=420).
Variable
Age (years)
Male
Abdominal pain/fullness
Fatigue
BMI (kg/m2)
Obesity (BMI > 30 kg/m2)
Glucose (mg/dl)
Diabetes (> 126 mg/dl)
Hyperglycemia (>110 mg/dl)
Hypertension (>85 / >135 mmHg)
Triglyceride (mg/dl)
Hypertriglyceridemia (>150mg/dl)
HDL cholesterol (mg/dl)
Low HDL chol. (>40/50 mg/dl)
AST (U/l)
Elevated AST (>31U/1)
Bilirubin (mg/dl)
Elevated bilirubin (>1.0 mg/dl))
Albumin (gm/dl)
Low albumin (<3.5 gm/dl)
Prothrombin (sec.)
Elevated prothrombin (>11.0 sec.)
Platelet count (xl09/l)
Low Platelet count (<150 xl09/l)
Biopsied
(n=65)
50 ±16
27/65 (42%)
25/64 (39%)
7/64 (13%)
32.6 ±7.4
41/64 (64%)
125 ±45
27/65 (42%)
34/65 (52%)
28/65 (43%)
217 ±207
37/59 (63%)
41 ±12
19/48 (40%)
69 ±82
42/64 (66%)
0.9 ±0.8
13/63 (21%)
4.1 ±0.6
6/53 (11%)
10.9 ±1.2
8/46 (17%)
248 ± 84
5 / 64 (8%)
Not Biopsied (n=355)
48 ±15
175/355 (49%)
88/343 (26%)
45/342 (11%)
33.5 ±5.9
253/345 (73%)
117 ±48
81/355 (23%)
120/355 (34%)
122/351 (35%)
229 ±152
225/324 (69%)
42 ±12
98 / 309 (32%)
48 ±43
232/348 (67%)
0.7 ± 0.4
34/324 (10%)
4.3 ±0.4
9/281 (3%)
10.4 ±1.1
11/161 (7%)
245 ± 60
8/338 (2%)
P
value
0.2
0.2
0.03
0.8
0.1
0.1
0.02
0.003
0.005
0.2
0.1
0.4
0.6
0.3
0.1
0.9
0.06
0.03
0.003
0.02
0.03
0.04
0.7
0.04
Footnote: Patients biopsied at post-mortem (n=20) excluded. Patients with
missing data excluded from the denominator.
Not biopsied
i — ~ ~ » - 1 ™ — i ^ — i — i — i — i
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
M * . ,, Time (years) No. at risk Not biopsied 355 336 330 324 307 260 218 188 152 121 89 57 41 32 23 15 8 Biopsied 65 63 59 58 54 46 36 29 24 22 20 14 13 8 8 8 6
Figure 2.4 Survival of patients diagnosed with N A F L D between 1980-1999 in
Olmsted County who underwent liver biopsy compared to those who did not
2.5 DISCUSSION
This study is the first to describe the natural history of N A F L D in a large cohort
of community-based patients. Mortality was significantly increased among
patients with NAFLD compared to the expected mortality ofthe general
population of same age and gender, and was predicted by presence of
IFG/diabetes, cirrhosis and older age. Death occurred in 12.6% of patients and
was most commonly due to malignancy and ischemic heart disease, which are also
the two commonest causes of death in the Minnesota general population. [192]
Liver disease was also an important contributor of death among patients with
NAFLD, being the third most common cause and accounting for 13% of all
deaths. In contrast, 'chronic liver disease and cirrhosis' is the thirteenth leading
cause of death among the Minnesota general population, accounting for less than
one percent of all deaths. Although the causes of death in our cohort were not age
and gender adjusted, this comparison clearly implies that the increased overall
mortality rate among NAFLD patients compared to the general population, was at
least partly due to complications of NAFLD. Nevertheless, the incidence of liver
related death was low (1.7%), as was the occurrence of cirrhosis (5%) and
cirrhosis related complications (3.1%>).
Previous studies examining the natural history of NAFLD originate from
specialist referral centres where patients had been selected to undergo liver
biopsy. [115,116,118] Morbidity and mortality rates in the largest of these studies,
were four to six fold higher than those among our community based patients, most
likely because to selection bias.fl 15] It is pertinent to note that patients selected
for biopsy in our population had significantly lower survival than non-biopsies
patients, reflecting an increased likelihood to biopsy patients with more advanced
disease. It is therefore difficult to extrapolate results of studies of biopsy proven
NAFLD to all NAFLD patients in the community. Thus the prognosis and public
health significance of community diagnosed NAFLD patients, is not as severe as
suggested by findings from referral centres. However, because NAFLD may
affect up to one in three people within the community, [4] the modestly increased
risk of all-cause mortality and low absolute risk of liver related death in these
patients still translates to a significant health burden.
Other hospital based studies ofthe histological subgroup of NAFLD
patients with NASH have documented progression to cirrhosis and HCC, but have
been limited by small numbers and/or average follow-up of less than 5
years.[l 16,118,127] Hui et al. followed 23 patients with cirrhotic stage NASH for
a median of 5 years, and found complications developed in 9/23 (39%) patients
and liver related death in five (22%) patients.[146] Among our cirrhotic patients,
13/21 (62%) developed complications and 7/21 (33%) died from liver related
causes during a median follow-up of 6.8 years. Our patients appeared to present
with more advanced disease, as indicated by lower serum albumin (3.8 ± 0.4 vs.
4.2 ± 0.5 gm/dl) and platelet counts (177 ± 84 vs. 194 ± 87 xl09/L) and higher
bilirubin levels (2.1 + 4.6 vs. 1.1 ± 0.8 mg/dl) than the Australian patients, which
may explain the difference in mortality rate and liver-related death between the
two patient populations.
At the other end ofthe spectrum of NAFLD, patients with a biopsy
demonstrating bland steatosis and no histological evidence of NASH, appear to
haye a benign prognosis.[114,117] A Danish cohort of 109 predominantly
morbidly obese subjects followed for nearly 17 years found the incidence of
cirrhosis to be less than 1%> with no liver related deaths recorded. [114] Although
not directly compared, the survival curve ofthe general population fell within the
95%) confidence interval ofthe survival curve of patients with steatosis. Given
our study design, we were unable to stratify risk of death in all patients according
to histology. However ofthe 10 patients in our study with biopsy diagnosed
bland steatosis, none developed cirrhosis and only two died, neither from liver
related causes.
In the present study, overall mortality was increased among NAFLD
patients compared to the general population ofthe same gender and age and was
associated with IFG/diabetes. Diabetes and IFG are recognized risk factors for the
three most common causes of death in our cohort; ischemic heart disease[193],
malignancy [194,195] and liver disease. [196,197] Thus insulin resistance which
results in IFG/diabetes, is important not only as a pathogenic factor for NAFLD,
but also as a risk factor for death in NAFLD patients. Therefore a diagnosis of
NAFLD should mandate screening for diabetes and glucose intolerance.
Subsequent aggressive management of insulin resistance may potentially reduce
the risk of mortality and morbidity among these patients with further research into
this area warranted. [198-200]
Cirrhosis due to NAFLD was also an independent risk factor for overall
death, conferring a three-fold increased risk. Therefore treatment of NAFLD to
prevent progression to cirrhosis may also decrease mortality in these patients.
Although the incidence of cirrhosis was relatively low (3%) and thus treatment of
all NAFLD patients would benefit only a few in terms of preventing cirrhosis,
treatment of NAFLD aimed at improving insulin resistance may also have
beneficial effects in decreasing the risk of death from cardiovascular
disease. [199,200]
Obesity or the metabolic syndrome as independent variables, were not risk
factors for overall death in our cohort. Obesity was present in the majority of our
patients, increasing the difficulty to detect a difference in survival. Studies are
conflicting as to whether the metabolic syndrome increases risk of all cause death
in the general population, however diabetes per se is consistently reported to
increase mortality,[201-203] which is in agreement with our findings.
Interestingly, the association between the metabolic syndrome and insulin
resistance is variable. In particular, the metabolic syndrome is relatively
insensitive for detecting subjects with insulin resistance who are not diabetic. [204]
Certain components ofthe metabolic syndrome such as diabetes are more closely
related to insulin resistance than others such as hypertension and
dysplipidemia.[204] Thus variables within the metabolic syndrome which are
more closely correlated with insulin resistance appear to be more critical in
determining prognosis among subjects with NAFLD.
Liver related death occurred in 1.7% ofthe cohort and was associated on
univariate analysis with age, cirrhosis and BMI. Obesity is a recognised risk
factor for cirrhosis related death or hospitalization among non-alcoholic
individuals,[196] as well as being a risk factor for advanced fibrosis among
patients with NASH.[128] Obesity may promote progressive hepatic fibrosis via
metabolic changes associated with insulin resistance such as reduced hepato-
protective adiponectin,[205] increased hepato-toxic FFA's[84] or up-regulation of
pro-fibrogenic factors.[185] Obesity may also increase the susceptibility to liver
related complications such as spontaneous bacterial peritonitis. [206] It should be
noted that the relatively small number of liver related deaths may have prevented
identification of other significant prognostic variables. In addition, potential
confounders cannot be excluded due to the inability to perform multi-variable
analysis.
Due to the retrospective nature of this study, a limitation is the inability to
detect all cases of NAFLD within Olmsted County. Initial inclusion in the study
required a physician diagnosis of 'fatty liver'. Although more than 90%) of the
Olmsted County population are assessed by a health care provider within a three
year periodf 189],thereby maximizing potential for diagnosis, it is likely that
increased physician awareness and utility of ultrasound led to increased
ascertainment, as reflected by the increase in incidence over the two decade study
period and the higher incidence of cirrhosis in the first decade. In addition, we
required confirmation by imaging or biopsy to maximize diagnostic certainty. As
the entire County did not undergo imaging or liver biopsy, our cohort is likely to
be biased towards patients with clinically relevant disease or with greater co
morbidities who are more likely to be investigated, potentially leading to an over-
estimation of morbidity or mortality risk. Despite this, many patients in this
cohort were diagnosed not in a specialist hepatology setting, but by community
practitioners. Diagnosis was often made incidentally during investigation of other
conditions or during annual health checks in which aminotransferases are
routinely performed, although it is unknown what proportion of subjects
undergoing routine check-ups underwent imaging studies or had serum
aminotransaminases measured. Thus although our cohort represents patients with
NAFLD diagnosed in the community rather than all NAFLD patients within the
community, this study is likely to represent the community burden and natural
history of N A F L D more accurately than any previous reports. It should also be
noted however, that the majority of our population was Caucasian reflecting the
ethnic composition in Olmsted County, Minnesota. It remains unclear if
morbidity and mortality rates in our study can be generalized to other racial
groups.
In summary, patients in the community diagnosed with NAFLD have a
modest increased risk of death compared to the general population.
Approximately one in 30 patients may develop cirrhosis or a liver related
complication. Liver cirrhosis and IFG/diabetes are risk factors for overall and
liver-related death among patients with NAFLD and therefore should be identified
and vigorously managed where possible. Although overall mortality, rate of liver-
related death and the development of cirrhosis in community-based patients is
strikingly less than that seen in referral patients, the potential community burden
of NAFLD remains considerable because ofthe high prevalence ofthe disease.
CHAPTER THREE
THE HISTOLOGICAL EVOLUTION OF NAFLD
3.1 INTRODUCTION
The previous chapter demonstrated that subjects diagnosed with NAFLD have an
increased risk of all-cause mortality, which was at least partly attributable to liver
disease with cirrhosis being a predictive factor for death. The incidence of
cirrhosis over the follow-up period was 3.1%, however it is unknown what stage
of disease these patients had at study entry. Thus the time course of fibrosis
progression remains unclear. Furthermore, the relatively small number of liver-
related deaths hindered identification of important prognostic variables.
Chronic hepatitis typically has a prolonged time course, which generally
takes decades to manifest as morbidity and mortality. Fibrosis stage, as
determined by liver biopsy, is recognized as the most objective indicator of liver
damage and is the best prognostic marker for morbidity and mortality in liver
disease of various etiologies. Studying the change in fibrosis stage over time
allows valuable insight into the sub-clinical course of liver disease. Few studies
have investigated the natural history of NAFLD by examining fibrosis stage
among patients with paired liver biopsies, and the largest series included only 22
patients (reviewed in Chapter One).[l 16-118,125-127,131] Due to small numbers,
conclusions remain limited. It is uncertain what proportion of patients with
NAFLD have progressive disease. In addition, the rate of disease progression or
change in other liver histological features over time is unknown. Thus it remains
unclear whether some factors predict higher rates of progression.
3.2 A I M S
To determine in a large number of patients with N A F L D ;
1) the proportion and rate of fibrosis change over time
2) the proportion and rate of change of other histological features over time
3) whether routinely determined clinical, laboratory or histological features
predict fibrosis progression or rate.
This study has been published in part (Adams et al, J Hepatol. 2005; 42: 132-
138).
3.3 METHODS
3.3.1 Case Ascertainment
The computerized pathology database ofthe Mayo Clinic has diagnoses indexed
and coded using the HICDA system. The database was searched using the
diagnostic codes for NAFLD as inclusion criteria and codes for other liver disease
as exclusion criteria (see Appendix One). A total of 1097 patients were identified
between the time-period 1985 to 1999. To determine the number of liver biopsies
each patient had undergone, the medical notes of each of these individuals were
examined. Patients who had undergone more than one liver biopsy, either at the
Mayo Clinic or at other institutions, had their medical records examined further to
determine whether they fulfilled the inclusion and exclusion criteria ofthe study
(see below).
Inclusion into the study required a diagnosis of NAFLD based upon: 1)
steatosis involving at least 10% of hepatocytes on biopsy, 2) ethanol consumption
of less than 140 grams per week. Ethanol consumption was prospectively
evaluated at each patient clinic visit by a patient history form detailing the type,
amount, frequency and duration of alcohol intake as well as screening for alcohol
abuse. These forms were reviewed by a nurse and physician at each patient visit.
Patients with evidence of other liver disease using standard clinical,
laboratory and histological criteria were excluded. Specifically, patients were
excluded if they had any ofthe following; hepatitis C antibody or RNA, hepatitis
B surface antigen, low alpha-one antitrypsin level or ZZ phenotype, low
ceruloplasmin level, positive anti-nuclear antibody or anti-smooth muscle
antibody with histological evidence of auto-immune hepatitis, positive anti-
mitochondrial antibody with histological features of primary biliary cirrhosis,
elevated ferritin or transferrin saturation with positive HFE genotype (C282Y
homozygous or H63D/C282Y compound heterozygous) or histological evidence
of iron overload. Patients were also excluded if they had a secondary cause of
NAFLD (see Table 1.1). One patient had serial biopsies prior to 1990 without
hepatitis C serology. He had no hepatitis C risk factors and his histology was not
compatible with hepatitis C infection and thus was included.
One hundred and twenty patients who had undergone serial liver biopsies
were identified as fulfilling the inclusion and exclusion criteria. Of these, 47 had
at least one of their liver biopsies performed at an outside institution. These
patients were sent a letter requesting authorization to obtain their liver biopsies for
review (see Appendix Three). A follow-up phone call was made if no response
was received. Eleven patients were unable to be contacted or refused permission.
Biopsies of six patients who had provided authorization, were unable to be located
or had been destroyed. The remaining 103 patients formed the study group.
3.3.2 Patients
Patient medical records were extensively reviewed to abstract data at the time of
the first and last liver biopsy. All patients had undergone a complete medical
history, physical examination and laboratory testing as part of their medical
evaluation. Data on the following clinical characteristics were collected; age,
gender, height, weight, blood pressure, medical history of hypertension,
dyslipidemia or diabetes and medication history. Laboratory parameters collected
included AST, ALT, bilirubin, alkaline phosphatase, albumin, prothrombin time,
platelet count, fasting lipids (total cholesterol, HDL cholesterol, triglyceride) and
fasting glucose. Fasting insulin levels were available for 40 individuals. Insulin
sensitivity was calculated in these individuals using the HOMA and quantitative
insulin-sensitivity check index (QUICKI) which correlate closely with the
euglycemic hyperinsulin clamp method of determining insulin
sensitivity.[207,208] The formula used to calculate the HOMA score was;
HOMA=(insulin*glucose)/22.5; and the formula used for the QUICKI was;
QUICKI=l/[log(insulin) + log(glucose)]. The metabolic syndrome was defined in
the study subjects at time of initial liver biopsy using the criteria proposed by the
National Cholesterol Education Program (ATP III), i.e., when at least three ofthe
five following features were present: [15 ^hyperglycemia (fasting blood glucose
>110 mg/dl, or known diabetes mellitus), hypertension (blood pressure >130/>85
or under treatment), hypertriglyceridemia (>150 mg/dl or under pharmacological
treatment), low-HDL cholesterol (<40 mg/dl for males and <50 mg/dl for
females), and obesity (BMI >30 kg/m2). As waist circumference was not
measured for most ofthe patients, a BMI >30 kg/m2 was used instead to define
obesity.[190]
Ofthe 103 patients, a repeat liver biopsy was performed as part of their
medical follow-up in 26 patients and as part of a clinical trial in 77. The additional
inclusion criteria required for the trial patients were; age 18-75 years, treatment
with UDCA or chenodeoxycholic acid in the 3 months prior to the study;
anticipated need for transplantation within 1 year or recurrent variceal bleeding,
spontaneous porto-systemic encephalopathy, diuretic-resistant ascites, or bacterial
peritonitis; pregnancy or lactation; persistent elevation of serum alanine
aminotransferase (ALT) or aspartate aminotransferase (AST) at least 1.5 times the upper
limits of normal for at least 3 months. The clinical indications for the repeat liver
biopsy in the 26 patients were; liver transplant explant (n=6), increased liver
enzymes (n=6), suspicion of cirrhosis (n=5), taken at laparoscopic
cholecystectomy (n=3), lymphoma staging (n=l), investigation of pyrexia of
unknown origin (n=l) and unknown (n=4). The 77 patients in clinical trials
included 50 participants in a placebo-controlled trial (27 randomized to placebo
and 23 to ursodeoxycholic acid),[209]and 27 participants in a pilot study of
clofibrate or ursodeoxycholic acid.[210] Whilst all overweight and obese patients
in the trials were "encouraged to lose weight", there was no formal dietician
review, exercise or lifestyle program. Both liver enzymes and histology were
unchanged after a year of treatment with clofibrate and changes in liver enzymes
and histological features were identical among patients treated with
ursodeoxycholic acid or placebo for two years. In addition, patients who received
pharmacotherapy as part of a clinical trial had the same change in histological
features and rate of disease progression as those who did not receive
pharmacotherapy (Table 3.1). Similarly, those biopsied for clinical reasons had
the same histological course as those biopsied as part of a clinical trial (Table 3.2).
77
Table 3.1 Change in liver histology was not different between patients on drug
treatment compared to untreated patients
Treated Untreated
(Ursodiol/Clofibrate) (Placebo/Clinical) value
N
Fibrosis Stage
Change between biopsies
Regressors /Stable /Progressors
Steatosis Grade
Change between biopsies
Regressors /Stable /Progressors
Inflammation Grade
Change between biopsies
Regressors /Stable /Progressors
Ballooning Grade
Change between biopsies
Regressors /Stable /Progressors
50
0.04+1.23
30/36/34
-0.4 ±0.8
50/38/12
-0.2 ± 0.7
28/60/12
-0.3 ± 0.6
33/63/4
53
0.34 ±1.43
28/32/40
-0.4 ±1.0
46/42/12
-0.1 ±0.7
23/62/15
-0.1 ±0.6
25/61/14
0.9
0.8
0.9
0.9
0.5
0.8
0.2
0.2
Footnote: Change in biopsies presented as mean ± SD; Regressors /Stable
/Progressors presented as percentage
78
Table 3.2 Change in liver histology was not different between patients enrolled
in clinical trials compared to patients biopsied for clinical indications
Trial Patients Non-Trial Patients P value
(Urso/Clofibrate) (Placebo/Clinical)
N 77 26
Fibrosis Stage
Change between biopsies
Regressors /Stable /Progressors
Steatosis Grade
Change between biopsies
Regressors /Stable /Progressors
Inflammation Grade
0.10 ±1.22
30/36/34
-0.5 ±0.8
49/41 /10
0.38 ±1.56
27 / 27 / 46
-0.3 ±1.0
48/36/16
Footnote: Change in biopsies presented as mean ± SD; Regressors /Stable
0.5
0.5
0.1
0.7
Change between biopsies
Regressors /Stable /Progressors
Ballooning Grade
Change between biopsies
Regressors /Stable /Progressors
-0.2 ± 0.7
28/60/12
-0.3 ± 0.6
33/61/6
0.0 ±0.6
16/65/19
0.0 ±0.6
20/64/16
0.1
0.3
0.1
0.1
/Progressors presented as percentage
Therefore, as neither medication (clofibrate nor ursodeoxycholic acid) or
the indication for liver biopsy affected liver histology, patients were pooled
together for the purpose of this study. The time elapsed from first to last biopsy
was not significantly different (p=0.2) between the 77 participants in a clinical
protocol and the 26 patients who had biopsies as part ofthe standard evaluation.
The study was approved by the Mayo Institutional Review Board and all patients
gave written informed consent for participation in medical research.
3.3.3 Liver Histology
Liver biopsy specimens were read under coded identification by a single liver
pathologist (Dr S. Sanderson, Department of Anatomical Pathology, Mayo Clinic)
who was unaware ofthe patient details or biopsy sequence. Biopsies were
routinely stained with hematoxylin-eosin, and Masson's trichrome. All biopsies
were a minimum of 15mm in length and had at least six portal tracts to make a
confident evaluation ofthe histological features and diagnosis. [211] Histological
features were interpreted according to the schema outlined by Brunt et al.[101]
Briefly, steatosis was graded on a 4-point scale: grade 0 = no steatosis; grade 1 =
steatosis involving <33% of hepatocytes, grade 2 = 33-66%», grade 3 >66%. The
severity of lobular inflammation was graded on a 4-point scale: grade 0 = no or
negligible inflammation, grade 1 = mild, grade 2 = moderate, grade 3 = severe.
Fibrosis was staged on a 5-point scale: stage 0 = no fibrosis, stage 1 = zone 3
perisinusoidal/ perivenular fibrosis, stage 2 = zone 3 and periportal fibrosis, stage
3 = septal/bridging fibrosis, stage 4 = cirrhosis. In addition, the following
histological features were scored: hepatocellular ballooning (0 = absent, 1 = mild,
2 = marked); Mallory's hyaline (0 = absent, 1 = occasional, 2 = several); and
hepatocellular iron (0-4+ as per Searle).[212] Severity of lobular inflammation,
hepatocellular necrosis, portal tract inflammation, pericellular fibrosis, portal
fibrosis, and bridging fibrosis were also recorded and scored as described by
Brunt etal.[101]
NASH was defined as either the presence of steatosis/?/^ mixed lobular
inflammation plus hepatocellular ballooning, as proposed during the AASLD
single topic conference,[3]or the presence of steatosis plus any stage of fibrosis.
Steatosis plus either lobular inflammation or ballooning (but not both) was termed
'steatosis with non-specific inflammation', whereas steatosis without lobular
inflammation, ballooning or fibrosis was termed 'bland steatosis'.
3.3.4 Statistical Analysis
Patients were divided into groups according to the change in fibrosis stage
between biopsies; either 'progressors' (increased in fibrosis stage), 'stable' (no
change) and 'regressors' (decreased in fibrosis stage). Fibrosis rate was
calculated by dividing the difference in fibrosis stage between first and last
biopsy, by the time between biopsies in years. Categorical data are presented as
number (percentage). Continuous data are presented as mean ± standard deviation
(SD), and medians (range). Continuous variables were compared using Students t
test and Mann-Whitney U test when appropriate. Frequency data were compared
using chi-squared test or Fisher's exact test where appropriate. The McNemar's
test was used to compare paired proportions. Spearman's rank correlation
coefficient was used as a measure of association. Predictors ofthe rate of
progression were assessed using univariate and multivariate linear regression
analysis. A two tailed p value of less than 0.05 was considered statistically
81
significant. Analyses were performed using Systat 10.1. (Systat Software Inc.
Richmond CA.)
3.4 RESULTS
3.4.1 Patient Characteristics
The 103 patients underwent a total of 227 biopsies, from which only the initial
and final biopsies were examined. The mean time interval between first and last
biopsy was 3.2 + 3.0 years (range 0.7-21 years). Baseline clinical features ofthe
patient population are summarized in Table 3.3. Approximately two thirds ofthe
cohort were female. Metabolic risk factors were common with approximately half
ofthe patients fulfilling diagnostic criteria for the metabolic syndrome.
Biochemical and haematological parameters are shown in Table 3.4. The
majority of patients had elevated aminotransaminase levels although evidence of
synthetic dysfunction was uncommon. Abnormal lipid profiles were observed in
approximately two thirds ofthe cohort with one third demonstrating impaired
fasting glycemia.
Histological features on initial biopsy are shown in Table 3.5. The
distribution of steatosis grade and fibrosis stage was relatively even throughout
the cohort. Sixteen subjects were cirrhotic on initial biopsy. The majority of
subjects had low grade inflammation and hepatocellular ballooning which is
typical of NASH.[101] Mallory's bodies were less common, being present in just
over one third. Ninety-six patients fulfilled criteria for NASH, 4 had steatosis
with non-specific inflammation and 3 had bland steatosis.
Table 3.3 Clinical features at time of initial liver biopsy (n = 103)
Variable Mean ± SD or n (%)
Age (years) 45 ± 11
Sex (female) 65 (63%)
BMI (kg/m2) 32.7 ±5.5
Obese 69 (67%)
Hypertensive 31 (30%)
Diabetic 42 (43%)
Hypertriglyceridemia 61 (59%)
Low HDL cholesterol 67 (65%)
Metabolic Syndrome 55 (53%)
83
Table 3.4 Laboratory features at time of initial liver biopsy
Variable
A L T (U/L)
A S T (U/L)
A S T / A L T ratio
Bilirubin (mg/dL)
Albumin (gm/dL)
Platelets (xl09/L)
Prothrombin (sec.)
Ferritin (|ig/L)
Glucose (mg/dL)
QUICKI
HOMA
Triglyceride (mg/dL)
H D L cholesterol (mg/dL)
Mean ± SD
or n (%)
97 ±56
75 ±50
0.9 ±0.5
0.8 ±0.5
4.3 ±0.5
211 ±76
10.9 ±1.1
280 ± 275
119 + 40
0.33 + 0.04
4.23 + 3.87
232 + 233
41 + 11
Median (range)
82(21-373)
60(13-258)
0.8(0.3-3.6)
0.7(0.2-3.2)
4.4(2.6-5.4)
212(45-423)
10.8(8.4-14.3)
211(6-1478)
102(78-273)
0.33(0.27-0.46)
2.93(0.36-14.13)
179(62-2085)
40(12-65)
Abnormal*
n (%)
89/96 (93%)
86/97 (89%)
26/96 (27%)
14/100(14%)
4/100(4%)
25/102(24%)
15/98(15%)
32/72 (44%)
37/101 (37%)
61/101 (60%)
66/96 (69%)
*Footnote: Normal ranges: A L T < 29 U/L females, < 45 U/L males;
AST/ALT>1; Bilirubin <1.2 mg/dl; Albumin 3.5-5.0 gm/dL; Platelets 150-400
xl09/L; Prothrombin 8.4-12.0 seconds; Ferritin <150 iag/L females, <300 |ag/L
males; Fasting Glucose < 1 lOmg/dl; Triglyceride <150 mg/dl; H D L cholesterol
>40mg/dl females, >50mg/dl males.
Table 3.5 Histological features at initial liver biopsy (n=103)
Fibrosis Stage
0 25 (24%)
1 21 (20%)
2 23 (22%)
3 18(18%)
4 16(16%)
Ballooning
0 8 (8%)
1 86 (83%)
2 9 (9%)
Inflammation Grade
0
1
2
3
9 (9%)
72 (70%)
20 (19%)
2 (2%)
Mallory's Hyaline
0
1
2
62 (60%)
32(31%)
9 (9%)
Steatosis Grade
1
2
3
40 (39%)
31 (30%)
32(31%)
Hepatocellular Iron*
0
1
2
77 (85%)
7 (8%)
6 (7%)
Footnote: Numbers (%) of patients shown. * Iron stains not available for 13
patients. No patients had iron grade three or four.
3.4.2 Fibrosis Progression
A total of 38 patients (37%) had progressive fibrosis during the time period
between the first and last biopsy; 35 patients (34%) did not change in fibrosis
stage and 30 (29%) regressed (Table 3.6). Fourteen patients (14%) progressed by
two stages or more and 4 patients progressed three stages or more. Nine patients
progressed to cirrhosis, with two of these having no fibrosis on initial biopsy.
Eight patients regressed by two or more stages, with two patients regressing three
stages. One patient had cirrhosis on initial biopsy and stage two on their
subsequent biopsy.
Sampling error could be expected to cause patients to progress as frequently
as they regress. However, this may be skewed by the proportion of patients with
stage F0 fibrosis who cannot 'regress' in fibrosis stage, or patients with F4
cirrhosis who cannot 'progress'. However, after excluding patients with F0 of F4
fibrosis, the proportion of patients who had progressive fibrosis remained similar
at 42% (26/62).
The likelihood of observing progressive fibrosis increased as the duration
between biopsies increased. One quarter (24%) of all patients biopsied more than
four years apart progressed by two or more stages, compared to 11% of patients
biopsied within four years. Two thirds (67%) of early stage patients (stage 0-2)
biopsied more than four years apart had progressive fibrosis, compared to 40% of
patients biopsied within two years (Table 3.7). Patients progressing three or more
stages had a significantly longer biopsy interval compared to the remaining
patients (12 ± 8 vs. 3 ± 2 years, p=0.003). In particular, two patients progressed
from no fibrosis to cirrhosis over 9.2 and 15.4 years each.
86
Table 3.6 Change in fibrosis stage between first and last biopsy (n = 103)
Final Fibrosis Stage
Initial
Fibrosis
Stage
F0
Fl
F2
F3
F4
F0
13
7
4
2
0
Fl
5
2
4
1
0
F2
3
.8
5
6
1
F3
2
4
7
5
5
F4
2
0
3
4
10
Footnote: Patient number in boxes. Lightly shaded boxes represent patients who
did not change in fibrosis stage between liver biopsies. Patients to the right ofthe
shaded boxes had progressive fibrosis. Patients to the left ofthe shaded boxes
regressed.
Table 3.7 Progression of fibrosis stage according to initial fibrosis stage and time
interval between biopsies
Biopsy Interval
< 24 Months 24 - 48 Months > 48 Months
Total Cohort (n = 103)
Progressors 11 (34%) 18(36%)
Stable 13 (41%) 13 (26%)
Regressors 8 (25%) 19 (38%)
Initial Stage 0 - 2 (n = 69)
Progressors
Stable
Regressors
Initial Stage 3 - 4 (n = 34)
Progressors 1 (14%) 2(11%) 1 (12%)
Stable 4(57%) 5(28%) 6(67%)
Regressors 2(29%) 11(61%) 2(22%)
9 (43%)
9 (43%)
3 (14%)
10 (40%)
9 (36%)
6 (24%)
16 (50%)
8 (25%)
8 (25%)
8 (67%)
3 (25%)
1 (8%)
3.4.3 Predictors of Fibrosis Progression
No demographic, clinical, metabolic or biochemical variables identified at the
time of initial biopsy differed significantly between patients who regressed,
remained stable or progressed in fibrosis stage (Table 3.8). Similarly, the
development of diabetes or weight gain between biopsies did not differ between
groups. Overall, BMI did not change significantly between the first and last liver
biopsy (32.7 ± 5.5 vs. 32.9 ± 5.9 kg/m2).
Ofthe baseline histological variables, only fibrosis stage was significantly
(inversely) associated with fibrosis progression (P = 0.003). The proportion of
NASH patients who progressed was not significantly different from patients
without NASH (34.4% vs. 53.8%, P = 0.2). Similar proportions of patients with
progressive fibrosis were observed between those with and without progressive
inflammation (50% vs. 35%, P = 0.4), progressive steatosis (25% vs. 38%, P =
0.5), and progressive ballooning (36% vs. 44%, P = 0.7).
The effect of alcohol and medications on liver histology was also examined.
One patient was taking metformin before the initial biopsy and continued on the
same dosage until the final biopsy. No patients were taking thiazolidinediones or
vitamin E. One patient was noted to increase her alcohol consumption from an
average of 10 gm/day to an average 30-40 gm/day between her two biopsies,
although her histological features remained essentially unchanged.
89
Table 3.8 Patients with progressive, stable or regressive fibrosis (n = 103)
n (%)
Age (years)
Sex (female)
Obese
Weight gain
Diabetes
Development of Diabetes
Hypertension
Hypertriglyceridemia
Low HDL
Metabolic Syndrome
ALT (U/L)
AST (U/L)
AST / ALT ratio
Bilirubin (mg/dL)
Albumin (gm/dL)
Platelets (xl09/L)
Prothrombin (sec.)
Ferritin (ng/L) Glucose (mg/dL)
QUICKI HOMA Triglyceride (mg/dL)
H D L (mg/dL)
Steatosis Grade •
Inflammatory Grade
Ballooning
Mallory's Hyaline
Iron Stain
Fibrosis Stage
Regressors
30 (29%)
46 ±2
20 (66%)
19(63%)
14 (47%)
8 (27%)
2 /22 (9%)
9 (30%)
20 (67%)
20 (74%)
16(53%)
91 ±57
80 ±71
1.0 ±0.8
0.7 ±0.1
4.4 ±0.1
206 ± 67
11.0±1.4
214 ±169
117 ±41 0.32 ±0.03 3.82 ±2.48
220 ±149
42 ±9
1.9 ±0.9 2(1-3)
1.3 ±0.6 1 (1-3)
1.1 ±0.4 1 (0-2)
0.6 ± 0.7 0 (0-2)
0.04 ± 0.04 0(0-1)
2.5 ±1.1 2.5 (1-4)
Stable
35 (34%)
47 ±2
19 (54%)
22 (65%)
21 (60%)
15 (43%)
2/20(10%)
10 (29%)
22 (65%)
24 (71%)
20 (57%)
94 ±51
74 ±44
0.9 ±0.4
0.9 ±0.6
4.2 ± 0.5
197 ±72
10.8 ±1.0
258 ±194
125 ±45
0.33 + 0.04 4.15 ±3.78
238 ±163
40 ±11
1.7 ±0.8 1 (1-3)
1.1 ±0.6 1 (0-2)
1.0 ±0.4 1 (0-2)
0.5 ±0.6 0 (0-2)
0.07 ± 0.26 0(0-1)
1.9+1.7 2 (0-4)
Progressors
38 (37%)
44 + 2
25 (66%)
28 (74%)
25 (66%)
20 (53%)
4/19(21%)
12 (32%)
19(51%)
23 (66%)
19(50%)
104 ±60
71 ±34
0.8 ±0.3
0.8 ±0.4
4.4 ±0.5
228 ± 85
10.9 ±1.0
354 + 381
113 ± 33
0.35 ±0.05 2.78 ±1.71
237 ±330
42 ±13
2.2 + 0.8 2(1-3)
1.1 ±0.5 1 (0-2)
1.0 ±0.3 1 (0-2)
0.3 ±0.6 0 (0-2)
0.27 ±0.14 0 (0-2)
1.2 ±1.0 1 (0-3)
P
*
0.4 0.5
0.4
0.8
0.1
0.5
0.8
0.2
0.6
0.8
0.4 0.8
0.8 0.3 0.4 0.2*
*
0.8 0.8 0.4 0.5 0.5
0.2 0.6*
0.1
0.2
0.1
0.1
0.9
0.003
Footnote: Clinical and laboratory measurements recorded at time of initial biopsy. Weight gain and development of diabetes documented between first and last biopsy. Data presented as number
(percentage), mean ± standard deviation or median (range) if non-parametric. 'Continuous variables normally distributed and thus analysed by A N O V A ; Remaining data were skewed, thus
analysed by Kruskal-Wallis test.
3.4.4 Rate of Fibrosis Progression
To account for different time periods between biopsies, the rate of fibrosis change
was calculated by dividing the difference in fibrosis stage between biopsies by the
time interval (in years) between biopsies. Rate of fibrosis change varied from -
2.05 to 1.70 stages/year with an overall mean rate of 0.02 ± 0.66 stages/year.
Seventeen patients progressed at greater than 0.5 stages/year and 13 patients
regressed by more than 0.5 stages/year. When patients with cirrhosis were
excluded (as they cannot progress), the rate of fibrosis change was 0.09 ± 0.67
stages/year.
3.4.5 Predictors of Rate of Fibrosis Progression
By univariate linear regression analysis with rate of fibrosis change as the
dependent variable, diabetes (P = 0.01), AST/ALT ratio (P = 0.02), steatosis grade
(P = 0.05) and fibrosis stage (P = 0.003) were the only significant variables (Table
3.9). Similarly, rate of fibrosis change was not significantly different between
patients with or without NASH (0.014 ± 0.69 vs. 0.19 ± 0.20 stages/year
respectively, P = 0.3). As subjects with cirrhosis cannot progress in fibrosis stage
between biopsies and patients without fibrosis cannot regress on subsequent
biopsy, analysis ofthe influence of fibrosis stage on fibrosis rate was repeated
after excluding these two groups. Fibrosis rate remained significantly associated
with initial fibrosis stage when cirrhotics were excluded (regression coefficient = -
0.18, standard error = 0.06, P = 0.005) as well as when cirrhotics and those
without fibrosis were excluded (regression coefficient = -0.22, standard error =
0.11,P = 0.03).
91
Table 3.9 Predictors of fibrosis rate by univariate regression analysis
Variable
Age (years)
Sex (male=l)
Obese
BMI
Diabetes
Hypertension
Hypertriglyceridemia
Low HDL
Metabolic Syndrome
ALT (U/L)
AST (U/L)
AST / ALT ratio
Bilirubin (mg/dL)
Albumin (gm/dL)
Prothrombin (sec.)
Ferritin (pig/L)
Glucose (mg/dL)
QUICKI
HOMA
Triglyceride (mg/dL)
H D L (mg/dL)
Steatosis Grade
Inflammatory Grade
Ballooning
Mallory's Hyaline
Iron Stain
Fibrosis Stage
Regression
Coefficient
-0.003
-0.10
0.03
0.007
0.38
-0.002
-0.04
-0.15
0.07
0.001
-0.001
-0.30
0.16
0.08
0.01
0.000
-0.001
1.20
-0.01
0.000
0.000
0.16
0.01
-0.14
-0.10
0.04
-0.17
Standard Error
0.006
0.12
0.14
0.012
0.13
0.14
0.14
0.15
0.13
0.001
0.001
0.13
0.14
0.13
0.06
0.000
0.002
2.21
0.03
0.000
0.006
0.08
0.11
0.16
0.10
0.12
0.04
P value
0.6
0.4
0.8
0.6
0.01
0.9
0.8
0.3
0.6
0.3
0.3
0.02
0.3
0.5
0.9
0.4
0.4
0.6
0.9
0.9
0.9
0.048
0.9
0.4
0.4
0.8
O.001
Variables significantly associated with fibrosis rate on univariate analysis
were then analysed in a multivariate model adjusted for age and BMI. By
multivariate linear regression analysis, only presence of diabetes and earlier
fibrosis stage were significantly associated with a higher rate of fibrosis
progression. When cirrhotics were excluded, the same variables (diabetes and
fibrosis stage) plus BMI remained significant (Table 3.10). The addition of
'ursodeoxycholic acid' as a co-variate did not change the results ofthe
multivariate analysis, further indicating a lack of effect of that drug on fibrosis
change.
3.4.6 Change in Aminotransferase Levels
There was a significant decrease in ALT (97 ± 56 vs. 74 ± 61 U/L, p<0.001) and
AST (75 ± 50 vs. 56 ± 51 U/L, P< 0.001) between first and last biopsies.
However, comparing patients with progressive fibrosis to those without, there was
no significant difference in the magnitude of change in ALT level (-13 ± 63 vs. -
29 ± 55 U/L respectively, P = 0.7) or AST level (-10 ± 63 vs. -24 ± 51 U/L
respectively, P = 0.3).
Improvement in ALT correlated with improvement in Mallory's hyaline
score (r = 0.25, P = 0.02) and overall inflammatory grade (r = 0.29, P = 0.006).
Improvement in AST correlated with improvement in steatosis grade (r = 0.3, P =
0.004), Mallory's hyaline (r = 0.3, P = 0.004), lobular inflammation (r = 0.27, P =
0.02) and overall inflammatory grade (r = 0.39, PO.001). Neither change in ALT
nor AST correlated with change in fibrosis stage (r = 0.07, P = 0.5 and r = 0.15, P
= 0.15 respectively).
93
Table 3.10 Predictors of fibrosis rate by multivariate linear regression analysis
-
Total
Cohort
(n = 103)
Excluding
Cirrhotics
(n = 87)
Variable
AST/ALT ratio
Age
Steatosis Grade
B M I
Diabetes
Fibrosis Stage
AST/ALT ratio
Steatosis Grade
Age
BMI
Diabetes
Fibrosis Stage
Regression
Coefficient
-0.15
0.01
0.12
0.02
0.35
-0.20
-0.16
0.10
0.01
0.04
0.39
-0.22
Standard
Error
0.12
0.01
0.08
0.01
0.13
0.05
0.13
0.08
0.01
0.01
0.01
0.06
P value
0.2
0.2
0.1
0.09
0.007
<0.001
0.2
0.2
0.1
0.008
0.005
0.001
3.4.7 Change in Other Histological Parameters
Whereas fibrosis tended to progress, there was a significant overall reduction in
severity of steatosis, inflammation, ballooning of hepatocytes and Mallory's
hyaline, between the first and last biopsy (Figure 3.1). Among non-cirrhotics, a
significant decrease in steatosis (2.0 ± 0.8 vs. 1.6 ± 0.8, PO.001), and ballooning
(1.0 ± 0.4 vs. 0.8 ± 0.4, P = 0.008) occurred between biopsies with a trend for a
decrease in inflammatory stage (1.1 ± 0.6 vs. 1.0 ± 0.5, P = 0.09). Analysis of
cirrhotic patients also revealed a significant decrease in steatosis (1.4 ± 0.6 vs. 1.0
± 0.5, P = 0.008) and ballooning (1.2 ± 0.5 vs. 0.6 ± 0.7, P = 0.01), with a trend
for inflammation (1.3 ± 0.7 vs. 1.0 ± 0.6, P = 0.096).
Ofthe three patients with bland steatosis on initial biopsy, two developed
NASH and one remained unchanged. None developed fibrosis. All ofthe 4
patients with 'steatosis and non-specific inflammation' fulfilled diagnostic criteria
for NASH on the second biopsy. Ofthe 96 original NASH patients, four lost
evidence of steatosis, four developed 'steatosis with non-specific inflammation'
and two developed bland steatosis. Ofthe four NASH patients who lost evidence
of steatosis, three lost between 2.7-12.5 kg between biopsies and one gained 3.1
kg-
95
Steatosis Ballooning Mallory Inflammation
Figure 3.1 Histological features at initial and final biopsy.
There was a significant improvement in severity of steatosis, ballooning,
Mallory's hyaline and inflammation between first (dark bars) and last (clear bars)
liver biopsies (P < 0.05 for all comparisons). Mean and standard error ofthe mean
shown.
3.5 DISCUSSION
In this study, which represents the largest reported series of N A F L D patients with
sequential liver biopsies, 37% had progressive fibrosis over a mean period of 3.2
years. The majority (67%) of patients that were biopsied after a four-year interval
increased in fibrosis stage. The rate of fibrosis change was slow, overall being
0.02 ± 0.66 stages per year (or 0.09 ± 0.67 stages/year in non-cirrhotics).
Assuming fibrosis progression occurs at a linear rate, it would take an average of
50 years to advance just one stage (or 11 years in non-cirrhotics). However,
considerable variability in the rate of fibrosis change was observed, with one in
six patients having relatively rapid progression of more than 0.5 stages per year,
and four patients progressing from no (stage 0) fibrosis to advanced (stage 3-4)
fibrosis over an average interval of only 12 years. A proportion ofthe variability
in rate of fibrosis change was accounted for by diabetes (and BMI when cirrhotics
were excluded).
Previous studies examining fibrosis change over time have been limited by
small numbers. [116,117,125-127,131] In addition, patients have generally
undergone sequential biopsies due to clinical indications, potentially biasing
results towards patients with more severe or atypical disease. The majority of
patients in the present study underwent a biopsy at a predetermined interval as
part of a clinical protocol, therefore limiting this type of selection bias. Although
nearly half of our patients were taking either ursodeoxycholic acid or clofibrate
between biopsies, these agents were ineffective in changing the histological
course. [209,210] Despite these trials representing the largest number of patients
involved in clinical treatment studies for NASH to date, it may be possible that a
real therapeutic effect on histology was missed because of type 2 error. However,
there was no evidence that the histological course was different between treated
(i.e., ursodeoxycholic acid or clofibrate) and untreated patients, or between
participants in clinical trials and those biopsied outside of clinical trials (Tables
3.1, 3.2). Hence, the data from this study is likely to reflect the spontaneous
histological course that occurs over time in patients with NAFLD.
Diabetes mellitus was a strong independent predictor of higher rates of
fibrosis progression. This is consistent with cross-sectional studies that have
identified diabetes and insulin sensitivity as indicators of advanced liver fibrosis
in NAFLD.[128,130] Diabetes and insulin resistance have been noted to be
associated with a rapid progression of fibrosis in other liver diseases such as
chronic hepatitis C.[213,214] In addition, diabetes is recognized as a risk factor
for the development of chronic liver disease and liver cancer as well as death from
these conditions.[197,215] Aside from diabetes, BMI was independently
associated with higher fibrosis progression among non-cirrhotic patients. Thus,
elevated BMI is not only associated with advanced NAFLD,[66,128,129] but is
also a risk factor for more rapidly progressive disease. This finding supports the
observation in Chapter Two that BMI is a risk factor for liver related death.
The mechanisms relating diabetes and obesity to rapid fibrosis progression
may be due to accompanying metabolic derangements such as hyperglycemia,
hyperinsulinemia and hypoadiponectinemia. Cell culture studies have
demonstrated that hyperglycemia stimulates hepatic stellate cell growth as well as
collagen and free oxygen radical production. [216] In addition, hyperglycemia and
hyperinsulinemia have been shown to induce hepatic stellate cell connective tissue
growth factor production.fi 85] Interestingly, collagenization ofthe space of
Disse has been reported among diabetics and obese individuals. [217,218] This
may have the potential effect of impairing oxygen delivery from the sinusoid to
the hepatocyte thus promoting oxygen stress. [219]
Obesity and diabetes are associated with low levels of adiponectin. [220]
Adiponectin has been demonstrated to have protective effects against hepatic
steatosis and fibrosis in animal models of NASH and liver injury.[205,221,222]
In human cross-sectional studies, adiponectin is consistently associated with
hepatic steatosis grade among patients with NAFLD, although the association
with necroinflammatory grade and fibrosis stage is less consistent.[90,223,224]
This may reflect the difficulty in observing a cause-effect relationship from cross-
sectional studies.
The identification of obesity and diabetes mellitus as risk factors for rapid
fibrosis progression provides important prognostic information for the clinician
managing patients with NAFLD. The threshold for liver biopsy in patients with
these adverse prognostic variables should be lower as they are at higher risk of
developing advanced fibrosis, the diagnosis of which has management
implications for variceal and HCC screening.[225] In addition, weight loss
should be sought aggressively and glycemic control optimized in an attempt to
prevent progressive fibrosis. Additionally, it would be reasonable to enrol these
higher risk patients in clinical treatment trials if available.
Fibrosis stage on initial biopsy was a strong predictor of fibrosis rate, with
lower fibrosis stage associated with a higher rate of fibrosis progression. This
may in part be explained by the fact that patients without fibrosis cannot regress,
whereas patients with cirrhosis cannot progress in fibrosis stage. Alternatively,
the staging scale of fibrosis may not be linear with time so that progression
through early stages of fibrosis occurs faster than progression through later stages.
The association between initial fibrosis stage and fibrosis rate weakened after
excluding cirrhotics and patients without fibrosis (stage 0), but still remained
significant (p = 0.03).
Only three patients had bland steatosis and none of them developed fibrosis,
which is consistent with other studies reporting a relatively benign course among
patients with bland steatosis.[114,115,117] A significant reduction in steatosis
grade over time was observed, including three cirrhotic patients who subsequently
lost all evidence of steatosis. This has been observed previously in isolated
cases.[116,133] It is not known why this occurs, although mechanisms such as
shunting of insulin secondary to portal hypertension or loss of mitochondria
altering intrahepatic fat metabolism have been proposed. [2] Weight loss has also
been reported to improve steatosis grade[226] and it is pertinent to note that three
ofthe four NASH subjects who lost evidence of steatosis also lost up to 12.5kg of
body weight. The observed loss of hepatic steatosis in patients with cirrhosis
emphasizes the importance of considering NAFLD as a differential diagnosis for
cryptogenic cirrhosis. [2] Failure to identify cases of NAFLD related cirrhosis
may lead to an underestimation of disease related morbidity and mortality and
thus an underestimation ofthe severity ofthe natural history.
Similar to the reduction in steatosis over time, severity of hepatocyte
ballooning, inflammation and Mallory's hyaline also decreased and did not
parallel changes in fibrosis. Furthermore, aminotransferase levels improved
regardless of whether or not fibrosis progressed. Thus, improvement in
aminotransferase levels appear to indicate improvement in steatosis and
inflammation but not fibrosis and therefore may provide false reassurance
regarding prognosis. This should be kept in mind when interpreting clinical trials
lacking histological endpoints, as well as when monitoring patients in a clinical
setting.
Sampling error on liver biopsy may lead to erroneous grading and staging
among patients with NAFLD. Variability of fibrosis stage in paired liver biopsies
has been reported to occur in 30-41%) of biopsies, although it is uncommon to
vary by more than one stage.[103,104] Although we cannot confidently quantify
the effect of sampling error on our results, sampling variability is likely to upstage
patients as frequently as downstage patients. Therefore with increasing numbers
of biopsies (or power), the 'noise' sampling variability creates becomes less
significant and trends and true associations can be observed. [227]
Twenty-nine percent of patients had a reduction in fibrosis stage between
first and last biopsies. Part of this may be spurious due to sampling error,
however it is important to note that fibrosis pathogenesis is a dynamic process
which is well documented to regress after modification of underlying hepatotoxic
insults. [228] Improvement of fibrosis stage was not correlated with change in
BMI, however this may have been because the absolute change in BMI between
biopsies was negligible. Nevertheless, some patients may have had modification
of risk factors not measured such as physical exercise, glycemia, lipids or blood
pressure which were not able to be measured in the study.
It is important to note that these patients were recruited from a tertiary
referral hospital and thus may not represent the histological evolution of patients
with NAFLD in the general community. For instance, compared to a large
population study identifying subjects with fatty liver, the present study subjects
had a higher prevalence of diabetes (43% versus 18%) and elevated
aminotransaminase levels (90% versus 2 1 % ) . [4] A minority of patients in this
study had simple steatosis, although it is unknown what proportion of individuals
in the general community have simple steatosis as opposed to NASH.
In summary, the natural history of hepatic fibrosis in NAFLD is to
progresses slowly overall with considerable variability between patients.
Aminotransferase levels, hepatic steatosis and inflammatory features improve or
resolve over time as fibrosis progresses. Among patients with early stage disease,
a repeat liver biopsy is more likely to reveal progressive fibrosis after a four-year
interval. Fibrosis progression is more likely to occur among diabetics and also
occurs more rapidly among diabetics, patients with higher BMI and those with
lower stages of fibrosis. These patients should be targeted for therapeutic
intervention to prevent disease progression.
CHAPTER FOUR
EVOLUTION OF NAFLD IN INSULIN RESISTANCE RELATED
CONDITIONS : HYPOPITUITARISM AND HYPOTHALMIC DISEASE
4.1 INTRODUCTION
Insulin resistance is strongly related to the mechanisms leading to triglyceride
accumulation within the liver in primary NAFLD.[153,229] Consequently, the
prevalence of NAFLD is increased in clinical conditions associated with insulin
resistance such as obesity and diabetes mellitus. [157] Indeed, obesity and
diabetes mellitus are not only risk factors for the presence of NAFLD, but as
demonstrated in Chapters Two and Three, they also have prognostic significance.
Specifically, obesity and/or diabetes mellitus were found to be risk factors for
rapid fibrosis progression, liver related death and overall death among patients
with NAFLD. Therefore, the natural history of NAFLD in patients with these
insulin resistance-related co-morbidities can be expected to be more severe than in
patients without these conditions. Whether the natural history of NAFLD is more
aggressive among subjects with other clinical conditions associated with insulin
resistance remains unknown and is explored in this chapter.
Hypopituitarism and hypothalamic disease are associated with insulin
resistance. [168,230] Distinction between primary pituitary or hypothalamic
disease is often difficult because of their physiological interdependence and close
anatomical proximity. For example, extension of a peri-sellar mass such as a
craniopharyngioma, may lead to damage ofthe hypothalamus, the pituitary, or the
vascular pituitary stalk through which stimulatory or inhibitory hypothalamic
hormones pass to the pituitary.[231]
Hypopituitarism leads to complex hormonal disturbances including growth
hormone (GH) deficiency, hypoadrenalism and hypogonadism.fi68,232,233]
These metabolic changes lead to a change in body composition with an increase in
fat mass with related insulin resistance.[168,232] In addition, low levels of
dehydroepiandrosterone observed in hypoadrenalism, may also directly interfere
with pancreatic insulin sensitivity, thus promoting insulin resistance. [234]
Furthermore, low testosterone levels which occur with hypogonadism may impair
muscle insulin sensitivity.[23 5] Thus there are multiple potential mechanisms
linking hypopituitarism to insulin resistance.
Hypothalamic lesions due to structural damage (e.g. craniopharyngioma) or
genetic abnormalities (e.g. Prader-Willi syndrome) may also lead to insulin
resistance which is often accompanied by morbid obesity. [230] Interestingly,
experimental obesity produced by hypothalamic injury in various animal models
is uniformly accompanied by liver damage which may progress to cirrhosis. [236]
It is hypothesized that the damaged or dysfunctional hypothalamus is insensitive
to leptin and insulin which subsequently promotes the development of obesity,
insulin resistance and hyperglycemia. [180,237] A similar phenotype is observed
in fa/fa (Zucker) rats and db/db mice, which are leptin resistant because of
mutations in the leptin receptor, as well as in leptin deficient ob/ob mice. [23 8]
These animal models develop obesity, insulin resistance and fatty liver.
Therefore patients with hypopituitarism and hypothalamic dysfunction exhibit a
clinical phenotype of obesity and insulin resistance. Furthermore, insulin
resistance associated fatty liver is observed in animal models of hypothalamic
injury. Thus these patients may be prone to firstly, developing NAFLD and
secondly, having an aggressive disease course.
4.2 AIMS
To determine in a longitudinal follow-up study of subjects with hypopituitarism
and/or hypothalamic disease;
1. the association between insulin resistance related metabolic risk factors and
NAFLD.
2. the liver-related morbidity and mortality associated with NAFLD.
This study has been published in part (Adams et al, Hepatolology 2004; 39: 909-
914).
4.3 METHODS
4.3.1 Case Ascertainment
Patients with a diagnosis of either panhypopituitarism, hypothalamic obesity or
craniopharyngioma in conjunction with a diagnosis of NAFLD were identified
using the Mayo Clinic computerized master diagnostic index. This database
consists of diagnoses imputed by treating physicians, of all patients visiting the
institution. As craniopharyngioma is the commonest tumour associated with
hypothalamo-pituitary deficiencies, subjects with this disease were included to
maximize subject ascertainment.fi 80] The time period ofthe search was between
January 1990 and December 2001. The specific H I C D A codes used are listed in
appendix four.
Inclusion criteria for the study were firstly; a diagnosis of NAFLD as
determined by liver biopsy or by fatty infiltration on imaging studies in
association with abnormal liver enzymes; secondly a clinical diagnosis of
hypopituitarism or hypothalamic obesity. Subjects with a clinical diagnosis of
hypopituitarism also required biochemical evidence of pituitary deficiency such as
low levels of thyroid stimulating hormone, serum Cortisol (or abnormal
corticotropin stimulation test), follicle stimulating hormone, luteinizing hormone
or insulin-like growth factor-1.
Exclusion criteria were; secondary causes of NAFLD (see Table 1.1,
Chapter One); weekly alcohol intake of greater than 140 grams; clinical,
serological or histological evidence of other liver disease. Specifically, patients
were excluded if they had the following; hepatitis C antibody or RNA, hepatitis B
surface antigen, low alpha-one antitrypsin level or ZZ phenotype, low
ceruloplasmin level, positive anti-nuclear antibody or anti-smooth muscle
antibody with histological evidence of auto-immune hepatitis, positive anti-
mitochondrial antibody with histological features of primary biliary cirrhosis,
elevated ferritin or transferrin saturation with positive HFE genotype or
histological evidence of iron overload.
During the 12 year study period, a total of 945 patients with a diagnosis of
panhypopituitarism, hypothalamic obesity or craniopharyngioma were entered
into the diagnostic index. Twenty-seven of these had a concomitant diagnosis of
NAFLD. Patient medical records were reviewed to assess whether subjects met
the study inclusion criteria.* Six were excluded because ofthe presence of liver
disease other than primary N A F L D (one with alcoholic liver disease, one with
drug-induced liver disease) or the absence of confirmed NAFLD (two had normal
liver ultrasounds, two had not had liver imaging).
4.3.2 Patients
After the cases were identified, the medical records were extensively reviewed, to
assess the time course of pituitary/hypothalamic dysfunction as well as the
development of metabolic disorders and NAFLD. Time of diagnosis of pituitary
or hypothalamic dysfunction was taken from date of brain surgery if applicable, or
time of clinical diagnosis. Clinical features of hormonal disturbance and need for
replacement therapy was recorded. Time of diagnosis of NAFLD was taken from
date of liver biopsy or date of clinical diagnosis if the patient was not biopsied.
Liver histology was reviewed by liver pathologists and staged and graded
according to the classification published by Brunt et al.[101]
Body mass index, liver enzymes, lipid and glucose profiles and blood
pressure or history of hypertension were recorded at time of diagnosis ofthe
pituitary/hypothalamic disorder and at regular intervals thereafter when available.
The metabolic syndrome was defined in the study subjects at time of diagnosis of
NAFLD using the criteria proposed by the National Cholesterol Education
Program (ATP III), i.e., when at least three ofthe five following features were
present: [15 ^hyperglycemia (fasting blood glucose >110 mg/dl, or known
diabetes mellitus), hypertension (blood pressure >130/>85 or under treatment),
hypertriglyceridemia (>150 mg/dl or under pharmacological treatment), low-HDL
cholesterol (<40 mg/dl for males and <50 mg/dl for females), and obesity (BMI
>30 kg/m ). As waist circumference was not measured for most ofthe patients, a
BMI >30 kg/m2 was used instead to define obesity.[190]
Liver related morbidity (variceal haemorrhage, ascites, hepatic
encephalopathy, hepato-pulmonary syndrome) and cause of death or need for liver
transplantation after diagnosis of NAFLD was noted. Patient follow-up was
extended to May 2003. The study was approved by the Mayo Institutional
Review Board and all patients or responsible guardians gave informed consent for
participation in medical research.
4.3.3 Statistical Analysis
Continuous data are presented as mean + standard deviation and frequency data
are presented as number (proportion) of patients with a condition. Metabolic
variables at time of diagnosis of pituitary/hypothalamic disease and time of
diagnosis of NAFLD were compared with Wilcoxon Signed Rank tests. The
proportion of patients with glucose intolerance over time was compared using Chi
square tables.
4.4 RESULTS
4.4.1 Patient Characteristics
Ofthe 21 patients, 14 (67%>) were female. Nineteen (90%>) were Caucasian and
two were Hispanic. The mean age at time of diagnosis ofthe
pituitary/hypothalamic disease was 30 ± 20 years (range 3-66). The mean age at
time of diagnosis of NAFLD, was 36 ± 22 years (range 9-78).
4.4.2 Pituitary/Hypothalamic Disease
Brain tumour was the commonest cause of hypothalamic/pituitary dysfunction
affecting 15 (71%) patients, including eight with craniopharyngioma, six with
pituitary adenoma and one with an astrocytoma. The remaining six patients had
idiopathic hypopituitarism (four patients), hypophysitis (one patient) and Prader-
Willi syndrome (one patient). All patients with tumours underwent surgery and
four received additional radiotherapy. All patients were receiving physiological
doses of glucocorticoid and thyroxine replacement, apart from two who were on
sex hormone replacement only. Nine patients were on vasopressin supplement.
Growth hormone deficiency was tested in nine patients and confirmed in seven.
Two children who had evidence of growth retardation received growth hormone
supplementation. Hyperphagia was present in five patients; four of whom had
evidence of hypopituitarism requiring thyroxine, glucocorticoid and and
vasopressin supplementation. The remaining patient with hyperphagia had Prader-
Willi syndrome and was on the oral contraceptive pill only.
4.4.3 Development of The Metabolic Syndrome
4.4.3.1 Body Mass Index
Serial BMI measurements from time of diagnosis of the pituitary/hypothalamic
disease were available in eleven patients. Ten of these eleven patients gained
weight. The mean BMI increased between diagnoses from 26.7 ± 8.6 kg/m2 to
36.4 ± 9.5 kg/m2 (p=0.004); at an average yearly rate of 2.2 ± 2.2 kg/m2. Serial
weights were not available for the remaining 10 patients, although one patient was
noted to have a marked weight gain after removal of their craniopharyngioma,
with a subsequent BMI recording of 35.3 kg/m2.
At time of diagnosis of NAFLD, 18 of 21 patients had BMI above normal.
Five (24%) patients were overweight (BMI 25.1- 29.9 kg/m2) and 13 (62%) were
obese (BMI > 30 kg/m2) with four (19%) of these being severely obese (BMI > 40
kg/m2).
4.4.3.2 Glucose Tolerance
Fasting blood glucose levels were available in 10 patients at time of diagnosis of
pituitary/hypothalamic disease and were elevated in only one patient who was
diabetic. At time of diagnosis of NAFLD, 13 (62 %>) of 21 patients had a history
of elevated fasting blood glucose levels including 10 with diabetes and three with
glucose intolerance. The proportion of patients with glucose intolerance or
diabetes increased significantly from diagnosis of pituitary disease to diagnosis of
NAFLD (10% vs. 62%, p=0.007). Mean glucose level increased non-significantly
from diagnosis of pituitary disease to diagnosis of NAFLD (101 ± 16 vs. 109 ± 23
mg/dl, p=0.2).
4.4.3.3 Dyslipidemia and Hypertension
Fasting triglyceride and cholesterol levels increased after diagnosis of
pituitary/hypothalamic disease as shown in Figures 4.1 and 4.2, respectively. At
time of diagnosis of NAFLD, 14 patients (67%) had hypertriglyceridemia (> 150
mg/dL), eleven (52%) had low HDL levels (< 40 mg/dL for females, < 50 mg/dL
for males) and seven (33%) patients had hypercholesterolemia (> 240 mg/dL). At
550
500 -
450 -
400 - .,
350 - ^xr^^^
300 - ^^^^^ -L
250 - if 1
200 -
150 •" 1 1 1 1 '
0-12 13-24 25-36 37-48
Months after Pituitary/Hypothalamic Diagnosis
Figure 4.1 Mean fasting levels of triglyceride in the first 48 months after
diagnosis of pituitary/hypothalamic disease (n=15). Regression line/standard
error bars shown.
_J
CO
CO *D
0>
o
Ill
300 -| 1
280 - T
260 - ^^' O T ' ^ " ^
240 - ^^^nL 4>
220 -
200 -" i 1 1 1 '
0-12 13-24 25-36 37-48
Months after Pituitary/Hypothalamic Diagnosis
Figure 4.2 Mean fasting levels of cholesterol in the first 48 months after
diagnosis of pituitary/hypothalamic disease (n=15). Regression line/standard
error bars shown.
5
o >_ 0) (0
o o
o
time of diagnosis of N A F L D , five patients were hypertensive. T w o of these were
diagnosed after their pituitary/hypothalamic disease was detected.
4.4.3.4 Acanthosis Nigricans
Three patients had acanthosis nigricans at time of diagnosis of their NAFLD.
Two of these patients had craniopharyngiomas removed at ages eight and twelve
with subsequent diagnoses of NAFLD at ages 12 and 26 years, respectively. The
third patient had Prader-Willi Syndrome who had NAFLD diagnosed at age 20.
4.4.3.5 The Metabolic Syndrome
At the time NAFLD was diagnosed, 15/21 (71%) of subjects fulfilled diagnostic
criteria for the metabolic syndrome.
4.4.4 Association with NAFLD
The 21 patients were diagnosed with NAFLD 6.4 ± 7.5 years (median 3 years)
after the diagnosis of pituitary/hypothalamic dysfunction. Liver enzymes were
available and within the normal range in six patients at time of diagnosis of
pituitary/hypothalamic disorder. All patients subsequently developed abnormal
liver enzymes by time of diagnosis of NAFLD (Table 4.1). AST levels were
raised in all patients and ALT levels were elevated in 13 of 17 (76%). Five
patients had elevated alkaline phosphatase levels for their age and gender,
including four patients with less than twice (1.3, 1.3, 1.5 and 1.8) the upper limit
of normal and one patient with more than twice (2.7) the upper limit of normal.
Imaging studies ofthe abdomen did not reveal any biliary abnormalities in the 21
patients. The patient who had alkaline phosphatase levels 2.7 times that of normal
eventually underwent liver transplantation with the explant liver demonstrating
cirrhosis without features of bile duct disease or cholestatic liver disease.
In ten patients, the diagnosis of NAFLD was confirmed by liver biopsy, six
patients were cirrhotic (29% of total cohort), two had NASH with fibrosis and two
had simple steatosis. Histological features of these patients are summarized in
Table 4.1. Patients diagnosed with cirrhosis were relatively young with the
median age being 28 years (range 15-51). The two patients receiving growth
hormone supplementation had mild disease with simple steatosis and NASH with
stage 1 fibrosis respectively.
<N
Q J P*H
<
2 o O n
O u
rd
03
>-. bO
o
J3
flj
O o • —
UJ >
0)
es
H
c/a O X) • I-H
o
c/a
o -i—i
C/3
'S "J
S-l
DQ
c/a O
<
>
0) bfl
o
c
"bfl
S
2 •J
< ^
z x
z J
D
cfl
03
z £
tu SH
cu > <u tZ3
<u t-cu > CD t>0
OO
cn in
o o
in
o
m
MD
MD CN
MD
o ^
<5t
cn O
o
t--CN
•NI"
o m
fl) m
s
cn
<u o3
B cu [in
cn in
<u on
s
03
ti o PH
rn
oo
CN
in
<u 15
o
t
***• CN
o J<
cn
o
•n
<n
cn
o
-1
cn
o
u u u u
o fc!
U
cu
c
1
u cd iH tu Ti o -a 7713
03 SH <U -r) O
s s
flj s-flj
> <u on
cu SH
<U > <L) CZ*
1>
o in
o in
o
oo ON CN
MD 1—1
cu cS
B <u fc
CN t̂
<U
cd
s <U PH
OO T—
<u 03
s <u PH
cn •*-J-
cu 03
s flj PH
t>
JJ *c3
s
in r~'
AJ "o3
s flj PH
r-H (N cn in MD t> oo ON —
r-H ON
•*' cn
ON ON
00
o oo ^ H
oo o
•n
— H
^D O
ON
o MD O
o CN
r̂ *-+
MD O in
in r-~ CN
CN r-H
<c z
ON
J Ov
o cn J
<
z o CN
r̂ J
oq in
MD
r̂ t̂
o
O -H ^H
cn in
JJ "g *rH flj
•>; P H
c r-H (N) O r-H rJ U
Cfl
O H
PC
c o 1 —
I tS C
Cfl
O 13 cu
-4—>
oo
fi /—
'5 J
5 a
3 SH
CQ
Cfl
O
OH
r*i
00
H
CU
Cl CU
o c <u
bfl
r̂ Z
-I
~bfl
a
z
Z -J D
Z J
oo o
cn
.J <
13 > H flj -4-4
c 1—1
cu bfl <
D w
/*—X Cfl SH
>. V—'
* OO
r-H
13
CN CN
oo CN
o
MD
cn
z
CN MD ^
CN
cn CN
OO
MD MD
CU
Is
B flj
PH
cn
CN
OO
cu
s
in
cn
JJ
13
a OJ p-H
cn
O N T-J-
o o
o o J CN
CN in
JJ 13
<u PH
CN
OO "̂ t MD ON
o o d d
Ji in
<
Z
cn cn
MD r-
<u 13
B aj PH
00
CN
MD ON
o o
CN
CN cn
o MD r-H
flj
13
B aj PH
ON
OO
MD O
JJ
13
B AJ
P*H
OO cn
cn
d
f̂" MD
d d
(N O ^ CN CN <N
cn CN
r^ CN
** in —
JJ 13
B OJ PH
d -H CN
cn
d -H MD
d
m
d +1 ON
d
CN CN +1 ON CN
cn
+1
CN
in
r^ +1
MD
±1 O cn
o CN
O r-CN CN
flj Q 00 +1
SH
J
D in •-tf
i
o
ON CN
H -J <
•J
D
cn
o • 4 — i
CN
cz*
CU bfl
SH
13
o c flj
-cl
H a)
• * *
o s o o tu
<N
00
CU
03 .«
M B SH Tl
o SH
OH
~bfl
in
in .flJ
<^~)
13
•If
"bfl
<4-l
O flj
bfl
<
II
SH
OJ
C flj
bfl
C 03 flj
bfl cs
bfl fi
-3 SH
O cu cu 03 Cfl cu
< B .B
CS > CU Cfl
03
13 rfi
OH Cfl
O -fi O H
CU fi 03
13 SH
a AJ
bfl 0
flj Cfl
<+H
o
SH
OJ OH
OH
3
z -1
D AJ"
I '53 > 03 •4—»
o c
<
Z SH
flj
-a SH
O Cfl
CJ
J3 13 rfi -4-4
O OH
rn rfi SH
O
-O 3 SH
— -S 13 •4-i
o -4-4
. rv
Cfl flj
13
03
"o3
g O Z Cfl T3 C O CJ
OH
o OH
>, xi t+H
o cfl 'cfl O 0 bfl ca
— aj
Cfl
03
C cu
B o flj
OH Cfl ;n bfl
_o "o t3 DH
<
Z -I—»
3 O
-4—»
c s-=1
flj
> 03 .HI
r f i
bfl
O -fi
03 CJ
"fi ° -4-> 03
m
c cu
• i—i
13 DH
13
o fi
c o t5 SH H—4
tS
c C+H
-d 03
-fi
flj
ti o
cfl O OH
Cfl O -fi fc 'o
T3 CU cfl O C ofl 03 cfl 03
.fi MD
C cu
'•s DH
sJ cu >
cu -4—4
I OH
X flj
fi 3 O Cfl
03 SH £ rfi 3
fi O -3 flj Cfl
O fi bfl o3
-3 cfl 3 O '> (U SH
OH
4.4.5 Liver Related Morbidity and Mortality
Long term follow-up from diagnosis of NAFLD was 66 ± 33 months (median 72
months, range 12 to 120) and was available for 18 patients, the other three being lost to
follow-up. Two patients underwent liver transplantation. One of these underwent
transplantation at age 25, after being diagnosed with idiopathic anterior pituitary failure
at age 16. The second patient underwent transplantation at age 46 after having a
craniopharyngioma removed at age 10.
Overall six patients (29%) died. Two deaths were liver related and occurred in
cirrhotics. One died from hepatocellular carcinoma and the other died post-liver
transplantation due to recurrent NASH and hepato-pulmonary syndrome. One other
cirrhotic died from a bleeding gastric ulcer. Two patients with simple steatosis died
from non-liver related causes (one from lymphoma and one from a bleeding Dieulafoy
lesion). One patient died from peritonitis unrelated to their liver disease.
4.5 DISCUSSION
The clinical association between features ofthe insulin resistance (metabolic) syndrome
(obesity, diabetes and dyslipidemia) and NAFLD was noted in the first descriptions of
the disease.[239] Subsequent metabolic studies have confirmed that insulin resistance
is common among NAFLD patients and is independent of severity of liver disease or
BMI. [79,240] Other conditions associated with insulin resistance such as hypertension,
hyperuricemia, lipodystrophy and polycystic ovarian disease have also been described
in association with NAFLD. [3] This case series expands the clinical association to
include patients with hypopituitarism and hypothalamic dysfunction. In these patients,
pituitary/hypothalamic disease tended to occur at a relatively young age. The resultant
hormonal dysfunction was followed by precipitous weight gain and development of
hyperglycemia, dyslipidemia and NAFLD. NAFLD developed relatively rapid
(average of 6.2 years) after the diagnosis of pituitary/hypothalamic dysfunction, and the
liver disease in these patients was severe; 29%> ofthe cohort developed cirrhosis and
three (14.2%) ofthe 21 were either transplanted or died from liver related causes. In
comparison, the occurrence of cirrhosis and liver related death in comparable hospital
based series of NAFLD patients range between 5-20% and 2-8%
respectively.[115,116,118,119] Similarly, cirrhosis and liver related death rates among
the community based cohort of NAFLD patients in Chapter Two, was substantially
lower at 5% and 1.7% respectively. Thus the natural history of NAFLD among
subjects with hypopituitarism and/or hypothalamic disease appears to be rapid and
severe.
Prior to publication of this study,[241] only three cases (one reported in German)
had described the development of NAFLD in association with hypopituitarism or
hypothalamic disease. [242-244] Subsequently, seven more children have been reported
to develop NAFLD after a diagnosis of hypopituitarism. [245-247] The propensity to
develop severe liver disease seen in the current series was also noted in the above case
reports. Five ofthe ten (50%) cases reported progressed to advanced fibrosis with one
case requiring liver transplantation and one case subsequently died at age 13 secondary
to his liver disease. The metabolic and hormonal changes accompanying
hypopituitarism and hypothalamic injury may provide insight into the pathogenesis of
this aggressive form of NAFLD.
The onset of hypopituitarism leads to central obesity, hyperlipidemia and insulin
resistance providing the impetus for the genesis of NAFLD. [168,231] These metabolic
changes are principally thought to be due to GH deficiency, although altered insulin like
growth factor-1, Cortisol and gonadotropin metabolism have also been
implicated.[170,248,249] Adult patients with anterior pituitary deficiency and
associated GH deficiency have fatty infiltration ofthe liver more frequently than
patients with anterior pituitary hormone deficiency without GH deficiency.[250] In
addition, patients with NAFLD have lower GH levels compared to unmatched
controls. [251] However, this may simply reflect the decrease in growth hormone that
occurs with obesity. [252] In addition it is not clear that the level of insulin resistance in
GH deficient patients is greater than in BMI matched healthy
controls.[170,179,253,254] This suggests that obesity may play a more important role
in the development of insulin resistance than GH deficiency per se. To confuse the
issue further, acromegaly and G H excess are associated with diabetes mellitus.[255]
Similarly GH supplementation in deficient patients can worsen insulin resistance and
increase lipolysis leading to increased free fatty acid concentrations.[248,256] These
metabolic changes would presumably instigate or worsen NALFD. One case report
however, demonstrated improvement in liver enzymes and hepatomegaly after growth
hormone supplementation in a previously deficient individual. [244] Therefore, the role
of growth hormone and insulin sensitivity independent of obesity is unclear. Growth
hormone homeostasis may be finely balanced with too much or too little resulting in a
change in body fat composition and insulin resistance. Alternatively, other hormones
such as leptin may be important pathogenic factors in the development of obesity and
insulin resistance.
Leptin levels are significantly increased among patients with hypopituitarism and
GH deficiency compared to BMI and body fat content matched controls.[179,257]
Presumably this reflects enhanced peripheral and/or central resistance to leptin.
Similarly hyperleptinemia out of proportion to the level of obesity occurs in patients
with hypothalamic damage post craniopharyngioma resection, suggesting central leptin
resistance. [180]
Resistance to leptin has been modelled in the fa/fa (Zucker) rat and the db/db
mouse, both of which have defective leptin receptors. Hypothalamic resistance to leptin
decreases inhibition of orexigenic hormones such as neuropeptide Y and Agouti-related
peptide, leading to net increase in fat mass, hyperinsulinemia and a decrease in
metabolic rate.[258] The resultant phenotype in these animals is one of hyperphagia,
obesity, insulin resistance and NAFLD. [238] At least five of our patients shared this
phenotype of hypothalamic obesity with hyperphagia and marked obesity. Hyperphagia
however is not always present in this syndrome[l69,230] and it is probable that more
patients in the current series suffered from hypothalamic obesity.
As argued above, the onset of hypopituitarism and hypothalamic dysfunction
may lead to central obesity, hyperlipidemia and insulin resistance, thus providing
impetus for the genesis of NAFLD. However, the immediate hormonal disturbances
occurring with hypopituitarism and hypothalamic disease may induce acute NAFLD
prior to the onset of obesity and glucose intolerance. The severe disease course seen in
this case series would support an additional unique metabolic disturbance aside from
obesity and insulin resistance, to be contributing to severe progressive NAFLD. Acute
elevations of serum aminotransaminases due to fatty liver have been noted as early as
three days post suprasellar tumour resection. [247] Leptin was noted to be markedly
elevated in these patients. Leptin has been demonstrated to have pro-inflammatory
effects and has been implicated in enhancing adipocyte production of TNFa. [259,260]
TNFa is an inflammatory mediator in the liver and also promotes insulin
resistance. [261] In addition, leptin has been characterized as a pro-fibrotic cytokine in
animal models of liver fibrosis and fatty liver.[262,263] In cross-sectional human
studies however, leptin is not correlated with fibrosis stage among patients with
NAFLD and the association between leptin and hepatic steatosis is
inconsistent. [264,265]
There are other potential explanations for NAFLD in these patients. All patients
except two were on corticosteroid replacement, which is a known cause of liver
steatosis. However it is rare for corticosteroid induced steatosis to progress to
steatohepatitis and cirrhosis, as was observed frequently in this case series. [17]
Furthermore, the dosing of corticosteroid used was physiological, replacing absent
endogenous corticoids. Thus, it is unlikely this physiological dosage of corticosteroids
had a significant role in the development and progression of NAFLD in our patients.
A weakness of this case-series is that the degree of insulin resistance and levels
of adipocytokines (e.g. leptin, adiponectin) were not measured. Thus, this series
describes only the association (instead of causation) of NAFLD and
hypothalamic/pituitary dysfunction. In addition, as this is a retrospective study, a degree
of selection and ascertainment bias cannot be completely ruled out. For instance,
abdominal ultrasound was not performed routinely in patients with
hypothalamic/pituitary dysfunction and thus some cases of NAFLD were likely to go
undetected. Further prospective studies with a more detailed metabolic/hormonal
evaluation of these patients would better characterize the pathogenesis of NAFLD
among patients with hypothalamic/pituitary dysfunction. Despite these limitations,
however, this data provides important implications for the work-up and management of
patients with hypothalamic/pituitary dysfunction. Specifically, these patients should be
routinely screened for NAFLD and associated metabolic abnormalities such as obesity,
glucose intolerance and dyslipidemia should be managed aggressively. Hepatologists
and endocrinologists should be aware of this novel evidence that hypothalamic/pituitary
dysfunction may be accompanied by rapidly progressive NAFLD with the potential for
significant liver related morbidity and mortality.
CHAPTER FIVE
EVOLUTION OF NAFLD IN INSULIN RESISTANCE RELATED
CONDITIONS: DIABETES MELLITUS
5.1 INTRODUCTION
Type 2 diabetes mellitus is the archetypal disease associated with insulin resistance.
Patients with type 2 diabetes mellitus exhibit reduced hepatic and muscle sensitivity to
insulin, which results in hyperglycemia when the compensatory increase in insulin
secretion by pancreatic (3-islet cells is insufficient. Aside from abnormal glucose
homeostasis, other metabolic pathways such as lipid metabolism are also disturbed.
Resistance of adipose tissue to the effects of insulin results in increased lipolysis with
subsequent FFA release. [165] Increased flux of FFA into the liver, coupled with
increased de novo hepatic lipogenesis and reduced FFA oxidation promote hepatic
triglyceride accumulation and the development of NAFLD.[73,165] Consequently,
diabetes mellitus is frequently observed in patients with NAFLD, being present in 18-
45% of cases.[4,53,150] Similarly, NAFLD is common among patients with type 2
diabetes with the prevalence ranging between 49-62%>.[60,183,184] In contrast,
patients with type 1 diabetes mellitus have low insulin levels without significant insulin
resistance and infrequently develop NAFLD.[266]
The preceding chapters demonstrated that type 2 diabetes mellitus was an adverse
prognostic factor affecting the natural history of patients with NAFLD. Chapter Two
demonstrated that among community diagnosed patients with NAFLD, a co-existing
diagnosis of diabetes increased the risk of death nearly three fold. [267] Chapter Three
revealed subjects with NAFLD and diabetes mellitus had significantly higher fibrosis
progression than those without diabetes. [268] Furthermore, Chapter Four revealed that
subjects with conditions associated with insulin resistance and diabetes
(hypopituitarism and/or hypothalamic dysfunction), had an aggressive form of NAFLD
frequently leading to advanced liver disease. Additional evidence from cross-sectional
studies has demonstrated that diabetes mellitus is associated with more severe
histological disease, being a risk factor for inflammation and advanced fibrosis in
patients with NAFLD.[66,128] In addition, Younossi and colleagues found diabetes
increased the risk of liver related death among a cohort of 132 individuals with
NAFLD.[150]
Why diabetes mellitus increases the risk of liver disease and its complications is
unknown. Possible mechanisms may relate to insulin resistance, hyperinsulinemia,
hyperglycemia or the associated adipocytokine disturbances which accompany type 2
diabetes. For example, hyperinsulinemia and hyperglycemia may induce liver related
damage by up-regulation of fibrogenic growth factors produced by hepatic stellate
cells.[l85,216] Alternatively, diabetes associated adipocytokine disturbances such as
increased TNFa and leptin levels or reduced adiponectin levels may play significant
roles in contributing to liver damage.[261]
Whilst it is clear that diabetes mellitus has an adverse effect on the natural history
of NAFLD, it is not clear whether NAFLD itself impacts on the natural history of
diabetes mellitus. NAFLD was demonstrated in Chapter Two to be independently
associated with an increased risk of all-cause mortality compared to the general
population. [267] The impact of NAFLD on all-cause mortality among diabetics is
unknown.
The worldwide prevalence of diabetes is increasing at a rapid rate, reported to be
7.4% in Australia and 7.9% in the United States in the year 2001. [181,269] The
lifetime risk of developing diabetes mellitus is significant, being estimated to be 32.8%>
for males and 38.5% for females living in the United States and born in the year
2000. [182] Of concern is that diabetes mellitus is associated with a reduced life
expectancy of between 11 and 19 years. [182] Research is urgently required to examine
for risk factors which influence the risk of death.
5.2 AIMS
To examine the effect of a diagnosis of NAFLD on all-cause mortality among patients
with type 2 diabetes mellitus.
5.3 METHODS
5.3.1 Study Setting
The study population was derived from residents of Olmsted County located in
southeastern Minnesota (described in detail in Chapter Two).
5.3.2 Case Ascertainment
Patients with diabetes mellitus were identified using the resources ofthe REP and the
Mayo Clinic Laboratory Information System (MLIS). The REP is a data linkage
system indexing all medical diagnoses of residents of Olmsted County and is described
in the Methods section of Chapter Two. The M L I S is a database of all blood test results
performed by the Mayo Clinic Laboratories since 1983. Mayo Clinic Laboratories
performs the vast majority of biochemical analyses in Olmsted County. The database
includes information on the geographical residence of patients allowing categorization
of blood results by County.
Patients with diabetes mellitus were identified by searching the REP master
diagnostic index using HICDA codes for diabetes and its complications (see Appendix
Five). As the diagnostic criteria for diabetes mellitus has changed over the past two
decades from a fasting glucose level > 140mg/dl to > 126mg/dl,[270] the MLIS
database was searched for all Olmsted County residents with an outpatient fasting blood
glucose (FBG) level >126 mg/dl to ensure as complete ascertainment as possible. Case
ascertainment was over a 20-year period from January 1st 1980 to December 31st 1999.
Patients with a concomitant diagnosis of NAFLD were identified using HICDA
codes for fatty liver, hepatic steatosis or steatohepatitis (see Appendix One). Medical
records of patients identified were extensively reviewed to ensure patients fulfilled the
case definitions (see below). After identifying patients with diabetes and NAFLD, a
random sampling ofthe remaining patients with diabetes was undertaken with
stratification for age, gender and date of diagnosis of diabetes (performed by Scott
Harmsen, Division of Biostatistics, Mayo Clinic).
5.3.3 Case Definitions
The diagnosis of diabetes required a fasting blood glucose level > 126 mg/dl or a
random glucose > 200 mg/dl in the presence of symptoms (polyuria, polydipsia, blurred
vision). [271] Date of diagnosis was from date ofthe first recorded elevated fasting
glucose level. Patients were excluded if they had type one diabetes or secondary causes
for hyperglycemia (non-fasting, glucocorticoids, acute illness, Cushing's disease,
lipodystrophy).
The diagnosis of NAFLD required confirmation of hepatic steatosis by either
abdominal imaging (ultrasound, computed tomography or magnetic resonance imaging)
or by liver biopsy. Patients with secondary causes for hepatic steatosis (medications,
metabolic disease, gastrointestinal bypass surgery) were excluded, as were patients with
evidence of other liver disease on clinical history or examination, laboratory studies,
imaging or biopsy. Patients at risk of viral hepatitis due to intravenous drug use or
blood product transfusion prior to 1992 (when blood donor hepatitis C antibody testing
was introduced), were excluded if they had not had hepatitis B or C serology performed
after their exposure. Information on alcohol consumption (type, amount, frequency,
duration as well as alcohol abuse screening questions) was collected prospectively as
part ofthe medical record, by a patient history form filled out by the patient and
reviewed by a nurse and physician at each visit. Patients with an average weekly
ethanol consumption >140 grams were excluded.
Patients with diabetes but not NAFLD were required to have normal liver
function tests (bilirubin, AST, ALT, albumin) documented on at least three occasions
over time, without any unexplained abnormal liver tests at time of diagnosis or during
follow-up. In addition, subjects were required to have normal hepatic parenchyma if
abdominal imaging was performed during follow-up.
The medical records of 720 subjects were reviewed. Reasons for exclusion from the
study were; date of diagnosis of diabetes pre-1980 (n=69), type 1 diabetes (n=41),
gestational diabetes (n=6), hyperglycemia due to medications or illness (n=15),
hyperglycemia not within diabetic range (n=36), non-NAFLD liver disease (n=9),
initial diabetes diagnosis outside of Olmsted County or non-County residents (n=53),
unexplained elevated liver tests (n=109) or less than three normal liver tests (n=45). Of
the remaining 337 subjects with diabetes mellitus included within the study, 116 were
diagnosed with NAFLD after their initial diagnosis of diabetes mellitus.
5.3.4 Data Abstraction
The following information was abstracted from the medical record at date of diagnosis
of diabetes (see Appendix Six for data abstraction sheet); age, gender, race, diabetes
treatment required within three months of diagnosis (diet, oral hypoglycaemic
medications, insulin), past medical history of hyperlipidemia, hypertension, ischemic
heart disease (acute coronary syndrome or angina), cerebrovascular disease (transient
ischemic attack or stroke), chronic renal failure, examination findings (height, weight,
blood pressure), medications, laboratory data (bilirubin, AST, ALT, alkaline
phosphatase, albumin, fasting glucose, glycosylated haemoglobin, total cholesterol,
HDL cholesterol, triglyceride level). In addition, for patients with diabetes but not
NAFLD, serial liver tests after time of diagnosis of diabetes and any hepatic imaging
studies were recorded. Follow-up and cause of death were obtained from medical
records and death certificates. Patient follow-up was extended to January 1st 2005.
5.3.5 Statistical Analysis
Statistical analysis was performed by myself in conjunction with Steven Harmsen. The
results are presented as number and percent if categorical or as mean and standard
deviation if continuous. Baseline characteristics of patients with or without NAFLD are
compared using students t test, Mann-Whitney U test, Chi squared tables or Fishers
exact test when appropriate. Cox proportional hazards modelling was used to evaluate
overall mortality with NAFLD treated as a time dependent covariate from date of
diagnosis. Variables that were significant at the 0.05 level in separate Cox proportional
hazard models were considered for the multivariable model. Analysis was performed
using SAS Release 8.2 (SAS Institute Inc., Cary, NC).
5.4 RESULTS
5.4.1 Patient Characteristics
A total of 337 subjects with diabetes mellitus were identified; 116 who had
subsequently been diagnosed with NAFLD either at time of diagnosis of diabetes or
during follow-up and 221 without a diagnosis of NAFLD during the follow-up period.
Subjects without NAFLD were significantly more likely to be male, older and ex or
current smokers, compared to those who developed NAFLD (Table 5.1). Subjects with
NAFLD were more likely to be obese and correspondingly had a higher mean BMI.
There was no difference in presence of hypertension, hypertriglyceridemia, low HDL
129
Table 5.1 Clinical characteristics at time of diagnosis of diabetes
Variable
Age (years)
Male/Female
BMI (kg/m2)
Obese
Diabetes:
Diet/Oral/Insulin
Hypertensive
High TG
Low HDL chol.
IHD
CVD
Malignancy History
Smoking Status
Non / Ex / Current
Bilirubin
A L T (IU/L)
A S T (IU/L)
Albumin (gm/dl)
Platelets (xl07L)
HbAlc
Triglyceride (mg/dl)
H D L chol. (mg/dl)
Total Cohort
(n=337)
58 ± 13
165/172
32.7 ±6.9
198/315 (64%)
233 / 82 / 22
(69% / 24% / 7%)
213/334 (63%)
187/257 (73%)
136/191 (71%)
55/334 (16%)
17/334 (5%)
26/335 (8%)
193/76/65
58%/23%/19%
0.6 ±0.4
33 ± 31
58 ±41
4.2 ±0.4
256 ±75
10.3 ±3.5
263 ± 207
44 ±23
DM + NAFLD
(n=116)
55 ± 13
43/73
34.2 ±7.5
77/110 (70%)
84/26/6
(72% / 22% / 5%)
72/116 (62%)
70/88 (80%)
56/73 (77%)
16/115 (14%)
5/115 (4%)
9/116 (8%)
78/19/18
68%/16%/16%
0.7 ±0.6
69 ±42
54 ±43
4.2 ±0.5
248 ± 77
10.1 ±3.7
300 ± 267
44 ±26
DM -NALFD
(n=221)
59±13
122/99
31.9 ±6.4
117/205 (57%)
149/56/16
(67% / 25% / 7%)
141/218 (65%)
117/169 (69%)
80/118 (68%)
39/219 (18%)
12/219 (5%)
17/218 (8%)
115/57/47
53%/26%/21%
0.6 ±0.4
28 ±16
21 ±5
4.2 ±0.3
261 ± 73
10.4 ±3.4
242 ±164
43 ±21
P
value
0.005
0.002
0.008
0.02
0.6
0.6
0.08
0.2
0.4
0.7
0.9
0.02
0.4
O.001
<0.001
0.9
0.2
0.6
0.07
0.7
cholesterol levels or past history of vascular disease or malignancy. Similarly, there
was no difference in the apparent severity of diabetes with HbAlc values and treatment
regimes similar between the two groups.
At time of diagnosis of diabetes mellitus, subjects with NAFLD had higher levels
of aminotransaminsases compared to those without NAFLD. However, bilirubin,
albumin and platelet levels were similar between the two groups. Subjects without a
diagnosis of NAFLD had an average total number of 12 AST or ALT levels within
normal limits during follow-up.
Ofthe 116 diabetic subjects diagnosed with NAFLD, 114 underwent abdominal
imaging studies (85 ultrasound and 29 computed tomography). Twenty three
underwent a liver biopsy and one patient underwent a fine needle aspiration which
confirmed hepatic steatosis but was insufficient for grading of necroinflammatory
activity or staging of fibrosis. Liver biopsy was consistent with NASH in all cases,
with either hepatocyte ballooning and lobular inflammation present or hepatic fibrosis.
During the follow-up period, 10/116 (8.6%) patients with NAFLD were diagnosed with
cirrhosis a mean of 5.8 (±5.6) years after their initial diagnosis of diabetes.
5.4.2 Follow-up
The mean (± standard deviation) duration of follow-up ofthe total cohort was 10.9 (±
5.2) years (median 9.9 years, range 0.1-25.0 years). The diagnosis of NAFLD followed
the initial diagnosis of diabetes after a mean of 0.9 (± 4.6) years in the 116 subjects.
The average follow-up from initial diagnosis of diabetes was shorter among subjects
who developed N A F L D than those that did not (9.2 ±5.2 years vs. 11.7 ± 5.0 years,
pO.001).
5.4.3 Mortality
Overall, 99/337 (29%) subjects died during follow-up; 27 with NAFLD and 72 without
NAFLD. Heart disease and malignancy accounted for approximately half of all deaths
and were the commonest causes of death overall as well as among patients with
NAFLD and without NAFLD (Table 5.2). Liver related death was the third leading
cause of death among diabetics with NAFLD, accounting for 5/27 (19%) of all deaths
in this group compared to 0/72 deaths among the group without NAFLD. The liver
related mortality rate for diabetics with NAFLD was 4.3% (5/116) over the follow-up
period. One patient died from liver failure secondary to cirrhosis complicated by a
HCC; the other patients died from liver failure (n=3) and variceal bleeding (n=l).
5.3.5 Effect of NAFLD on Mortality
To examine the effect of NAFLD on survival among patients with diabetes mellitus,
NAFLD was entered as a time dependent covariate into a multivariable Cox
proportional hazards model (Table 5.3). The model was adjusted for age and gender
because ofthe significant differences in these variables between diabetics diagnosed
with NAFLD and those not diagnosed with NAFLD (Table 5.3). In addition, as
treatment of diabetes mellitus has changed over the 25 year study period potentially
influencing outcome, the model was also adjusted for the date of diabetes diagnosis.
Thus NAFLD was significantly associated with an increased risk of death (hazard ratio
Table 5.2 Causes of death among patients with diabetes mellitus with or without
NAFLD.
Heart Disease
Malignancy
Infection
Cerebrovascular
Liver Related
Other
Unknown
Overall
Total Cohort
(n=337)
31 (31%)
23 (23%)
15 (15%)
5 (5%)
5 (5%)
18(18%)
2 (2%)
99
DM + NAFLD
(n=116)
5 (19%)
9 (33%)
3(11%)
0 (0%)
5 (19%)
4(15%)
1 (3%)
27 (23%)
DM - NAFLD
(n=221)
26 (36%)
13 (18%)
12(17%)
5 (7%)
0
14 (20%)
1 (2%)
72 (33%)
1
Table 5.3 Effect of N A F L D on overall survival among subjects with diabetes mellitus
Variable
Age 50-60*
Age 60-70*
Age >70*
Male
D M date
NAFLD
Estimate
0.91
2.09
2.87
0.24
0.00
0.49
S.E.
0.49
0.44
0.45
0.21
0.00
0.23
P value
0.06
<0.0001
O.0001
0.2
0.2
0.03
H.R.
2.49
8.11
17.78
1.28
1.00
1.63
9 5 % C.I.
0.95-6.55
3.40-19.33
7.39-42.79
0.84-1.92
1.00-1.00
1.04-2.56
* relative to age <50 years
1.63, 95%> confidence interval 1.04-2.56). N A F L D remained an independent risk factor
for mortality when controlling for other potential confounding variables such as
smoking (HR 1.6,95% CI 1.03-2.6), hypertension (HR 1.7, 95% CI 1.1-2.7), obesity
(HR 1.8, 95% CI 1.1-2.8), hyperlipidemia (HR 2.1, 95% CI 1.2-3.6), history of cancer
(HR 1.6, 95% CI 1.02-2.5), or ischemic heart disease (HR 1.5, 95% CI 1.0-2.4).
To determine whether patients with NAFLD had a greater risk of a specific cause
of death, multivariable Cox proportional hazards modelling was used with the end-
points of death due to heart disease, vascular disease (combination of death from
ischemic heart disease and cerebrovascular disease), malignancy and liver related death.
After adjustment for age, gender and date of diabetes diagnosis, patients with NAFLD
were at greater risk of death from malignancy (HR 2.40, 95% CI 1.02-5.66, p=0.04).
There was no increase in risk of death from heart disease (HR 0.9, 95% CI 0.34-2.36,
p=0.8) or vascular disease (HR 0.8, 95% CI 0.30-1.96, p=0.6). The hazard ratio for
liver related death could not be calculated as no subjects without NAFLD died from
liver related causes.
5.4 DISCUSSION
It is not surprising that NAFLD is common among subjects with type 2 diabetes
mellitus considering the two conditions share insulin resistance as a common
underlying pathogenic factor.[60,183,184] As NAFLD is common and generally
asymptomatic in these patients, the clinical significance of NAFLD may be overlooked.
However, this study demonstrated that a diagnosis of N A F L D was associated with a
significantly increased risk of death among patients with diabetes mellitus.
Liver related death accounted for 19%> of all deaths of patients with NAFLD and
diabetes whereas no subjects without NAFLD died from liver related causes. This
suggests NAFLD may be directly responsible for a proportion ofthe increased risk of
overall death by leading to cirrhosis with fatal complications of HCC, liver failure and
variceal bleeding. Supporting this is evidence from an Italian population-based cohort
study that found patients with diabetes mellitus had an increased risk of death from
cirrhosis compared to the general population with a SMR of 2.5 (95% C.I. 1.96-
3.20). [272] The authors reported that the relative risk of death from cirrhosis was the
second highest after death from diabetes and was higher than the risk of death from
cardiovascular complications (SMR 1.34, 95% C.I. 1.23-1.44). Given the propensity for
diabetics to develop NAFLD, it is likely that a significant proportion ofthe liver related
deaths in this study were secondary to NAFLD.
Patients with NAFLD were also at greater risk of dying from malignancy
compared to subjects without NAFLD. Similarly, the proportion of deaths due to cancer
among community-diagnosed patients with NAFLD in Chapter Two, was moderately
higher than the general population (28% vs. 24.5%). This is similar to findings from a
Danish population-based study which compared cancer mortality rates in 1800 subjects
discharged from hospital with a diagnosis of NAFLD compared to the general
population. [122] Although this study has several methodological issues as discussed in
Chapter One, it found an increased risk of hepatic and non-hepatic related malignancy
among subjects with NAFLD.
The potential mechanisms through which N A F L D may lead to an increased risk
of malignancy are unclear. NAFLD associated cirrhosis may be complicated by HCC
and this occurred in one patient, however this patient was included among the liver-
related deaths rather than deaths secondary to malignancy. Subjects with NAFLD in
this study were significantly heavier and had a higher prevalence of obesity than
subjects without NAFLD. Obesity is an established risk factor for a range of
malignancies although the underlying mechanisms leading to the increased risk are
unknown. [194] One underlying abnormality common to NAFLD and obesity is
hyperinsulinemia and insulin resistance. In fact, hepatic steatosis increases hepatic
insulin resistance and may exacerbate hyperinsulinemia.[78,240] In vitro studies have
demonstrated that insulin is mitogenic to colonic mucosa [273] and hyperinsulinemia
has recently been identified as an independent risk factor for colorectal malignancy at a
population level. [274] Subsequently, the metabolic features associated with NAFLD
may have increased the risk of malignancy in our cohort.
One ofthe strengths ofthe study is that it is population based and thereby
minimizes the potential for selection bias and increases the ability to generalize the
results to the community. Ascertainment was maximized by using two sources to
identify potential patients - the REP and the MLIS. In addition, the REP allows
evaluation of prospectively collected data and permits a long period of follow-up.
Inherent in any medical record based review is the potential for the clinician to not
record or miss the diagnosis of NAFLD. This was minimized by also examining
biochemical results overtime. Although it is well recognized that liver enzymes may be
normal among patients with NAFLD,[275] it is also well known that liver enzymes
fluctuate over time among subjects with NAFLD[276] and thus the chance of
misclassifying an undiagnosed patient with NAFLD was minimized.
In summary, the diagnosis of NAFLD has an adverse prognostic influence on the
natural history of patients with diabetes mellitus. The risk of death was increased 63%
among individuals who had a concomitant diagnosis of NAFLD. Liver related death
was responsible for a substantial proportion of deaths among patients with NAFLD.
Therefore patients with diabetes mellitus should be evaluated for evidence of NAFLD
and considered for therapy. Further research is required to assess whether treatment of
NAFLD leads to a reduction in liver and overall mortality rates among patients with
diabetes. "
CHAPTER SIX
GENERAL DISCUSSION
6.1 FINDINGS, SIGNIFICANCE AND IMPLICATIONS
This thesis has added significantly to the knowledge ofthe natural history of NAFLD
by examining the clinical and histological evolution of NAFLD in the largest well-
defined cohort of patients published to date. It has established that the diagnosis of
NAFLD is associated with an increased risk of death in the general population as well
as among subjects with type 2 diabetes. Importantly, this was due at least in part, to
mortality from liver disease. NAFLD was found to result in progressive liver fibrosis
with considerable variability in the speed of histological progression. Other
histological features such as steatosis and necro-inflammatory activity decreased over
time, as did serum aminotransaminases. Adverse prognostic factors affecting the
natural history of subjects with NAFLD were clinical associations of insulin resistance,
namely obesity, diabetes, hypothalamic dysfunction and hypopituitarism.
A diagnosis of NAFLD among community-based subjects was associated with a
modestly increased (1.3 fold) risk of all-cause death compared to the general population
of comparable age and gender (see Chapter Two). A portion ofthe increased risk could
reasonably be accounted for by death due to liver disease, as this caused 13% of all
deaths in NAFLD patients as compared to 0.9% of deaths in the general population. It
is not clear whether deaths due to vascular disease also contributed to the increased risk
of death among NAFLD patients. The proportion of deaths due to ischemic heart
disease in N A F L D patients was marginally higher than deaths due to heart disease in
the general population (25% vs. 23.5%). This is somewhat surprising given the
relatively high prevalence of vascular risk factors such as obesity and diabetes and
elevated cardiovascular risk scores in patients with NAFLD. [277] However, these
mortality figures are not adjusted for age and gender and thus direct comparisons need
to be made with some caution. In addition, the proportion of deaths from heart disease
in the general population included ischemic and non-ischemic aetiologies. Thus the
mortality rate among NAFLD patients from ischemic causes may be significantly
higher than the general population. However, it is interesting to note that patients with
advanced liver disease have a low prevalence of cardiovascular disease, which is
hypothesized to be related to the accompanying decrease in blood pressure and
cholesterol levels.[278] Thus, some NAFLD patients who developed cirrhosis may
have been 'protected' against death from cardiovascular disease but died from liver
related causes instead. It is important to acknowledge that the vascular risk factors of
diabetes and obesity also lead to NAFLD, which will result in death due to liver disease
in some individuals. Therefore liver-related death from NAFLD, in addition to death
from vascular disease and diabetes, should be included as a part ofthe public health
burden ofthe metabolic conditions of obesity and diabetes.
The increased risk of death among subjects diagnosed with NAFLD is
concerning because the prevalence of NAFLD is as high as 31%. [4] From a community
perspective, the magnitude of morbidity and mortality secondary to NAFLD and other
complications of these metabolic diseases are potentially enormous. These findings
have implications regarding resource allocation for health spending. In particular, it is
likely the costs associated with treatment of N A F L D and its complications are likely to
escalate given that the prevalence of pathogenic factors such as obesity and diabetes
continues to increase rapidly. [181,239] Resources need to be allocated towards primary
prevention strategies aimed at treating metabolic factors to prevent the onset of NAFLD
and community awareness programs should highlight liver disease as yet another
potentially life-threatening complication of obesity and diabetes.
In addition to being a risk factor for death in the general population, a diagnosis
of NAFLD was found to be associated with an increased risk of death among patients
with diabetes mellitus (Chapter Four). This appeared to be because of an increased risk
of death from cirrhosis and malignancy. Thus specialists and general practitioners need
to be aware of these significant findings that highlight the adverse prognosis associated
with a diagnosis of NAFLD in the diabetic population. As NAFLD is generally
asymptomatic and often associated with normal serum aminotransaminases, it may go
undiagnosed until the advanced stages of disease. Therefore, it would be reasonable to
screen patients with diabetes mellitus for NAFLD by ultrasonography. If NAFLD is
confirmed, a liver biopsy should be considered for staging purposes, particularly in the
presence of other adverse prognostic factors such as obesity, where there is a clinical
suspicion of cirrhosis, or in the presence of a AST/ALT ratio greater than one, which is
suggestive of advanced fibrosis. [128] As data from this thesis demonstrates an
increased risk of mortality from liver disease among diabetics with NAFLD, it would
be reasonable to implement treatment strategies in these patients aimed at preventing
the progression of liver disease, particularly if there is evidence of steatohepatitis or
fibrosis on liver biopsy. Patients should be counselled about avoiding excess alcohol,
which m a y accelerate liver damage. In addition, consideration should be given to
introducing an insulin-sensitizing agent or swapping from sulphonylureas to a
thiozoladinedione or metformin as these have demonstrated some benefit in NAFLD in
pilot trials. [279-281]
The absolute rate of cirrhosis and liver related death among NAFLD patients was
found to be relatively low, at 3.1% and 1.7% respectively over a seven and a half year
period (Chapter One). This is substantially lower than the only previous study which
examined outcomes in patients with the full histological spectrum of NAFLD. [115] In
this hospital based study which followed 98 subjects for an average of eight years, the
occurrence of cirrhosis and liver related death was significantly greater at 20% and 9%
respectively. As discussed in Chapter Two, the more than five fold higher rates of
NAFLD related morbidity and mortality in this study relate primarily to selection bias
and inclusion of incident cases of cirrhosis in the cohort. Reliance on these morbidity
and mortality rates would falsely elevate the community health burden related to
NAFLD and lead to incorrect patient prognostication.
Consistent with the slow clinical progression of NAFLD, the histological course
was one of insidious fibrosis progression (Chapter Three). Fibrosis progressed slowly
at an average of approximately one fibrosis stage per decade (when excluding cirrhotics
from the analysis). Assuming a linear rate of fibrosis progression, it would therefore
take a four-decade period from which to progress from no fibrosis to cirrhosis. This
compares with a median three-decade period for similar progression secondary to
hepatitis C.[282] Therefore, NAFLD is a relatively low grade indolent liver disease.
Of concern however, is that the onset of NAFLD may be during childhood in some
individuals.[64] Childhood obesity has doubled over the past decade and currently has a
prevalence between 5-15% in developed countries,[283] with NAFLD present in up to
50% of obese children. [284] Therefore, an increasing proportion ofthe population will
have NAFLD for sufficient duration to be at risk of developing advanced fibrosis. This
may translate into an increase in the incidence of NAFLD cirrhosis in the future.
In contrast to the progressive worsening of fibrosis over time, other histological
features such as steatosis and necroinflammation became less evident. In particular,
some patients with cirrhosis lost all evidence of steatosis. Thus individuals with a
history of obesity and diabetes who present with cryptogenic cirrhosis could reasonably
be considered to have prior NAFLD. Histological features such as glycogenated nuclei
or Mallory's hyaline may also point to previous NAFLD in these patients.[2]
Recognition of previous NAFLD in patients with cryptogenic cirrhosis is important to
gain accurate estimates of disease related morbidity and mortality. In addition, these
patients warrant aggressive management of metabolic disorders post-transplant and
possible use of insulin sensitizing agents due to the high risk of developing recurrent
disease. [142]
The natural history of aminotransaminases over time was also to improve. It is
unlikely this simply represents 'regression to the mean' as the improvement correlated
with the improvement in steatosis and necro-inflammation but not fibrosis. Thus single
arm treatment trials using serum aminotransaminases as endpoints should be interpreted
with caution. Improvement may reflect the natural history of disease rather than
efficacy ofthe treatment agent. Similarly, aminotransaminases are an unreliable tool
for monitoring disease progression. Therefore liver biopsy is currently the only reliable
method available for monitoring disease progression. Results from chapter three
indicated that the chance of observing progressive fibrosis among patients with early
stage disease was maximised after a four-year interval. Therefore to monitor for
disease progression, a repeat liver biopsy should be performed at least four years after
the initial biopsy.
Although the histological progression of NAFLD was found to be relatively slow
and the rates of liver morbidity and mortality relatively low, there was considerable
variability in the course of disease between individuals. For example, some individuals
progressed from no liver fibrosis to cirrhosis in the time span of 9 to 15 years. In
addition, the youngest patient in this thesis to develop NAFLD related cirrhosis was
only 15 years old (Chapter Four). In contrast, some patients were documented to be
histologically non-progressive and some actually had improvement in hepatic fibrosis
over time. Thus, an important part of this thesis was to determine prognostic factors
that may help explain and predict the variability that was observed in the natural history
of NAFLD.
Clinical conditions linked with insulin resistance, namely diabetes, obesity and
hypopituitarism/hypothalamic disease were clearly associated with a more severe
disease course. Diabetes was associated with an increased risk (2.6 fold) of overall
mortality in NAFLD patients. The risk of liver-related death was also higher among
diabetics (4.3%) when compared to those without diabetes (0%>). Similarly, increasing
BMI was associated with an increased risk (1.1 fold per unit kg/m ) of liver-related
death. Diabetes and BMI were also significantly predictive of higher rates of fibrosis
progression (Chapter Three). Finally, hypopituitarism/hypothalamic disease was
associated with an aggressive disease course that frequently and rapidly progressed to
cirrhosis and often resulted in death or required transplantation (Chapter Five).
Identification of diabetes, obesity and hypopituitarism/hypothalamic disease as
prognostic factors affecting the natural history of subjects with NAFLD, has
considerable impact upon the clinical management ofthe liver disease of these patients.
It is clear from this thesis that the vast majority of patients with NAFLD do not develop
end-stage liver disease. Therefore patients without these adverse indicators can be
reassured the likelihood of developing cirrhosis is very low. The need for liver biopsy
to provide further prognostic information based on histological features (NASH vs.
steatosis) is unlikely to be required for these patients. Conversely, a liver biopsy in
patients with these adverse prognostic indicators may further help stratify the risk for
these patients and the requirement for specific treatment aimed at reversing liver
damage.
Subjects without adverse prognostic factors are at low risk of developing end-
stage liver disease and therefore specific pharmaceutical intervention aimed at reversing
their liver disease is unlikely to be of benefit, particularly as treatment is likely to be
chronic, expensive and have potential side-effects. The presence of these prognostic
variables should prompt the clinician to initiate management of these disorders with the
aim to prevent progressive liver disease. For example, weight loss among morbidly
obese individuals induced by bariatric surgery improves insulin sensitivity and
histological features of NASH including hepatic fibrosis. [226] Improving glycemic
control in patients with diabetes mellitus would intuitively appear to decrease the risk of
progressive liver disease, however this has not been systematically examined. Insulin
sensitizing agents such as the thiozoladinediones and metformin, improve histological
features of NAFLD in pilot trials. [279-281] Although patients with diabetes were
excluded from these trials, it would be reasonable to assume that they are likely to also
be efficacious in this population. Specific treatment recommendations in patients with
hypopituitarism/hypothalamic disease is difficult because ofthe relative lack of insight
into disease pathogenesis. However, these patients clearly tend to be insulin resistant
with accompanying obesity, glucose intolerance and dyslipidemia. Therefore
management of these conditions would be a logical first step. The commencement of
insulin sensitizing agents in patients with obesity, diabetes and/or
hypopituitarism/hypothalamic disease is difficult to recommend definitively due to the
absence of randomized controlled trials. However, due to the adverse natural history
identified in these patients, it would be reasonable to trial these agents with close
monitoring or to enter these into subjects into a clinical trial where possible.
6.2 FUTURE DIRECTIONS
A number of variables which help prognosticate the natural history of NAFLD have
been identified in this thesis. These variables are of undoubted benefit to the clinician
in patient counselling and management. However, even with the identification of these
prognostic variables, the prediction of patients who will develop liver related
complications secondary to NAFLD remains a significant clinical challenge. Obesity
and diabetes mellitus are common in the general population and among subjects with
NAFLD. Thus, a relatively large proportion of individuals with NAFLD have these
adverse prognostic indicators, although the absolute risk of developing liver morbidity
and mortality in these individuals is still relatively low. For example, the incidence of
liver related death among diabetics in Chapter Five was only 4.3%> after approximately
eight years of follow-up. In comparison to diabetes and obesity,
hypopituitarism/hypothalamic disease is relatively rare among subjects with NAFLD,
and thus will be of little prognostic use for the majority of patients with NAFLD.
Therefore further prognostic markers would be helpful in order to stratify patients who
are likely to have a more aggressive natural history and thus would benefit from closer
clinical monitoring and specific liver directed therapy.
Histological features offer valuable prognostic information. Simple steatosis is
associated with benign outcomes as observed in Chapters Two and Three, as well as in
other studies. [114,117] In addition, fibrosis stage predicts liver related morbidity and
mortality in patients with chronic liver disease. Performing a liver biopsy is currently
the gold standard for assessing liver histology and was found to be superior to liver
aminotransaminases for monitoring of disease progression. However, liver biopsy has a
number of drawbacks including expense and complications of bleeding, biliary leakage
and death. [285] Furthermore, liver biopsy samples a tiny fraction ofthe whole liver
and is thus subject to sampling variability as well as intra- and inter-observer variability
in its interpretation. [103] A non-invasive marker able to distinguish between simple
steatosis and NASH as well as determine fibrosis stage would be a valuable adjunct for
the clinician in managing NAFLD patients.
The 'Hepascore' is an algorithm of serum markers (bilirubin, gamma glutamyl
transferase, alpha-2 macroglobulin, hyaluronic acid) in combination with age and
gender which has recently been demonstrated to be accurate in predicting the presence
of significant fibrosis (portal fibrosis, bridging fibrosis and cirrhosis) and the absence of
advanced fibrosis (bridging fibrosis and cirrhosis) among patients with chronic hepatitis
C.[286] Application of this score in the NAFLD population would add a useful
prognostic marker in addition to the clinical variables identified in this thesis. Ideally,
the Hepascore would be able to distinguish patients with simple steatosis and NASH, as
well as predict the presence of significant fibrosis. For example, a patient with diabetes
mellitus, obesity and a Hepascore predictive of significant fibrosis would be a candidate
for aggressive management of their metabolic risk factors, regular monitoring and
consideration for a clinical trial or an insulin-sensitizing agent. In contrast, a patient
with diabetes and/or obesity but a Hepascore predictive of simple steatosis may simply
need their metabolic risks addressed and occasional review.
In addition to determining non-invasive markers of liver histology in NAFLD,
further understanding ofthe pathogenesis of disease progression may lead to insight
into ways of stratifying risk of liver morbidity and mortality. The reasons why
diabetes, obesity and hypopituitarism/hypothalamic disease are associated with adverse
outcomes in NAFLD is likely to be related to the underlying metabolic disturbances
associated with these conditions. Insulin resistance is an underlying feature of all of
these conditions and is likely to play an important role in leading to progressive disease.
However, other hormonal disturbances involving growth hormone, adiponectin and
leptin may also be important. As previously reviewed, disordered regulation of these
hormones among subjects with hypopituitarism/hypothalamic disease may have been
responsible for the aggressive course of their liver disease.
Adiponectin remains an intriguing potential modulator of disease progression.
Adiponectin is a hormone almost exclusively produced by adipose tissue. In contrast to
other adipocytokines, levels are decreased rather than increased among subjects with
obesity and diabetes mellitus.[220] Adiponectin has insulin sensitizing actions and also
inhibits pro-inflammatory cytokines such as TNFa. The administration of adiponectin
to animal models of NAFLD ameliorates hepatic steatosis. [205] In addition, animals
subjected to various hepatotoxic agents are protected from developing liver fibrosis in
the presence of adiponectin. [221,222] Among obese humans, adiponectin is inversely
correlated with ALT levels and is also negatively associated with necroinflammatory
activity in patients with NASH. [90,205] Furthermore, expression of adiponectin
mRNA and its type 2 receptor is decreased among patients with NASH compared to
those with benign simple steatosis. Thus adiponectin appears to be insulin sensitizing
and hepatoprotective and reduced levels may promote the development and progression
of NAFLD. Currently the human data examining adiponectin and NAFLD is cross-
sectional and thus it cannot be concluded that there is a cause-effect relationship.
Longitudinal data examining the expression of adiponectin in relation to the clinical and
histological progression of NAFLD would aid in determining its role. It is interesting
to note that the two clinical prognostic risk factors found in this thesis (obesity and
diabetes) are associated with low levels of adiponectin. Thus these factors would need
to be carefully controlled for when examining the prognostic significance of
adiponectin in NAFLD.
Identification of these prognostic clinical factors of obesity, diabetes mellitus and
hypopituitarism/hypothalamic disease should instigate renewed efforts to examine and
dissect the underlying physiological and metabolic disturbances in these patients to gain
greater insight into the pathogenesis of progressive NAFLD. Further knowledge ofthe
pathogenesis of NAFLD will not only help predict the natural history of disease but will
aid in the search to find effective treatment options.
6.3 CONCLUSIONS
With the continuing worldwide increase in obesity and diabetes, general practitioners
and specialists are increasingly encountering patients with NAFLD. Knowledge ofthe
natural history of this disease is vital for directing patient management. In particular,
counselling and prognostication, the need for liver biopsy, monitoring and treatment all
depend on the probability ofthe patient encountering NAFLD related morbidity and/or
mortality. In addition, knowledge of disease morbidity and mortality is important from
a public health perspective as it predicts the potential health burden and guides resource
allocation and need for prevention strategies. Prior studies ofthe natural history of
NAFLD have had significant limitations including selection biases, limited numbers
and relatively short follow-up.
This thesis has substantially advanced the knowledge ofthe natural history of
NAFLD and its impact on the patient. Significantly, a diagnosis of NAFLD in the
general community was found to be associated with an increased risk of all-cause death
which was predicted by glucose intolerance and cirrhosis. Thus it is important for the
clinician to carefully evaluate for these conditions when managing a patient with
NAFLD. From a public health perspective, it is critical to address the rising prevalence
of obesity and diabetes; two pathogenic factors which lead to N A F L D and its associated
risk of death.
The liver related morbidity and mortality of NAFLD was revealed to be lower
than reported in previous studies. NAFLD progressed slowly overall with
complications of cirrhosis, HCC and liver failure occurring in a small proportion of
individuals. It was observed however, that considerable variability existed in the
evolution of NAFLD. Importantly, patients with NAFLD had an increased risk of rapid
disease progression and death if they had co-existing clinical conditions associated with
insulin resistance such as diabetes mellitus, obesity and hypopituitarism/hypothalamic
disease. In addition, patients with diabetes mellitus had an increased risk of overall and
liver-related death if they had co-existing NAFLD. Thus it is important to raise
awareness among clinicians ofthe risk of NAFLD related morbidity and mortality in
patients with these conditions. Specific therapy should be considered in these patients
with the aim of preventing progressive liver disease.
APPENDIX ONE
HICDA codes utilized for ascertainment of subjects in Olmsted County diagnosed with
N A F L D (1980-2000).
Inclusion Diagnoses H I C D A Code
Fatty liver, steatohepatitis
Steatosis
Cryptogenic cirrhosis
05710-42-43
02790-44-1
05712-16-1
Exclusion Diagnoses H I C D A Code
Viral hepatitis B
Viral hepatitis C
Abnormal serology for hepatitis B,C,D,E
Chronic hepatitis B and C
Alcohol abuse
Alcoholism
Alcoholic cirrhosis
Alcoholic liver disease
Cholestatic liver disease
Primary biliary cirrhosis
Primary sclerosing cholangitis
Haemochromatosis
Wilson's disease
Alpha one anti-trypsin deficiency
Auto-immune hepatitis
00701
00709
34728-12-2 through 5
05730-31-3 and 4
03139
03132
05710-11-0 though 5
05710-12-13-40
05769-11
05711
05760-31
02732
02733-11-0
02739-11
05732-25
APPENDIX TWO
Data abstraction sheet for Chapter Two: Residents of Olmsted County diagnosed with
NAFLD between 1980-2000.
For all categories, X = not done or unknown Demoqraph
v1 - v7
v8-v17
v18-v25
v26
v27
Islander,
2= Black, 3=
ics
/ /
Hispanic, 4=White,
5= American Indian/Alaskan, 6= Other
v28
Clinical Details at time of Diagnosis
V29 - v31
V32
1=Yes
v33
v34
v35 - v38
v39
v40 - v42
v43
. kg/m2
_̂ _̂
_̂ _ ._— — —̂—
gm/weekET
v44 years
v45
Mayo Clinic Number
OMC
DOB (mm/dd/yyyy)
Gender 1= Male, 2= Female
Race 1 = Asian / Pacific
Other
Body Mass Index (BMI)
Stigmata of Liver Disease 0 = No,
Tattoo 0=No, 1=Yes
Blood Transfusion 0=No, 1=Yes
Year if yes.
IVDU 0=No, 1=Yes
O H intake
Diabetes 0= No, 1=IGT
2= Diet controlled, 3= Oral
hypoglycemics, 4= Insulin requiring
type 2, 5= Type 1
Duration of diabetes
Hyperlipidemic 0= No, 1= Yes
v46 Hypertensive 0= No, 1=Yes
v47 0= No abdo discomfort, 1 = Abdo discomfort
v48 0= No fatigue, 1= Fatigue
v49 0 = Nonsmoker, 1= Exsmoker, 2= current smoker
v50 Ischemic Heart Disease 0=No,1=Yes
v51 Cerebrovascular Disease 0=No,1=Yes
v52 Chronic renal failure 0=No,1=Yes
v53 Previous/current malignancy 0= No, 1= Yes
v54 Type. 1= Bowel, 2= Lung, 3=Hematological, 4=prostate,
5=breast, 6= skin, 7=other
v55 Other
v56 - v63 / / Date of diagnosis (MM/DD/YYYY)
Laboratory Data at time of diagnosis
v64 Liver biopsy 0= No, 1=Yes
v65 Reason for Biopsy 1= abnormal LFTs,
2= hepatomegaly, 3= suspicion of cirrhosis
4= abnormal appearance at surgery, 5= Staging
6= Post mortem, 7= Other
v66 Other
v67 - v74 / / Date of liver biopsy
(MM/DD/YYYY)
v75 Steatosis grade 1-3
v76 Inflammatory grade 0-3
v77 _ Fibrosis stage 0-4
v78 - v79 _. Hepatic Iron Index
v80 - v83 Hepatic Iron Concentration
V84 _ Hepatic Iron Stain 0= Negative, 1= Positive
v85 - 87 mcg/gm liver Copper dry weight
V88 __ Imaging 0= No, 1= Ultrasound, 2= CT, 3= MRI
v89-v96 / / Date of Imaging (MM/DD/YYYY)
V97 Findings 0= Normal, 1= Fatty infliltration
V98 0= No ascites, 1= Ascites
v99 0= No Splenomegaly, 1= Splenomegaly
V100 0= Normal contour, 1= Nodular liver
V101 0= No mass, 1= mass
V102-V103 . mg/dL Bilirubin
V104-V106 U/L AST
v107-v109 U/L ALT
v110 - v113 U/L Alk. Phosphatase
v114-v115 . g/dL Albumin
v116-v118 . sec. Prothrombin time
v119-v122 mcg/L Ferritin
v123 - v124 % Transferrin saturation
v125-v127 G a m m a globulin level
v128-v130 mg/dL Immunoglobulin G
v131-v133 mg/dL Serum glucose
v134 - v136 mg/dL Total Cholesterol
v137-v140 mg/dL Triglyceride
vl41-v143 mg/dL HDL cholesterol
Laboratory Data at anytime v144 HBV serology 1= sAb +ve, 2= sAg +ve, 3= sAg -ve
v145 _ Hepatitis C 1= Ab-ve ,2= Ab or RNA +ve
v146-v148 . mg/dL Ceruloplasmin
v149-v153 : Anti-nuclear Ab
v154-v157 : Anti-smooth muscle Ab
v158-v161 : Anti-mitochondrial Ab
v162 - v164 mg/dL Alpha-one antitrypsin level
V 1 6 5 Alpha-one antitrypsin phenotype 1=ZZ, 2=Other
V166 Liver biopsy Hepatocyte ballooning 1= yes, 2= no
V167 Liver biopsy Mallorys hyaline 1= yes, 2= no
V 1 6 8 x109/L Platelet count
v169-176 / / Date of first abnormal liver test
(MM/DD/YYYY)
Follow-up Data
v1 -v8 • / / Date of last follow-up (MM/DD/YYYY)
v9 Death 0= no, 1 = yes
v10-v17 / / DateofDeath
v18 Cause of death 1= liver failure, 2= variceal bleed,3= HCC,
4= other malignancy, 5= IHD, 6= CVD, 7= Infection, 8= other, 9= don't know
v19 Other
v20 Cirrhosis 0= no, 1 = biopsy proven, 2= clinical
diagnosis
v21-v28 / / Date of diagnosis of cirrhosis
v29 Ascites 0= no, 1=yes
v30 - v37 / / Date of first diagnosis of ascites
v38 Hepatic encephalopathy 0= no, 1=yes
v39 - v46 / / Date of first diagnosis of encephalopathy
v47 Variceal haemorrhage 0= no, 1=yes
v48 - v55 / / Date of first variceal haemorrhage
v56 Jaundice 0= no, 1=yes
v57 - v63 / / Date of first diagnosis of jaundice
v65 Hepatocellular carcinoma 0= no, 1=yes
v66 - v73 / / Date of diagnosis of H C C
v74 Liver transplantation 0= no, 1=yes
v75 - v82 / / Date of liver transplantation
v83 New Diagnosis of Diabetes 0=no, 1=yes
V84-91 / / Date
v 92 New Diagnosis of Hypertension 0=no,
1=yes
v93 -100 / / Date
v 101 New Diagnosis of Hyperlipidemia 0=no,
1=yes
V102-109 / / Date
APPENDIX THREE
Letter sent to subjects requesting of authorization to access to their liver biopsy
specimen from an outside institution.
Division of Gastroenterology and Hepatology Mayo Clinic 200 First Street S W Rochester M N 55905 USA Date
Dear
Our records show that you have had at least two liver biopsies, which have shown fat in the liver. This is now known as nonalcoholic fatty liver disease (NAFLD). W e are performing a research study to see if liver biopsies in patients with N A F L D change over time. This is important as it will help us determine whether having repeat biopsies is useful.
At least one of your biopsies was performed at a medical facility other than Mayo Clinic. W e are asking for your authorization for Mayo Clinic to borrow the biopsy specimen from , where it is currently located. W e would review the biopsy and return it once w e have finished. W e would have the biopsy slides for approximately one month. W e will only review the liver biopsy slide. N o other medical or personal information will be requested. Information provided from review ofthe biopsy will be kept strictly confidential and your privacy will be maintained. Any future care at Mayo Clinic will not be influenced by your decision. There is no charge for allowing us to use this information.
If you authorize the release ofthe liver biopsy to Mayo Clinic for review, please sign the attached medical release form and the 'Authorization to Use and Disclose Protected Health Information' form, and return them in the accompanying stamped and addressed envelope. W e sincerely hope that you are willing to take part in this study.
Thank you for your time.
Yours sincerely
Dr Leon Adams M.B.B.S. & Dr Paul Angulo M.D. Division of Gastroenterology and Hepatology
APPENDIX FOUR
Specific H I C D A codes used to identify subjects with hypopituitarism, hypothalamic
disease and craniopharyngioma
Inclusion Diagnoses HICDA Code
Hypopituitarism 02262-16-17
Hypothalamic Obesity 01943-22
Craniopharyngioma 02531-41 to 44 and 62
APPENDIX FIVE
H I C D A codes used to ascertain patients with a diagnosis of diabetes mellitus
Inclusion Diagnoses H I C D A Code
Diabetes mellitus - not otherwise specified; insulin
dependent; diet controlled; hypoglycemia adverse event;
impaired hypoglycemic awareness.
Diabetes mellitus suspect
Hyperglycemia
Diabetes mellitus - neurological complications
Diabetes mellitus - ophthalmologic complication
Diabetes mellitus - ulcer
Diabetes mellitus - gangrene
Diabetes mellitus - necrobiosis
Diabetes mellitus complications - dermatitis, lipoatrophy,
cardiomyopathy.
Diabetes mellitus - glomerulosclerosis, nephropathy,
glomerulonephritis
Diabetes mellitus - latent
Diabetes mellitus - diarrhoea
Adverse effect - antidiabetic agent; insulin
Abnormal serum glucose, abnormal glucose tolerance test,
abnormal glucose
0250-0110 to 0250-0118
02500120
02500212,0250-0213
02502110 to 0250-2115
0250-2210
0250-2310
0250-2410
0250-2420
0250-2610 to 0250-2614
0250-6110,0250-6140,
0250-6150,0377-4410,
0377-4420
0250-7110
0782-1120
0962-3110,0962-3211
3472-1000,3472-1110,
3472-1111
A P P E N D I X SIX
Data abstraction sheet for Chapter Five: Residents of Olmsted County diagnosed with
diabetes mellitus between 1980-2000.
Demographics
- - Mayo Clinic Number
O M C
_ _ / _ _ / DOB (mm/dd/yyyy)
Gender 1 = Male, 2= Female
Race 1 = Asian / Pacific Islander, 2= Black,
3= Hispanic, 4=White, 5= American
Indian/Alaskan, 6= Other
Other
Clinical Details at time of Diagnosis of diabetes
_/_ _/ Date of diagnosis of diabetes
Type of Diabetes 0= No, 1= IGT, 2= Diet controlled, 3= Oral
hypoglycemics, 4= Insulin requiring type 2,
5= Type 1.
Type of oral hypoglycemics
Hyperlipidemic 0= No, 1=Yes
Lipid lowering medications 0= No, 1= Yes
Type
_ Hypertensive 0= No, 1=Yes
mmHg SBP
_mmHg DBP
Anti-hypertensive medications 0= No, 1= Yes
Type
. kg/m2 Body Mass Index (BMI)
gm/day ETOH intake
0 = Nonsmoker, 1= Exsmoker, 2= current smoker
Ischemic Heart Disease 0=No,1=Yes
Cerebrovascular Disease 0=No,1=Yes
Chronic renal failure 0=No,1=Yes
Previous/current malignancy 0=No,1=Yes
Type. 1= Bowel, 2= Lung, 3=Hematological, 4=prostate,
5=breast, 6= skin, 7=other
Other
/ / Date of cancer diagnosis (MM/DD/YYYY)
Laboratory Data at time of diagnosis of diabetes (within 3 months)
_ . mg/dL Bilirubin
U/L AST
U/L ALT
U/L Alk. Phosphatase
_. __ g/dL Albumin
sec. Prothrombin time
mcg/L Ferritin
_ % Transferrin saturation
mg/dL Serum glucose
_.__% HbA1c
mg/dL Total Cholesterol
_ mg/dL Triglyceride
mg/dL HDL cholesterol
x109/L Platelet count
Elevated AST due to non-NAFLD, 0=no, 1=yes.
Imaging date showing no hepatic fat
(MM/DD/YYYY) 1=US, 2=CT,
3=MRI
(MM/DD/YYYY) 1=US, 2=CT,
3=MRI
Normal Aminotransaminase Levels every 6 months
Total normal AST readings Total normal ALT readings
Follow-up Data
v1 -v8 / / Date of last follow-up(MM/DD/YYYY)
v9 Death 0= no, 1=yes
v10-v17 / / Date of Death
v18 Cause of death 1= diabetes, 2= cardiovascular disease,
3= cerebrovascular disease, 4= malignancy, 5= infection 6=
chronic renal failure, 7= other, 8= don't know
v19 Other
1
REFERENCES
1. Fletcher RF, SW. Wagner,EH. Clinical Epidemiology/The Essentials. Second ed.
Baltimore: Lippincott, Williams and Wilkins; 1996.
2. Caldwell SH, Crespo DM. The spectrum expanded: cryptogenic cirrhosis and the
natural history of non-alcoholic fatty liver disease. J Hepatol 2004;40:578-84.
3. Neuschwander-Tetri BA, Caldwell SH. Nonalcoholic steatohepatitis: summary of an
AASLD Single Topic Conference. Hepatology 2003;37:1202-19.
4. Browning JD, Szczepaniak LS, Dobbins R, Nuremberg P, Horton JD, Cohen JC,
Grundy SM, Hobbs HH. Prevalence of hepatic steatosis in an urban population in the
United States: impact of ethnicity. Hepatology 2004;40:1387-95.
5. Lieber CS. Medical disorders of alcoholism. N Engl J Med 1995;333:1058-65.
6. Sanyal AJ. AGA technical review on nonalcoholic fatty liver disease.
Gastroenterology 2002; 123:1705-25.
7. Becker U, Deis A, Sorensen TI, Gronbaek M, Borch-Johnsen K, Muller CF, Schnohr
P, Jensen G. Prediction of risk of liver disease by alcohol intake, sex, and age: a
prospective population study. Hepatology 1996;23:1025-9.
8. Bellentani S, Saccoccio G, Costa G, Tiribelli C, Manenti F, Sodde M, Saveria Croce
L, Sasso F, Pozzato G, Cristianini G, Brandi G. Drinking habits as cofactors of risk for
alcohol induced liver damage. The Dionysos Study Group. Gut 1997;41:845-50.
9. Farrell GC. Drugs and steatohepatitis. Semin Liver Dis 2002;22:185-94.
10. Boldrini R, Devito R, Biselli R, Filocamo M , Bosman C. Wolman disease and
cholesteryl ester storage disease diagnosed by histological and ultrastructural
examination of intestinal and liver biopsy. Pathol Res Pract 2004;200:231-40.
11. Lombardi B, Pani P, Schlunk FF. Choline-deficiency fatty liver: impaired release
of hepatic triglycerides. J Lipid Res 1968;9:437-46.
12. Guntupalli SR, Steingrub J. Hepatic disease and pregnancy: an overview of
diagnosis and management. Crit Care Med 2005;33:S332-9.
13. Broussard CN, Aggarwal A, Lacey SR, Post AB, Gramlich T, Henderson JM,
Younossi ZM. Mushroom poisoning—from diarrhea to liver transplantation. Am J
Gastroenterol 2001;96:3195-8.
14. Angulo P. Nonalcoholic fatty liver disease. N Engl J Med 2002;346:1221-31.
15. Hamaguchi M, Kojima T, Takeda N, Nakagawa T, Taniguchi H, Fujii K, Omatsu
T, Nakajima T, Sarui H, Shimazaki M, Kato T, Okuda J, Ida K. The metabolic
syndrome as a predictor of nonalcoholic fatty liver disease. Ann Intern Med
2005;143:722-8.
16. Marchesini G, Bugianesi E, Forlani G, Cerrelli F, Lenzi M, Manini R, Natale S,
Vanni E, Villanova N, Melchionda N, Rizzetto M. Nonalcoholic fatty liver,
steatohepatitis, and the metabolic syndrome. Hepatology 2003;37:917-23.
17. Dourakis SP, Sevastianos VA, Kaliopi P. Acute severe steatohepatitis related to
prednisolone therapy. Am J Gastroenterol 2002;97:1074-5.
18. Nemoto Y, Saibara T, Ogawa Y, Zhang T, Xu N, Ono M, Akisawa N, Iwasaki S,
Maeda T, Onishi S. Tamoxifen-induced nonalcoholic steatohepatitis in breast cancer
patients treated with adjuvant tamoxifen. Intern Med 2002;41:345-50.
19. Hamada N, Ogawa Y, Saibara T, Murata Y, Kariya S, Nishioka A, Terashima M ,
Inomata T, Yoshida S. Toremifene-induced fatty liver and NASH in breast cancer
patients with breast-conservation treatment. Int J Oncol 2000; 17:1119-23.
20. Cooper L, Palmer M, Oien K. Cirrhosis with steatohepatitis following longterm
stilboestrol treatment. J Clin Pathol 2003;56:639.
21. Babany G, Uzzan F, Larrey D, Degott C, Bourgeois P, Rene E, Vissuzaine C,
Erlinger S, Benhamou JP. Alcoholic-like liver lesions induced by nifedipine. J Hepatol
1989;9:252-5.
22. Arranto AJ, Sotaniemi EA. Histologic follow-up of alpha-methyldopa-induced
liver injury. Scand J Gastroenterol 1981;16:865-72.
23. Miller KD, Cameron M, Wood LV, Dalakas MC, Kovacs JA. Lactic acidosis and
hepatic steatosis associated with use of stavudine: report of four cases. Ann Intern Med
2000;133:192-6.
24. Freneaux E, Labbe G, Letteron P, The Le D, Degott C, Geneve J, Larrey D,
Pessayre D. Inhibition ofthe mitochondrial oxidation of fatty acids by tetracycline in
mice and in man: possible role in micro vesicular steatosis induced by this antibiotic.
Hepatology 1988;8:1056-62.
25. Chitturi S, George J. Hepatotoxicity of commonly used drugs: nonsteroidal anti
inflammatory drugs, antihypertensives, antidiabetic agents, anticonvulsants, lipid-
lowering agents, psychotropic drugs. Semin Liver Dis 2002;22:169-83.
26. Lewis JH, Ranard RC, Caruso A, Jackson LK, Mullick F, Ishak KG, Seeff LB,
Zimmerman HJ. Amiodarone hepatotoxicity: prevalence and clinicopathologic
correlations among 104 patients. Hepatology 1989;9:679-85.
27. Pieterse A S , Rowland R, Dunn D. Perhexiline maleate induced cirrhosis. Pathology
1983;15:201-3.
28. Langman G, Hall PM, Todd G. Role of non-alcoholic steatohepatitis in
methotrexate-indueed liver injury. J Gastroenterol Hepatol 2001;16:1395-401.
29. Fernandez FG, Ritter J, Goodwin JW, Linehan DC, Hawkins WG, Strasberg SM.
Effect of steatohepatitis associated with irinotecan or oxaliplafin pretreatment on
resectability of hepatic colorectal metastases. J Am Coll Surg 2005;200:845-53.
30. Lonardo A, Adinolfi LE, Loria P, Carulli N, Ruggiero G, Day CP. Steatosis and
hepatitis C virus: mechanisms and significance for hepatic and extrahepatic disease.
Gastroenterology 2004; 126:586-97.
31. Ristig M, Drechsler H, Powderly WG. Hepatic steatosis and HIV infection. AIDS
Patient Care STDS 2005;19:356-65.
32. Nazim M, Stamp G, Hodgson HJ. Non-alcoholic steatohepatitis associated with
small intestinal diverticulosis and bacterial overgrowth. Hepatogastroenterology
1989;36:349-51.
33. Lanza-Jacoby S, Rosato EL. Regulatory factors in the development of fatty
infiltration ofthe liver during gram-negative sepsis. Metabolism 1994;43:691-6.
34. Tarugi P, Lonardo A, Ballarini G, Erspamer L, Tondelli E, Bertolini S, Calandra S.
A study of fatty liver disease and plasma lipoproteins in a kindred with familial
hypobetalipoproteinemia due to a novel truncated form of apolipoprotein B (APO B-
54.5). J Hepatol 2000;33:361-70.
35. Garg A. Acquired and inherited lipodystrophies. N Engl J Med 2004;350:1220-34.
36. Amarapurkar D N , Patel N D , Amarapurkar A D . Panniculitis and liver disease
(hepatic Weber Christian disease). J Hepatol 2005;42:149-50.
37. Glasgow JF, Middleton B. Reye syndrome—insights on causation and prognosis.
Arch Dis Child 2001;85:351-3.
38. Roberts EA, Schilsky ML. A practice guideline on Wilson disease. Hepatology
2003;37:1475-92.
39. Srinivasan R, Hadzic N, Fischer J, Knisely AS. Steatohepatitis and unsuspected
micronodular cirrhosis in Dorfman-Chanarin syndrome with documented ABHD5
mutation. J Pediatr 2004;144:662-5.
40. Takagi H, Hagiwara S, Hashizume H, Kanda D, Sato K, Sohara N, Kakizaki S,
Takahashi H, Mori M, Kaneko H, Ohwada S, Ushikai M, Kobayashi K, Saheki T. Adult
onset type II citrullinemia as a cause of non-alcoholic steatohepatitis. J Hepatol
2006;44:236-9.
41. Lundqvist G, Flodin U, Axelson O. A case-control study of fatty liver disease and
organic solvent exposure. Am J Ind Med 1999;35:132-6.
42. Cotrim HP, Carvalho F, Siqueira AC, Lordelo M, Rocha R, De Freitas LA.
Nonalcoholic fatty liver and insulin resistance among petrochemical workers. Jama
2005;294:1618-20.
43. Mahler H, Pasi A, Kramer JM, Schulte P, Scoging AC, Bar W, Krahenbuhl S.
Fulminant liver failure in association with the emetic toxin of Bacillus cereus. N Engl J
Med 1997;336:1142-8.
44. Holzbach RT. Hepatic effects of jejunoileal bypass for morbid obesity. Am J Clin
Nutr 1977;30:43-52.
45. Abdo A, Meddings J, Swain M . Liver abnormalities in celiac disease. Clin
Gastroenterol Hepatol 2004;2:107-12.
46. Bargiggia S, Maconi G, Kill M, Molteni P, Ardizzone S, Parente F, Todaro I, Greco
S, Manzionna G, Porro GB. Sonographic prevalence of liver steatosis and biliary tract
stones in patients with inflammatory bowel disease: study of 511 subjects at a single
center. J Clin Gastroenterol 2003;36:417-20.
47. Quigley EM, Marsh MN, Shaffer JL, Markin RS. Hepatobiliary complications of
total parenteral nutrition. Gastroenterology 1993;104:286-301.
48. Buchman AL, Dubin MD, Moukarzel AA, Jenden DJ, Roch M, Rice KM,
Gornbein J, Ament ME. Choline deficiency: a cause of hepatic steatosis during
parenteral nutrition that can be reversed with intravenous choline supplementation.
Hepatology 1995;22:1399-403.
49. Mottin CC, Moretto M, Padoin AV, Swarowsky AM, Toneto MG, Glock L,
Repetto G. The role of ultrasound in the diagnosis of hepatic steatosis in morbidly
obese patients. Obes Surg 2004;14:635-7.
50. Joy D, Thava VR, Scott BB. Diagnosis of fatty liver disease: is biopsy necessary?
Eur J Gastroenterol Hepatol 2003;15:539-43.
51. Clark JM, Brancati FL, Diehl AM. The prevalence and etiology of elevated
aminotransferase levels in the United States. Am J Gastroenterol 2003;98:960-7.
52. Ruhl CE, Everhart JE. Determinants ofthe association of overweight with elevated
serum alanine aminotransferase activity in the United States. Gastroenterology
2003;124:71-9.
53. Weston SR, Leyden W , Murphy R, Bass N M , Bell BP, Manos M M , Terrault NA.
Racial and ethnic distribution of nonalcoholic fatty liver in persons with newly
diagnosed chronic liver disease. Hepatology 2005;41:372-9.
54. Browning JD, Kumar KS, Saboorian MH, Thiele DL. Ethnic differences in the
prevalence of cryptogenic cirrhosis. Am J Gastroenterol 2004;99:292-8.
55. Ryan CK, Johnson LA, Germin Bl, Marcos A. One hundred consecutive hepatic
biopsies in the workup of living donors for right lobe liver transplantation. Liver
Transpl 2002;8:1114-22.
56. Hilden M, Christoffersen P, Juhl E, Dalgaard JB. Liver histology in a 'normal'
population—examinations of 503 consecutive fatal traffic casualties. Scand J
Gastroenterol 1977;12:593-7.
57. Ground KE. Liver pathology in aircrew. Aviat Space Environ Med 1982;53:14-8.
58. Nomura H, Kashiwagi S, Hayashi J, Kajiyama W, Tani S, Goto M. Prevalence of
fatty liver in a general population of Okinawa, Japan. Jpn J Med 1988;27:142-9.
59. Shen L, Fan JG, Shao Y, Zeng MD, Wang JR, Luo GH, Li JQ, Chen SY.
Prevalence of nonalcoholic fatty liver among administrative officers in Shanghai: an
epidemiological survey. World J Gastroenterol 2003;9:1106-10.
60. Jimba S, Nakagami T, Takahashi M, Wakamatsu T, Hirota Y, Iwamoto Y, Wasada
T. Prevalence of non-alcoholic fatty liver disease and its association with impaired
glucose metabolism in Japanese adults. Diabetic Med 2005.
61. Fan JG, Zhu G, Li XJ, Chen L, Li L, Dai F, Li F, Chen SY. Prevalence of and risk
factors for fatty liver in a general population of Shanghai China. J Hepatol
2005;43:508-13.
62. Kim H C , Choi SH, Shin H W , Cheong JY, Lee K W , Lee H C , Huh K B , Kim DJ.
Severity of ultrasonographic liver steatosis and metabolic syndrome in Korean men and
women. World J Gastroenterol 2005; 11:5314-21.
63. Yano E, Tagawa K, Yamaoka K, Mori M. Test validity of periodic liver function
tests in a population of Japanese male bank employees. J Clin Epidemiol 2001;54:945-
51.
64. Tominaga K, Kurata JH, Chen YK, Fujimoto E, Miyagawa S, Abe I, Kusano Y.
Prevalence of fatty liver in Japanese children and relationship to obesity. An
epidemiological ultrasonographic survey. Dig Dis Sci 1995;40:2002-9.
65. Kagansky N, Levy S, Keter D, Rimon E, Taiba Z, Fridman Z, Berger D, Knobler
H, Malnick S. Non-alcoholic fatty liver disease—a common and benign finding in
octogenarian patients. Liver Int 2004;24:588-94.
66. Wanless IR, Lentz JS. Fatty liver hepatitis (steatohepatitis) and obesity: an autopsy
study with analysis of risk factors. Hepatology 1990; 12:1106-10.
67. Dixon JB, Bhathal PS, O'Brien PE. Nonalcoholic fatty liver disease: predictors of
nonalcoholic steatohepatitis and liver fibrosis in the severely obese. Gastroenterology
2001;121:91-100.
68. Abrams G, Kunde S, Lazenby A, Clements R. Portal fibrosis and hepatic steatosis
in morbidly obese subjects: a spectrum of nonalcoholic fatty liver disease. Hepatology
2004;40:475-83.
69. Dallal R, Mattar S, Lord J, Watson A, Cottom D, Eid G, Hamad G, Rabinovitz M,
Schauer P. Results of Japaroscopic gastric bypass in patients with cirrhosis. Obes Surg
2004;14:47-53.
70. Kojima S, Watanabe N, Numata M , Ogawa T, Matsuzaki S. Increase in the
prevalence of fatty liver in Japan over the past 12 years: analysis of clinical
background. J Gastroenterol 2003;38:954-61.
71. Ioannou GN, Boyko EJ, Lee SP. The prevalence and predictors of elevated serum
aminotransferase activity in the United States in 1999-2002. Am J Gastroenterol
2006;101:76-82.
72. Adams LA, Angulo P, Lindor KD. Nonalcoholic fatty liver disease. Cmaj
2005;172:899-905.
73. Browning JD, Horton JD. Molecular mediators of hepatic steatosis and liver injury.
J Clin Invest 2004;114:147-52.
74. Azzout-Marniche D, Becard D, Guichard C, Foretz M, Ferre P, Foufelle F. Insulin
effects on sterol regulatory-element-binding protein-lc (SREBP-lc) transcriptional
activity in rat hepatocytes. Biochem J 2000;350 Pt 2:389-93.
75. Charlton M, Sreekumar R, Rasmussen D, Lindor K, Nair KS. Apolipoprotein
synthesis in nonalcoholic steatohepatitis. Hepatology 2002;35:898-904.
76. Letteron P, Sutton A, Mansouri A, Fromenty B, Pessayre D. Inhibition of
microsomal triglyceride transfer protein: another mechanism for drug-induced steatosis
in mice. Hepatology 2003;38:133-40.
77. Schattenberg JM, Wang Y, Singh R, Rigoli RM, Czaja MJ. Hepatocyte CYP2E1
overexpression and steatohepatitis lead to impaired hepatic insulin signaling. J Biol
Chem 2005.
78. Samuel V T , Liu ZX, Q u X, Elder B D , Bilz S, Befroy D, Romanelli AJ, Shulman
GL Mechanism of hepatic insulin resistance in non-alcoholic fatty liver disease. J Biol
Chem 2004;279:32345-53.
79. Pagano G, Pacini G, Musso G, Gambino R, Mecca F, Depetris N, Cassader M,
David E, Cavallo-Perin P, Rizzetto M. Nonalcoholic steatohepatitis, insulin resistance,
and metabolic syndrome: further evidence for an etiologic association. Hepatology
2002;35:367-72.
80. Suzuki A, Angulo P, Lymp J, St Sauver J, Muto A, Okada T, Lindor K.
Chronological development of elevated aminotransferases in a nonalcoholic population.
Hepatology 2005;41:64-71.
81. Day CP, James OF. Steatohepatitis: a tale of two "hits"? Gastroenterology
1998;114:842-5.
82. Haque M, Sanyal AJ. The metabolic abnormalities associated with non-alcoholic
fatty liver disease. Best Pract Res Clin Gastroenterol 2002;16:709-31.
83. Feldstein AE, Canbay A, Angulo P, Taniai M, Burgart LJ, Lindor KD, Gores GJ.
Hepatocyte apoptosis and fas expression are prominent features of human nonalcoholic
steatohepatitis. Gastroenterology 2003;125:437-43.
84. Feldstein AE, Werneburg NW, Canbay A, Guicciardi ME, Bronk SF, Rydzewski
R, Burgart LJ, Gores GJ. Free fatty acids promote hepatic lipotoxicity by stimulating
TNF-alpha expression via a lysosomal pathway. Hepatology 2004;40:185-94.
85. Valenti L, Fracanzani AL, Dongiovanni P, Santorelli G, Branchi A, Taioli E,
Fiorelli G, Fargion S. Tumor necrosis factor alpha promoter polymorphisms and insulin
resistance in nonalcoholic fatty liver disease. Gastroenterology 2002;122:274-80.
86. Namikawa C, Shu-Ping Z, Vyselaar JR, Nozaki Y, Nemoto Y, Ono M , Akisawa N,
Saibara T, Hiroi M, Enzan H, Onishi S. Polymorphisms of microsomal triglyceride
transfer protein gene and manganese superoxide dismutase gene in non-alcoholic
steatohepatitis. J Hepatol 2004;40:781-6.
87. Dixon JB, Bhathal PS, Jonsson JR, Dixon AF, Powell EE, O'Brien PE. Pro-fibrotic
polymorphisms predictive of advanced liver fibrosis in the severely obese. J Hepatol
2003;39:967-71.
88. Wigg AJ, Roberts-Thomson IC, Dymock RB, McCarthy PJ, Grose RH, Cummins
AG. The role of small intestinal bacterial overgrowth, intestinal permeability,
endotoxaemia, and tumour necrosis factor alpha in the pathogenesis of non-alcoholic
steatohepatitis. Gut 2001;48:206-11.
89. Ueki K, Kondo T, Tseng YH, Kahn CR. Central role of suppressors of cytokine
signaling proteins in hepatic steatosis, insulin resistance, and the metabolic syndrome in
the mouse. Proc Natl Acad Sci U S A 2004;101:10422-7.
90. Hui JM, Hodge A, Farrell GC, Kench JG, Kriketos A, George J. Beyond insulin
resistance in NASH: TNF-alpha or adiponectin? Hepatology 2004;40:46-54.
91. Leclercq IA, Farrell GC, Schriemer R, Robertson GR. Leptin is essential for the
hepatic fibrogenic response to chronic liver injury. J Hepatol 2002;37:206-13.
92. Anania FA. Leptin, liver, and obese mice-fibrosis in the fat lane. Hepatology
2002;36:246-8.
93. Bray GA, York DA. The MONA LISA hypothesis in the time of leptin. Recent
Prog Horm Res 1998;53:95-117; discussion -8.
94. Targher G, Bertolini L, Zoppini G, Zenari L, Falezza G. Relationship of non
alcoholic hepatic steatosis to Cortisol secretion in diet controled Type 2 diabetic
patients. Diabetic Med 2005.
95. Bujalska IJ, Kumar S, Stewart PM. Does central obesity reflect "Cushing's disease
ofthe omentum"? Lancet 1997;349:1210-3.
96. McCuskey RS, Ito Y, Robertson GR, McCuskey MK, Perry M, Farrell GC. Hepatic
microvascular dysfunction during evolution of dietary steatohepatitis in mice.
Hepatology 2004;40:386-93.
97. Wanless IR, Shiota K. The pathogenesis of nonalcoholic steatohepatitis and other
fatty liver diseases: a four-step model including the role of lipid release and hepatic
venular obstruction in the progression to cirrhosis. Semin Liver Dis 2004;24:99-106.
98. Diehl AM, Goodman Z, Ishak KG. Alcohollike liver disease in nonalcoholics. A
clinical and histologic comparison with alcohol-induced liver injury. Gastroenterology
1988;95:1056-62.
99. Schwimmer JB, Deutsch R, Rauch JB, Behling C, Newbury R, Lavine JE. Obesity,
insulin resistance, and other clinicopathological correlates of pediatric nonalcoholic
fatty liver disease. J Pediatr 2003;143:500-5.
100. Schwimmer JB, Behling C, Newbury R, Deutsch R, Nievergelt C, Schork NJ,
Lavine JE. Histopathology of pediatric nonalcoholic fatty liver disease. Hepatology
2005;42:641-9.
101. 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;94:2467-74.
102. Kleiner D E , Brunt E M , Van Natta M , Behling C, Contos MJ, Cummings O W ,
Ferrell LD, Liu YC, Torbenson MS, Unalp-Arida A, Yeh M, McCullough AJ, Sanyal
AJ. Design and validation of a histological scoring system for nonalcoholic fatty liver
disease. Hepatology 2005;41:1313-21.
103. Ratziu V, Charlotte F, Heurtier A, Gombert S, Giral P, Bruckert E, Grimaldi A,
Capron F, Poynard T. Sampling variability of liver biopsy in nonalcoholic Fatty liver
disease. Gastroenterology 2005;128:1898-906.
104. Janiec DJ, Jacobson ER, Freeth A, Spaulding L, Blaszyk H. Histologic variation
of grade and stage of non-alcoholic fatty liver disease in liver biopsies. Obes Surg
2005;15:497-501.
105. Joseph AE, Saverymuttu SH, al-Sam S, Cook MG, Maxwell JD. Comparison of
liver histology with ultrasonography in assessing diffuse parenchymal liver disease.
Clin Radiol 1991;43:26-31.
106. Mathiesen UL, Franzen LE, Aselius H, Resjo M, Jacobsson L, Foberg U, Fryden
A, Bodemar G. Increased liver echogenicity at ultrasound examination reflects degree
of steatosis but not of fibrosis in asymptomatic patients with mild/moderate
abnormalities of liver transaminases. Dig Liver Dis 2002;34:516-22.
107. Limanond P, Raman SS, Lassman C, Sayre J, Ghobrial RM, Busuttil RW, Saab S,
Lu DS. Macrovesicular hepatic steatosis in living related liver donors: correlation
between CT and histologic findings. Radiology 2004;230:276-80.
108. Mendler MH, Bouillet P, Le Sidaner A, Lavoine E, Labrousse F, Sautereau D,
Pillegand B. Dual-energy CT in the diagnosis and quantification of fatty liver: limited
clinical value in comparison to ultrasound scan and single-energy CT, with special
reference to iron overload. J Hepatol 1998;28:785-94.
109. Saadeh S, Younossi ZM, Remer EM, Gramlich T, Ong JP, Hurley M, Mullen KD,
Cooper JN, Sheridan MJ. The utility of radiological imaging in nonalcoholic fatty liver
disease. Gastroenterology 2002;123:745-50.
110. Levenson H, Greensite F, Hoefs J, Friloux L, Applegate G, Silva E, Kanel G,
Buxton R. Fatty infiltration ofthe liver: quantification with phase-contrast MR imaging
at 1.5 T vs biopsy. AJR Am J Roentgenol 1991;156:307-12.
111. Ataseven H, Yildirim MH, Yalniz M, Bahcecioglu IH, Celebi S, Ozercan IH. The
value of ultrasonography and computerized tomography in estimating the
histopathological severity of nonalcoholic steatohepatitis. Acta Gastroenterol Belg
2005;68:221-5.
112. Caldwell SH, Hespenheide EE. Subacute liver failure in obese women. Am J
Gastroenterol 2002;97:2058-62.
113. Kuwabara H, Yoshii Y, Mori H, Fujiwara S, Eiraku S, Kojima H, Miyaji K,
Hongo Y, Katsu K. Nonalcoholic steatohepatitis-related cirrhosis with subacute liver
failure: an autopsy case. Dig Dis Sci 2003;48:1668-70.
114. Dam-Larsen S, Franzmann M, Andersen IB, Christoffersen P, Jensen LB,
Sorensen TI, Becker U, Bendtsen F. Long term prognosis of fatty liver: risk of chronic
liver disease and death. Gut 2004;53:750-5.
115. 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;116:1413-9.
116. Powell EE, Cooksley W G , Hanson R, Searle J, Halliday JW, Powell L W . The
natural history of nonalcoholic steatohepatitis: a follow-up study of forty-two patients
for up to 21 years. Hepatology 1990;11:74-80.
117. Teli MR, James OF, Burt AD, Bennett MK, Day CP. The natural history of
nonalcoholic fatty liver: a follow-up study. Hepatology 1995;22:1714-9.
118. Lee RG. Nonalcoholic steatohepatitis: a study of 49 patients. Hum Pathol
1989;20:594-8.
119. Cortez-Pinto H, Baptista A, Camilo ME, De Moura MC. Nonalcoholic
steatohepatitis—a long-term follow-up study: comparison with alcoholic hepatitis in
ambulatory and hospitalized patients. Dig Dis Sci 2003;48:1909-13.
120. Powell EE, Jonsson JR, Clouston AD. Steatosis: Co-factor in other liver diseases.
Hepatology 2005;42:5-13.
121. Jepsen P, Vilstrup H, Mellemkjaer L, Thulstrup AM, Olsen JH, Baron JA,
Sorensen HT. Prognosis of patients with a diagnosis of fatty liver~a registry-based
cohort study. Hepatogastroenterology 2003;50:2101-4.
122. Sorensen T, Mellemkjaer L, Jepsen P, Thulstrup A, Baron J, Olsen J, Vilstrup H.
Risk of cancer in patients hospitalized with fatty liver. J Clin Gastroenterol
2003;36:356-59.
123. Ludwig J, Viggiano TR, McGill DB, Oh BJ. Nonalcoholic steatohepatitis: Mayo
Clinic experiences with a hitherto unnamed disease. Mayo Clin Proc 1980;55:434-8.
124. Petrides AS, Stanley T, Matthews DE, Vogt C, Bush AJ, Lambeth H. Insulin
resistance in cirrhosis: prolonged reduction of hyperinsulinemia normalizes insulin
sensitivity. Hepatology 1998;28:141-9.
125. Fassio E, Alvarez E, Dominguez N, Landeira G, Longo C. Natural history of
nonalcoholic steatohepatitis: a longitudinal study of repeat liver biopsies. Hepatology
2004;40:820-6.
126. Harrison SA, Torgerson S, Hayashi PH. The natural history of nonalcoholic fatty
liver disease: a clinical histopathological study. Am J Gastroenterol 2003;98:2042-7.
127. Evans CD, Oien KA, MacSween RN, Mills PR. Non-alcoholic steatohepatitis: a
common cause of progressive chronic liver injury? J Clin Pathol 2002;55:689-92.
128. Angulo P, Keach JC, Batts KP, Lindor KD. Independent predictors of liver
fibrosis in patients with nonalcoholic steatohepatitis. Hepatology 1999;30:1356-62.
129. Ratziu V, Giral P, Charlotte F, Bruckert E, Thibault V, Theodorou I, Khalil L,
Turpin G, Opolon P, Poynard T. Liver fibrosis in overweight patients. Gastroenterology
2000;118:1117-23. .
130. Bugianesi E, Manzini P, D'Antico S, Vanni E, Longo F, Leone N, Massarenti P,
Piga A, Marchesini G, Rizzetto M. Relative contribution of iron burden, HFE
mutations, and insulin resistance to fibrosis in nonalcoholic fatty liver. Hepatology
2004;39:179-87.
131. Hui AY, Wong VW, Chan HL, Liew CT, Chan JL, Chan FK, Sung JJ.
Histological progression of non-alcoholic fatty liver disease in Chinese patients.
Aliment Pharmacol Ther 2005;21:407-13.
132. Sakugawa H, Nakasone H, Nakayoshi T, Kawakami Y, Yamashiro T, Maeshiro T,
Kobashigawa K, Kinjo F, Saito A. Clinical characteristics of patients with cryptogenic
liver cirrhosis in Okinawa, Japan. Hepatogastroenterology 2003;50:2005-8.
133. Abdelmalek M , Ludwig J, Lindor K D . T w o cases from the spectrum of
nonalcoholic steatohepatitis. J Clin Gastroenterol 1995;20:127-30.
134. Caldwell SH, Oelsner DH, Iezzoni JC, Hespenheide EE, Battle EH, Driscoll CJ.
Cryptogenic cirrhosis: clinical characterization and risk factors for underlying disease.
Hepatology 1999;29:664-9.
135. Poonawala A, Nair SP, Thuluvath PJ. Prevalence of obesity and diabetes in
patients with cryptogenic cirrhosis: a case-control study. Hepatology 2000;32:689-92.
136. Duseja A, Nanda M, Das A, Das R, Bhansali A, Chawla Y. Prevalence of obesity,
diabetes mellitus and hyperlipidaemia in patients with cryptogenic liver cirrhosis. Trop
Gastroenterol 2004;25:15-7.
137. Ong J, Younossi ZM, Reddy V, Price LL, Gramlich T, Mayes J, Boparai N.
Cryptogenic cirrhosis and posttransplantation nonalcoholic fatty liver disease. Liver
Transpl 2001;7:797-801.
138. Sutedja DS, Gow PJ, Hubscher SG, Elias E. Revealing the cause of cryptogenic
cirrhosis by posttransplant liver biopsy. Transplant Proc 2004;36:2334-7.
139. Targher G. Associations between liver histology and early carotid atherosclerosis
in subjects with nonalcoholic fatty liver disease. Hepatology 2005;42:974-5; discussion
5.
140. Brea A, Mosquera D, Martin E, Arizti A, Cordero JL, Ros E. Nonalcoholic fatty
liver disease is associated with carotid atherosclerosis: a case-control study. Arterioscler
Thromb Vase Biol 2005;25:1045-50.
141. Volzke H, Robinson D M , Kleine V, Deutscher R, Hoffmann W , Ludemann J,
Schminke U, Kessler C, John U. Hepatic steatosis is associated with an increased risk
of carotid atherosclerosis. World J Gastroenterol 2005 ;11:1848-53.
142. Charlton M, Kasparova P, Weston S, Lindor K, Maor-Kendler Y, Wiesner RH,
Rosen CB, Batts KP. Frequency of nonalcoholic steatohepatitis as a cause of advanced
liver disease. Liver Transpl 2001;7:608-14.
143. Contos MJ, Cales W, Sterling RK, Luketic VA, Shiffman ML, Mills AS, Fisher
RA, Ham J, Sanyal AJ. Development of nonalcoholic fatty liver disease after orthotopic
liver transplantation for cryptogenic cirrhosis. Liver Transpl 2001;7:363-73.
144. Bullock RE, Zaitoun AM, Aithal GP, Ryder SD, Beckingham IJ, Lobo DN.
Association of non-alcoholic steatohepatitis without significant fibrosis with
hepatocellular carcinoma. J Hepatol 2004;41:685-6.
145. Shimada M, Hashimoto E, Taniai M, Hasegawa K, Okuda H, Hayashi N,
Takasaki K, Ludwig J. Hepatocellular carcinoma in patients with non-alcoholic
steatohepatitis. J Hepatol 2002;37:154-60.
146. Hui JM, Kench JG, Chitturi S, Sud A, Farrell GC, Byth K, Hall P, Khan M,
George J. Long-term outcomes of cirrhosis in nonalcoholic steatohepatitis compared
with hepatitis C. Hepatology 2003;38:420-7.
147. Ratziu V, Bonyhay L, Di Martino V, Charlotte F, Cavallaro L, Sayegh-Tainturier
MH, Giral P, Grimaldi A, Opolon P, Poynard T. Survival, liver failure, and
hepatocellular carcinoma in obesity-related cryptogenic cirrhosis. Hepatology
2002;35:1485-93.
148. Marrero JA, Fontana RJ, Su GL, Conjeevaram HS, Emick D M , Lok AS. N A F L D
may be a common underlying liver disease in patients with hepatocellular carcinoma in
the United States. Hepatology 2002;36:1349-54.
149. Bugianesi E, Leone N, Vanni E, Marchesini G, Brunello F, Carucci P, Musso A,
De Paolis P, Capussotti L, Salizzoni M, Rizzetto M. Expanding the natural history of
nonalcoholic steatohepatitis: from cryptogenic cirrhosis to hepatocellular carcinoma.
Gastroenterology 2002; 123:134-40.
150. Younossi ZM, Gramlich T, Matteoni CA, Boparai N, McCullough AJ.
Nonalcoholic fatty liver disease in patients with type 2 diabetes. Clin Gastroenterol
Hepatol 2004;2:262-5'.
151. Executive Summary of The Third Report of The National Cholesterol Education
Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood
Cholesterol In Adults (Adult Treatment Panel III). JAMA 2001;285:2486-97.
152. Willner IR, Waters B, Patil SR, Reuben A, Morelli J, Riely CA. Ninety patients
with nonalcoholic steatohepatitis: insulin resistance, familial tendency, and severity of
disease. Am J Gastroenterol 2001;96:2957-61.
153. Marchesini G, Brizi M, Morselli-Labate AM, Bianchi G, Bugianesi E,
McCullough AJ, Forlani G, MelchiondaN. Association of nonalcoholic fatty liver
disease with insulin resistance. Am J Med 1999;107:450-5.
154. Marchesini G, Pagotto U, Bugianesi E, De Iasio R, Manini R, Vanni E, Pasquali
R Melchionda N, Rizzetto M. Low ghrelin concentrations in nonalcoholic fatty liver
disease are related to insulin resistance. J Clin Endocrinol Metab 2003;88:5674-9.
155. Alberti K G , Zimmet PZ. Definition, diagnosis and classification of diabetes
mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus
provisional report of a WHO consultation. Diabet Med 1998;15:539-53.
156. Ford ES. Prevalence ofthe metabolic syndrome defined by the international
diabetes federation among adults in the US. Diabetes Care 2005;28:2745-8.
157. Bedogni G, Miglioli L, Masutti F, Tiribelli C, Marchesini G, Bellentani S.
Prevalence of and risk factors for nonalcoholic fatty liver disease: the Dionysos
nutrition and liver study. Hepatology 2005;42:44-52.
158. Kim HJ, Lee KE, Kim DJ, Kim SK, Ahn CW, Lim SK, Kim KR, Lee HC, Huh
KB, Cha BS. Metabolic significance of nonalcoholic fatty liver disease in nonobese,
nondiabetic adults. Arch Intern Med 2004;164:2169-75.
159. Nielsen S, Guo Z, Johnson CM, Hensrud DD, Jensen MD. Splanchnic lipolysis in
human obesity. J Clin Invest 2004; 113:1582-8.
160. Kelley DE, McKolanis TM, Hegazi RA, Kuller LH, Kalhan SC. Fatty liver in type
2 diabetes mellitus: relation to regional adiposity, fatty acids, and insulin resistance. Am
J Physiol Endocrinol Metab 2003;285:E906-16.
161. Stranges S, Dorn D, Freudenheim J, Farinaro E, Russell M, Nochajski T, Trevisan
M. Body fat distribution, relative weight, and liver enzyme levels: a population based
study. Hepatology 2004;39:754-63.
162. Bellentani S, Saccoccio G, Masutti F, Croce LS, Brandi G, Sasso F, Cristanini G,
Tiribelli C. Prevalence of and risk factors for hepatic steatosis in Northern Italy. Ann
Intern Med 2000;132:112-7.
163. Silverman E M , Sapala JA, Appelman HD. Regression of hepatic steatosis in
morbidly obese persons after gastric bypass. Am J Clin Pathol 1995;104:23-31.
164. Marceau P, Biron S, Hould FS, Marceau S, Simard S, Thung SN, Krai JG. Liver
pathology and the metabolic syndrome X in severe obesity. J Clin Endocrinol Metab
1999;84:1513-7.
165. Lewis GF, Carpentier A, Adeli K, Giacca A. Disordered fat storage and
mobilization in the pathogenesis of insulin resistance and type 2 diabetes. Endocr Rev
2002;23:201-29.
166. Assy N, Kiata K, Mymin D, Ley C, Rosser B, Minuk G. Fatty infiltration of liver
in hyperlipidemic patients. Dig Dis Sci 2000;45:1929-34.
167. Ehrmann DA. Polycystic ovary syndrome. N Engl J Med 2005;352:1223-36.
168. Murray RD, Shalet SM. Insulin sensitivity is impaired in adults with varying
degrees of GH deficiency. Clin Endocrinol (Oxf) 2005;62:182-8.
169. Pinkney J, Wilding J, Williams G, MacFarlane I. Hypothalamic Obesity in
Humans: what do we know and what can be done ? Obesity Reviews 2002;3:27-34.
170. Mersebach H, Svendsen OL, Hoist JJ, Astrup A, Feldt-Rasmussen U.
Comparisons of leptin, incretins and body composition in obese and lean patients with
hypopituitarism and healthy individuals. Clin Endocrinol (Oxf) 2003;58:65-71.
171. Apridonidze T, Essah PA, Iuorno MJ, Nestler JE. Prevalence and characteristics
ofthe metabolic syndrome in women with polycystic ovary syndrome. J Clin
Endocrinol Metab 2005;90:1929-35.
172 Schwimmer JB, Khorram C, Chiu V, Schwimmer W. Abnormal aminotransferase
activity in women with polycystic ovary syndrome. Fertil Steril 2005;83:494-7.
173. Javor ED, Ghany M G , Cochran EK, Oral EA, DePaoli A M , Premkumar A,
Kleiner DE, Gorden P. Leptin reverses nonalcoholic steatohepatitis in patients with
severe lipodystrophy. Hepatology 2005;41:753-60.
174. Caples S, Garni A, Somers V. Obstructive sleep apnoea. Ann Intern Med
2005;142:187-97.
175. Tanne F, Gagnadoux F, Chazouilleres O, Fleury B, D. W, Lasnier E, B. L,
Poupon R, Serfaty L. Chronic liver injury during obstructive sleep apnoea. Hepatology
2005;41:1290-6.
176. Chin K, Nakamura T, Takahashi K, Sumi K, Ogawa Y, Masuzaki H, Muro S,
Hattori N, Matsumoto H, Niimi A, Chiba T, Nakao K, Nishima M, Ohi M, Nakamura
T. Effects of obstructive sleep apnoea syndrome on serum aminotransferase levels in
obese patients. Am J Med 2003;114:370-6.
177. Johansson JO, Fowelin J, Landin K, Lager I, Bengtsson BA. Growth hormone-
deficient adults are insulin-resistant. Metabolism 1995;44:1126-9.
178. Srinivasan S, Ogle GD, Garnett SP, Briody JN, Lee JW, Cowell CT. Features of
the metabolic syndrome after childhood craniopharyngioma. J Clin Endocrinol Metab
2004;89:81-6.
179. Ozbey N, Algun E, Turgut AS, Orhan Y, Sencer E, Molvalilar S. Serum lipid and
leptin concentrations in hypopituitary patients with growth hormone deficiency. Int J
Obes Relat Metab Disord 2000;24:619-26.
180. Roth C, Wilken B, Hanefeld F, Schroter W, Leonhardt U. Hyperphagia in children
with craniopharyngioma is associated with hyperleptinaemia and a failure in the
downregulation of appetite. Eur J Endocrinol 1998;138:89-91.
181. Mokdad A H , Ford ES, Bowman BA, Dietz W H , Vinicor F, Bales VS, Marks JS.
Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA
2003;289:76-9.
182. Narayan KM, Boyle JP, Thompson TJ, Sorensen SW, Williamson DF. Lifetime
risk for diabetes mellitus in the United States. Jama 2003;290:1884-90.
183. Gupte P, Amarapurkar D, Agal S, Baijal R, Kulshrestha P, Pramanik S, Patel N,
Madan A, Amarapurkar A, Hafeezunnisa. Non-alcoholic steatohepatitis in type 2
diabetes mellitus. J Gastroenterol Hepatol 2004;19:854-8.
184. Akbar DH, Kawther AH. Nonalcoholic fatty liver disease in Saudi type 2 diabetic
subjects attending a medical outpatient clinic: prevalence and general characteristics.
Diabetes Care 2003;26:3351-2.
185. Paradis V, Perlemuter G, Bonvoust F, Dargere D, Parfait B, Vidaud M, Conti M,
Huet S, Ba N, Buffet C, Bedossa P. High glucose and hyperinsulinemia stimulate
connective tissue growth factor expression: a potential mechanism involved in
progression to fibrosis in nonalcoholic steatohepatitis. Hepatology 2001;34:738-44.
186. Ford ES, Giles WH, Dietz WH. Prevalence ofthe metabolic syndrome among US
adults: findings from the third National Health and Nutrition Examination Survey.
JAMA 2002;287:356-9.
187. Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity
among US adults, 1999-2000. JAMA 2002;288:1723-7.
188. Bureau UC. Table DP-1, Profile of General Demographics. 2000:1117.
189. Melton LJ, 3rd. History ofthe Rochester Epidemiology Project. Mayo Clin Proc
1996;71:266-74.
190. Organisation W H . Report of a W H O consultation: definition of metabolic
syndrome in definition, diagnosis and classification of diabetes mellitus. World Health
Organisation, Department of Noncommunicable Disease Surveillance, Geneva. 1999.
191. Ederer F. The Relative Survival Rate: a statistical methodology. National Cancer
Institute Monograph 1961;6:101-21.
192. Health MDo. Mortality Table 4.2, MN Fifteen Leading Causes of Death.
Minnesta Health Statistics Annual Summary 2000:23.
193. Haffner SM, Lehto S, Ronnemaa T, Pyorala K, Laakso M. Mortality from
coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with
and without prior myocardial infarction. N Engl J Med 1998;339:229-34.
194. Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ. Overweight, obesity, and
mortality from cancer-in a prospectively studied cohort of U.S. adults. N Engl J Med
2003;348:1625-38.
195. Saydah SH, Loria CM, Eberhardt MS, Brancati FL. Abnormal glucose tolerance
and the risk of cancer death in the United States. Am J Epidemiol 2003;157:1092-100.
196. Ioannou GN, Weiss NS, Kowdley KV, Dominitz JA. Is obesity a risk factor for
cirrhosis-related death or hospitalization? A population-based cohort study.
Gastroenterology 2003; 125:1053-9.
197. El-Serag HB, Tran T, Everhart JE. Diabetes increases the risk of chronic liver
disease and hepatocellular carcinoma. Gastroenterology 2004;126:460-8.
198. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker
EA, Nathan DM. Reduction in the incidence of type 2 diabetes with lifestyle
intervention or metformin. N Engl J Med 2002;346:393-403.
199. Eriksson KF, Lindgarde F. N o excess 12-year mortality in men with impaired
glucose tolerance who participated in the Malmo Preventive Trial with diet and
exercise. Diabetologia 1998;41:1010-6.
200. Effect of intensive blood-glucose control with metformin on complications in
overweight patients with type 2 diabetes (UKPDS 34). UK Prospective Diabetes Study
(UKPDS) Group. Lancet 1998;352:854-65.
201. Malik S, Wong ND, Franklin SS, Kamath TV, L'ltalien GJ, Pio JR, Williams GR.
Impact ofthe metabolic syndrome on mortality from coronary heart disease,
cardiovascular disease, and all causes in United States adults. Circulation
2004;110:1245-50.
202. Ford ES. The metabolic syndrome and mortality from cardiovascular disease and
all-causes: findings from the National Health and Nutrition Examination Survey II
Mortality Study. Atherosclerosis 2004;173:309-14.
203. Hunt KJ, Resendez RG, Williams K, Haffner SM, Stern MP. National Cholesterol
Education Program versus World Health Organization metabolic syndrome in relation
to all-cause and cardiovascular mortality in the San Antonio Heart Study. Circulation
2004;110:1251-7.
204. Kahn R, Buse J, Ferrannini E, Stern M. The Metabolic Syndrome: Time for a
critical appraisal. Diabetes Care 2005;28:2289-305.
205. Xu A, Wang Y, Keshaw H, Xu LY, Lam KS, Cooper GJ. The fat-derived
hormone adiponectin alleviates alcoholic and nonalcoholic fatty liver diseases in mice. J
Clin Invest 2003;112:91-100.
206. Sorrentino P, Tarantino G, Conca P, Perrella A, Perrella O. Clinical presentation
and prevalence of spontaneous bacterial peritonitis in patients with cryptogenic
cirrhosis and features of metabolic syndrome. Can J Gastroenterol 2004;18:381-6.
207. Bonora E, Targher G, Alberiche M, Bonadonna RC, Saggiani F, Zenere MB,
Monauni T, Muggeo M. Homeostasis model assessment closely mirrors the glucose
clamp technique in the assessment of insulin sensitivity: studies in subjects with various
degrees of glucose tolerance and insulin sensitivity. Diabetes Care 2000;23:57-63.
208. Katz A, Nambi SS, Mather K, Baron AD, Follmann DA, Sullivan G, Quon MJ.
Quantitative insulin sensitivity check index: a simple, accurate method for assessing
insulin sensitivity in humans. J Clin Endocrinol Metab 2000;85:2402-10.
209. Lindor KD, Kowdley KV, Heathcote EJ, Harrison ME, Jorgensen R, Angulo P,
Lymp JF, Burgart L, Colin P. Ursodeoxycholic acid for treatment of nonalcoholic
steatohepatitis: results of a randomized trial. Hepatology 2004;39:770-8.
210. Laurin J, Lindor KD, Crippin JS, Gossard A, Gores GJ, Ludwig J, Rakela J,
McGill DB. Ursodeoxycholic acid or clofibrate in the treatment of non-alcohol-induced
steatohepatitis: a pilot study. Hepatology 1996;23:1464-7.
211. Bravo AA, Sheth SG, Chopra S. Liver biopsy. N Engl J Med 2001;344:495-500.
212. Searle J, Kerr J, Halliday JW, Powell LW. Pathology ofthe Liver. 2nd ed.
Edinburgh, Scotland: Churchill Livingston; 1987.
213. Ortiz V, Berenguer M, Rayon JM, Carrasco D, Berenguer J. Contribution of
obesity to hepatitis C-related fibrosis progression. Am J Gastroenterol 2002;97:2408-
14.
214. Hui JM, Sud A, Farrell G C , Bandara P, Byth K, Kench JG, McCaughan G W ,
George J. Insulin resistance is associated with chronic hepatitis C virus infection and
fibrosis progression [corrected]. Gastroenterology 2003;125:1695-704.
215. Trombetta M, Spiazzi G, Zoppini G, Muggeo M. Review article: type 2 diabetes
and chronic liver disease in the Verona diabetes study. Aliment Pharmacol Ther
2005;22 Suppl 2:24-7.
216. Sugimoto R, Enjoji M, Kohjima M, Tsuruta S, Fukushima M, Iwao M, Sonta T,
Kotoh K, Inoguchi T, Nakamuta M. High glucose stimulates hepatic stellate cells to
proliferate and to produce collagen through free radical production and activation of
mitogen-activated protein kinase. Liver Int 2005;25:1018-26.
217. Marubbio AT, Jr., Buchwald H, Schwartz MZ, Varco R. Hepatic lesions of central
pericellular fibrosis in morbid obesity, and after jejunoileal bypass. Am J Clin Pathol
1976;66:684-91.
218. Bernuau D, Guillot R, Durand AM, Raoux N, Gabreau T, Passa P, Feldmann G.
Ultrastructural aspects ofthe liver perisinusoidal space in diabetic patients with and
without microangiopathy. Diabetes 1982;31:1061-7.
219. Huet PM, Villeneuve JP. Microcirculation ofthe aging liver: is getting old like
having cirrhosis? Hepatology 2005;42:1248-51.
220. Kadowaki T, Yamauchi T. Adiponectin and adiponectin receptors. Endocr Rev
2005;26:439-51.
221. Masaki T, Chiba S, Tatsukawa H, Yasuda T, Noguchi H, Seike M, Yoshimatsu H.
Adiponectin protects LPS-induced liver injury through modulation of TNF-alpha in
KK-Ay obese mice. Hepatology 2004;40:177-84.
222. Kamada Y, Tamura S, Kiso S, Matsumoto H, Saji Y, Yoshida Y, Fukui K, Maeda
N, Nishizawa H, Nagaretani H, Okamoto Y, Kihara S, Miyagawa J, Shinomura Y,
Funahashi T, Matsuzawa Y. Enhanced carbon tetrachloride-induced liver fibrosis in
mice lacking adiponectin. Gastroenterology 2003;125:1796-807.
223. Bugianesi E, Pagotto U, Manini R, Vanni E, Gastaldelli A, de Iasio R, Gentilcore
E, Natale S, Cassader M, Rizzetto M, Pasquali R, Marchesini G. Plasma adiponectin in
nonalcoholic fatty liver is related to hepatic insulin resistance and hepatic fat content,
not to liver disease severity. J Clin Endocrinol Metab 2005;90:3498-504.
224. Musso G, Gambino'R, Biroli G, Carello M, Faga E, Pacini G, De Michieli F,
Cassader M, Durazzo M, Rizzetto M, Pagano G. Hypoadiponectinemia predicts the
severity of hepatic fibrosis and pancreatic Beta-cell dysfunction in nondiabetic
nonobese patients with nonalcoholic steatohepatitis. Am J Gastroenterol
2005;100:2438-46.
225. Adams PC, Arthur MJ, Boyer TD, DeLeve LD, Di Bisceglie AM, Hall M, Levin
TR, Provenzale D, Seeff L. Screening in liver disease: report of an AASLD clinical
workshop. Hepatology 2004;39:1204-12.
226. Dixon JB, Bhathal PS, Hughes NR, O'Brien PE. Nonalcoholic fatty liver disease:
Improvement in liver histological analysis with weight loss. Hepatology 2004;39:1647-
54.
227. Scheuer PJ. Liver biopsy size matters in chronic hepatitis: bigger is better.
Hepatology 2003;38:1356-8.
228. Friedman SL. Liver fibrosis - from bench to bedside. J Hepatol 2003 ;38 Suppl
LS38-53.
229. Sanyal AJ, Campbell-Sargent C, Mirshahi F, Rizzo W B , Contos MJ, Sterling RK,
Luketic VA, Shiftman ML, Clore JN. Nonalcoholic steatohepatitis: association of
insulin resistance and mitochondrial abnormalities. Gastroenterology 2001; 120:1183-
92.
230. Lustig R. Hypothalamic Obesity: the sixth cranial endocrinopathy. The
Endocrinologist 2002;12:210-17.
231. Arafah BM. Medical management of hypopituitarism in patients with pituitary
adenomas. Pituitary 2002;5:109-17.
232. Shabsigh R, Katz M, Yan G, Makhsida N. Cardiovascular issues in hypogonadism
and testosterone therapy. Am J Cardiol 2005;96:67-72.
233. Dhatariya K, Bigelow ML, Nair KS. Effect of dehydroepiandrosterone
replacement on insulin sensitivity and lipids in hypoadrenal women. Diabetes
2005;54:765-9.
234. Medina MC, Souza LC, Caperuto LC, Anhe GF, Amanso AM, Teixeira VP,
Bordin S, Carpinelli AR, Britto LR, Barbieri RL, Borella MI, Carvalho CR.
Dehydroepiandrosterone increases beta-cell mass and improves the glucose-induced
insulin secretion by pancreatic islets from aged rats. FEBS Lett 2006;580:285-90.
235. Holmang A, Bjorntorp P. The effects of testosterone on insulin sensitivity in male
rats. Acta Physiol Scand 1992;146:505-10.
236. Brobeck J. Mechanism ofthe development of obesity in animals with
hypothalamic lesions.'Physiol Rev 1946;26:541-59.
237. Schwartz MW, Porte D, Jr. Diabetes, obesity, and the brain. Science
2005;307:375-9.
238. Koteish A, Diehl A M . Animal models of steatosis. Semin Liver Dis 2001 ;21:89-
104.
239. Zelman S. The liver in obesity. Arch Intern Med 1952;90:141-56.
240. Chitturi S, Abeygunasekera S, Farrell GC, Holmes-Walker J, Hui JM, Fung C,
Karim R, Lin R, Samarasinghe D, Liddle C, Weltman M, George J. NASH and insulin
resistance: Insulin hypersecretion and specific association with the insulin resistance
syndrome. Hepatology 2002;35:373-9.
241. Adams LA, Feldstein A, Lindor KD, Angulo P. Nonalcoholic fatty liver disease
among patients with hypothalamic and pituitary dysfunction. Hepatology 2004;39:909-
14.
242. Altuntas B, Ozcakar B, Bideci A, Cinaz P. Cirrhotic outcome in patients with
craniopharyngioma. J Pediatr Endocrinol Metab 2002;15:1057-8.
243. Basenau D, Stehphani U, Fischer G. [Development of complete liver cirrhosis in
hyperphagia-induced fatty liver]. Klin Padiatr 1994;206:62-4.
244. Takano S, Kanzaki S, Sato M, Kubo T, Seino Y. Effect of growth hormone on
fatty liver in panhypopituitarism. Arch Dis Child 1997;76:537-8.
245. Nakajima K, Hashimoto E, Kaneda H, Tokushige K, Shiratori K, Hizuka N,
Takano K. Pediatric nonalcoholic steatohepatitis associated with hypopituitarism. J
Gastroenterol 2005;40:312-5.
246. Jonas MM, Krawczuk LE, Kim HB, Lillehei C, Perez-Atayde A. Rapid recurrence
of nonalcoholic fatty liver disease after transplantation in a child with hypopituitarism
and hepatopulmonary syndrome. Liver Transpl 2005;11:108-10.
247. Evans H M , Shaikh M G , McKiernan PJ, Hockley A D , Crowne EC, Kirk JM, Kelly
DA. Acute fatty liver disease after suprasellar tumor resection. J Pediatr Gastroenterol
Nutr 2004;39:288-91.
248. Jeffcoate W. Growth hormone therapy and its relationship to insulin resistance,
glucose intolerance and diabetes mellitus: a review of recent evidence. Drug Saf
2002;25:199-212.
249. Holt RI, Simpson HL, Sonksen PH. The role ofthe growth hormone-insulin-like
growth factor axis in glucose homeostasis. Diabet Med 2003;20:3-15.
250. Ichikawa T, Hamasaki K, Ishikawa H, Ejima E, Eguchi K, Nakao K. Non
alcoholic steatohepatitis and hepatic steatosis in patients with adult onset growth
hormone deficiency. Gut 2003;52:914.
251. Lonardo A, Loria P, Leonardi F, Ganazzi D, Carulli N. Growth hormone plasma
levels in nonalcoholic fatty liver disease. Am J Gastroenterol 2002;97:1071-2.
252. Langendonk JG, Meinders AE, Burggraaf J, Frolich M, Roelen CA, Schoemaker
RC, Cohen AF, Pijl H. Influence of obesity and body fat distribution on growth
hormone kinetics in humans. Am J Physiol 1999;277:E824-9.
253. Beshyah SA, Johnston DG. Cardiovascular disease and risk factors in adults with
hypopituitarism. Clin Endocrinol (Oxf) 1999;50:1-15.
254. Page RC, Levy J, Turner RC. Obesity associated insulin resistance occurs in
hypopituitary subjects. Diabet Med 1994;11:862-5.
255. Hennessey JV, Jackson IM. Clinical features and differential diagnosis of pituitary
tumours with emphasis on acromegaly. Baillieres Clin Endocrinol Metab 1995;9:271-
314.
256. Svensson J, Bengtsson BA. Growth hormone replacement therapy and insulin
sensitivity. J Clin Endocrinol Metab 2003;88:1453-4.
257. al-Shoumer KA, Anyaoku V, Richmond W, Johnston DG. Elevated leptin
concentrations in growth hormone-deficient hypopituitary adults. Clin Endocrinol (Oxf)
1997;47:153-9.
258. Sainsbury A, Cooney GJ, Herzog H. Hypothalamic regulation of energy
homeostasis. Best Pract Res Clin Endocrinol Metab 2002;16:623-37.
259. Loffreda S, Yang SQ, Lin HZ, Karp CL, Brengman ML, Wang DJ, Klein AS,
Bulkley GB, Bao C, Noble PW, Lane MD, Diehl AM. Leptin regulates
proinflammatory immune responses. Faseb J 1998;12:57-65.
260. Bullo M, Garcia-Lorda P, Megias I, Salas-Salvado J. Systemic inflammation,
adipose tissue tumor necrosis factor, and leptin expression. Obes Res 2003 ;11:525-31.
261. Diehl AM, Li ZP, Lin HZ, Yang SQ. Cytokines and the pathogenesis of non
alcoholic steatohepatitis. Gut 2005;54:303-6.
262. Ikejima K, Takei Y, Honda H, Hirose M, Yoshikawa M, Zhang Y, Lange T,
Fukada T, Takei Y, Sato N. Leptin receptor mediated signaling regulates hepatic
fibrogenesis and remodelling of extracellular matric in the rat. Gastroenterology
2002;122:1399-410.
263. Saxena NK, Saliba G, Floyd JJ, Anania FA. Leptin induces increased alpha2(I)
collagen gene expression in cultured rat hepatic stellate cells. J Cell Biochem
2003;89:311-20.
264. Chitturi S, Farrell G, Frost L, Kriketos A, Lin R, Fung C, Liddle C, Samarasinghe
D, George J. Serum leptin in NASH correlates with hepatic steatosis but not fibrosis: a
manifestation of lipotoxicity? Hepatology 2002;36:403-9.
265. Angulo P, Alba LM, Petrovic LM, Adams LA, Lindor KD, Jensen MD. Leptin,
insulin resistance, and liver fibrosis in human nonalcoholic fatty liver disease. J Hepatol
2004;41:943-9.
266. Perseghin G, Lattuada G, De Cobelli F, Esposito A, Costantino F, Canu T, Scifo
P, De Taddeo F, Maffi P, Secchi A, Del Maschio A, Luzi L. Reduced intrahepatic fat
content is associated with increased whole-body lipid oxidation in patients with type 1
diabetes. Diabetologia 2005;48:2615-21.
267. Adams LA, Lymp JF, St Sauver J, Sanderson SO, Lindor KD, Feldstein A,
Angulo P. The natural history of nonalcoholic fatty liver disease: a population-based
cohort study. Gastroenterology 2005; 129:113-21.
268. Adams LA, Sanderson S, Lindor KD, Angulo P. The histological course of
nonalcoholic fatty liver disease: a longitudinal study of 103 patients with sequential
liver biopsies. J Hepatol 2005;42:132-8.
269. Zimmet P. Diabesity in Australia: An affair ofthe heart. Heart Lung Circ 2003; 12
Suppl 2:S95-8.
270. American Diabetes Association: clinical practice recommendations 1997.
Diabetes Care 1997;20 Suppl LS1-70.
271. Diagnosis and classification of diabetes mellitus. Diabetes Care 2005 ;28 Suppl
LS37-42.
272. de Marco R, Locatelli F, Zoppini G, Bonora E, Muggeo M . Cause-specific
mortality in type 2 diabetes. The Verona Diabetes Study. Diabetes Care 1999;22:756-
61.
273. Tran TT, Medline A, Bruce WR. Insulin promotion of colon tumors in rats.
Cancer Epidemiol Biomarkers Prev 1996;5:1013-5.
274. Schoen RE, Tangen CM, Kuller LH, Burke GL, Cushman M, Tracy RP, Dobs A,
Savage PJ. Increased blood glucose and insulin, body size, and incident colorectal
cancer. J Natl Cancer Inst 1999;91:1147-54.
275. Sorrentino P, Tarantino G, Conca P, Perrella A, Terracciano M, Vecchione R,
Gargiulo G, Gennarelli N, Lobello R. Silent nonalcoholic fatty liver disease-a clinical-
histological study. J Hepatol 2004;41:751-57.
276. Ipekci S, Basaranoglu M, Sonsuz A. The fluctuation of serum levels of
aminotranferase in patients with nonalcoholic steatohepatitis. J Clin Gastroenterol
2003;36:371.
277. VillanovaN, Moscatiello S, Ramilli S, Bugianesi E, Magalotti D, Vanni E, Zoli
M, Marchesini G. Endothelial dysfunction and cardiovascular risk profile in
nonalcoholic fatty liver disease. Hepatology 2005;42:473-80.
278. Marchesini G, Ronchi M, Forlani G, Bugianesi E, Bianchi G, Fabbri A, Zoli M,
Melchionda N. Cardiovascular disease in cirrhosis~a point-prevalence study in relation
to glucose tolerance. Am J Gastroenterol 1999;94:655-62.
279. Bugianesi E, Gentilcore E, Manini R, Natale S, Vanni E, Villanova N, David E,
Rizzetto M, Marchesini G. A randomized controlled trial of metformin versus vitamin E
or prescriptive diet in nonalcoholic fatty liver disease. A m J Gastroenterol
2005;100:1082-90.
280. Neuschwander-Tetri BA, Brunt EM, Wehmeier KR, Oliver D, Bacon BR.
Improved nonalcoholic steatohepatitis after 48 weeks of treatment with the PPAR-
gamma ligand rosiglitazone. Hepatology 2003;38:1008-17.
281. Promrat K, Lutchman G, Uwaifo GI, Freedman RJ, Soza A, Heller T, Doo E,
Ghany M, Premkumar A, Park Y, Liang TJ, Yanovski J A, Kleiner DE, Hoofnagle JH.
A pilot study of pioglitazone treatment for nonalcoholic steatohepatitis. Hepatology
2004;39:188-96.
282. Massard J, Ratziu V, Thabut D, Moussalli J, Lebray P, Benhamou Y, Poynard T.
Natural history and predictors of disease severity in chronic hepatitis C. J Hepatol
2006;44 Suppl LSI9-24.
283. Speiser PW, Rudolf MC, Anhalt H, Camacho-Hubner C, Chiarelli F, Eliakim A,
Freemark M, Gruters A, Hershkovitz E, Iughetti L, Krude H, Latzer Y, Lustig RH,
Pescovitz OH, Pinhas-Hamiel O, Rogol AD, Shalitin S, Sultan C, Stein D, Vardi P,
Werther GA, Zadik Z; Zuckerman-Levin N, Hochberg Z. Childhood obesity. J Clin
Endocrinol Metab 2005;90:1871-87.
284. Guzzaloni G, Grugni G, Minocci A, Moro D, Morabito F. Liver steatosis in
juvenile obesity: correlations with lipid profile, hepatic biochemical parameters and
glycemic and insulinemic responses to an oral glucose tolerance test. Int J Obes Relat
Metab Disord 2000;24:772-6.
285. Gilmore IT, Burroughs A, Murray-Lyon IM, Williams R, Jenkins D, Hopkins A.
Indications, methods, and outcomes of percutaneous liver biopsy in England and Wales:
an audit by the British Society of Gastroenterology and the Royal College of Physicians
of London. Gut 1995;36:437-41.
286. Adams LA, Bulsara M, Rossi E, DeBoer B, Speers D, George J, Kench J, Farrell
G, McCaughan GW, Jeffrey GP. Hepascore: an accurate validated predictor of liver
fibrosis in chronic hepatitis C infection. Clin Chem 2005;51:1867-73.