Abstract
AIM: To study clinical and histopathological features of nonalcoholic fatty liver disease (NAFLD) in patients with and without type 2 diabetes mellitus (T2DM) using updated nonalcoholic steatohepatitis clinical research network (NASH-CRN) grading system.
METHODS: We retrospectively analyzed data of 235 patients with biopsy proven NAFLD with and without T2DM. This database was utilized in the previously published study comparing ethnicity outcomes in NAFLD by the same corresponding author. The pathology database from University of Chicago was utilized for enrolling consecutive patients who met the criteria for NAFLD and their detailed clinical and histopathology findings were obtained for comparison. The relevant clinical profile of patients was collected from the Electronic Medical Records around the time of liver biopsy and the histology was read by a single well-trained histopathologist. The updated criteria for type 2 diabetes have been utilized for analysis. Background data of patients with NASH and NAFLD has been included. The mean differences were compared using χ2 and t-test along with regression analysis to evaluate the predictors of NASH and advanced fibrosis.
RESULTS: Patients with NAFLD and T2DM were significantly older (49.9 vs 43.0, P < 0.01), predominantly female (71.4 vs 56.3, P < 0.02), had higher rate of metabolic syndrome (88.7 vs 36.4, P < 0.01), had significantly higher aspartate transaminase (AST)/alanine transaminase (ALT) ratio (0.94 vs 0.78, P < 0.01) and Fib-4 index (1.65 vs 1.06, P < 0.01) as markers of NASH, showed higher mean NAFLD activity score (3.5 vs 3.0, P = 0.03) and higher mean fibrosis score (1.2 vs 0.52, P < 0.01) compared to patients with NAFLD without T2DM. Furthermore, advanced fibrosis (32.5 vs 12.0, P < 0.01) and ballooning (27.3 vs 13.3, P < 0.01) was significantly higher among patients with NAFLD and T2DM compared to patients with NAFLD without T2DM. On multivariate analysis, T2DM was independently associated with NASH (OR = 3.27, 95%CI: 1.43-7.50, P < 0.01) and advanced fibrosis (OR = 3.45, 95%CI: 1.53-7.77, P < 0.01) in all patients with NAFLD. There was a higher rate of T2DM (38.1 vs 19.4, P < 0.01) and cirrhosis (8.3 vs 0.0, P = 0.01) along with significantly higher mean Bilirubin (0.71 vs 0.56, P = 0.01) and AST (54.2 vs 38.3, P < 0.01) and ALT (78.7 vs 57.0, P = 0.01) level among patients with NASH when compared to patients with steatosis alone. The mean platelet count (247 vs 283, P < 0.01) and high-density lipoprotein cholesterol level (42.7 vs 48.1, P = 0.01) was lower among patients with NASH compared to patients with steatosis.
CONCLUSION: Patients with NAFLD and T2DM tend to have more advanced stages of NAFLD, particularly advanced fibrosis and higher rate of ballooning than patients with NAFLD without T2DM.
Keywords: Non-alcoholic steatohepatitis, Non-alcoholic fatty liver disease, Advanced fibrosis, Non-alcoholic fatty liver disease activity score, Type 2 diabetes, Liver biopsy
Core tip: This retrospective cohort study shows that type 2 diabetes mellitus (T2DM) is a reliable predictor for both nonalcoholic steatohepatitis (NASH) and advanced fibrosis. Patients with nonalcoholic fatty liver disease (NAFLD) and uncontrolled T2DM tend to have advanced fibrosis and higher rate of ballooning histologically. It is important to recognize the differences between composite NAS and individual histological features while interpreting liver biopsies among NAFLD patients. Early diagnosis of NASH and advanced fibrosis in patients with NAFLD has important clinical significance especially to prevent further progression of liver disease to cirrhosis, hepatocellular carcinoma and other related complications. Thus, optimization of risk factors for NAFLD such as metabolic syndrome, uncontrolled T2DM and dyslipidemia is of paramount importance.
INTRODUCTION
Nonalcoholic fatty liver disease (NAFLD) includes a histological spectrum of liver diseases ranging from simple steatosis to nonalcoholic steatohepatitis (NASH) and the latter histological entity can progress to cirrhosis in patients without significant alcohol consumption[1]. NAFLD is considered to be the hepatic manifestation of metabolic syndrome (MS)[2]. Type 2 diabetes mellitus (T2DM) is not only associated with NAFLD, but has also been shown to be an independent risk factor for the development of NASH[2-4]. While screening for diabetes and other risk factors for NAFLD is easily performed, evaluation and management of NAFLD is still challenging. In spite of having a well-defined understanding of the stages of NAFLD progression, the exact pathogenesis is still unclear[5].
Despite the availability of various non-invasive tools, liver biopsy is still regarded as the gold standard for accurate measurement of histopathological features of NAFLD[6,7]. Identifying the unique histopathological features of NAFLD among patients with T2DM is not only important to stage the disease progression but also help us understand the impact of diabetes in progression of NAFLD. At present, there are only few studies describing histopathological differences among NAFLD patients with and without T2DM in a multiethnic United States population cohort[5,8]. Moreover, these studies utilized the Brunt scoring system[8,9] and were limited in sample size[8] or included pooled data from various centers[5].
Therefore, the primary aim of our study was to evaluate and compare the clinical, laboratory and detailed histological findings using the updated NASH clinical research network (CRN) scoring system among biopsy proven NAFLD patients with and without T2DM using a single histopathologist from a major tertiary health care center. We also explored the risk of NASH and advanced fibrosis in both groups in addition to comparing the background data of NAFLD and NASH separately.
MATERIALS AND METHODS
Study design
The pathology database from the University of Chicago Medical Center (UCMC) containing the terms “steatosis”, “steatohepatitis” and/or “fat” from June 1, 1995 to June 30, 2005 was retrospectively analyzed and 683 positive biopsy reports were consecutively identified for further analysis. UCMC’s computerized medical records were then retrospectively reviewed to obtain patient-related demographic, clinical, and laboratory data. Patients either lacking adequate information and/or having presence of other concomitant liver diseases including hepatitis B and C, iron overload, medication-related steatosis, significant alcohol use (current daily alcohol consumption of 40 g/d or more in males and 20 g/d or more in females) and liver transplant were excluded. Data on T2DM and MS were collected based upon American Diabetes Association[10] and National Cholesterol Education Program ATP III criteria[11]. Obesity was defined as body mass index (BMI) ≥ 30 kg/m2 since data on waist circumference measures of central obesity were unavailable[12]. Out of the 683 patients with fatty liver histology, only 238 patients were found to be meeting the criteria for NAFLD and on further review, 3 patients had inadequate data regarding presence or absence of T2DM, thereby leaving a total of 235 patients eligible for further analysis. Moreover, all liver biopsies were scored by a single pathologist using the updated NASH CRN scoring system[6]. The pathologist was blinded to the clinical and laboratory data of the patients. The clinical, laboratory and histological data were all computerized and stored securely. The indication for liver biopsy as elicited in previous study[13] among the 238 patients with NAFLD was abnormal liver function tests followed by abnormal intraoperative appearance of liver and subsequently followed by abnormal imaging and associated abdominal pain respectively. The reasons for ineligibility are also demonstrated clearly in the original study[13]. This study was originally reviewed and approved by the institutional review board of the University of Chicago Medical Center. The informed consent was waived as part of the previous study[13] and there were no patient identifiers in the current study.
Case definitions and liver histology
NAFLD was defined histologically by the presence of minimum 5% of steatosis on liver biopsy. Steatosis was scored as 1, 2, or 3 for 5%-33%, 34%-66%, and > 66% steatosis, respectively. Fibrosis was scored as 0, 1, 2, 3 and 4 for no fibrosis, perisinusoidal or periportal, perisinusoidal and portal/periportal, bridging, and cirrhosis, respectively. Lobular inflammation was scored as 0, 1, 2, or 3 based on presence of no inflammation, < 2 foci per 200 × field, 2-4 foci per 200 × field, and > 4 foci per 200 × field, respectively. Ballooning was scored as 0, 1, and 2 for no balloon cell, few balloon cells, and many cells/prominent ballooning cells, respectively. Lastly, Mallory’s hyaline was scored as 0 for “none to rare” or 1 for “many”. A NAFLD activity score (NAS) of ≥ 5 was considered NASH while fibrosis score of ≥ 2 was considered advanced fibrosis[6].
Statistical analysis
Patients with NAFLD were sub-divided according to presence or absence of T2DM. Background data for NASH and NAFLD was analyzed separately. Results are expressed as mean ± SD for continuous variables and as frequencies for categorical variables. A t-test for unequal variance was performed to compare the means of continuous variables. Categorical variables were compared by χ2 test. Separate logistic regression analyses were performed to study the variables associated with presence of NASH and fibrosis. Variables which were significant on univariate analysis were included in the multivariate analysis and independent variables with P > 0.1 were excluded sequentially from the models. The odds ratios and associated P-values of the remaining variables are reported. Two-sided P values < 0.05 were considered statistically significant. Data analyses were performed using Stata (StataCorp, College Station, TX). The statistical methods of this study were reviewed by Matt Briggs, consultant statistician at New York Methodist Hospital.
RESULTS
Demographic, comorbid and biochemical features
Patients with NAFLD and T2DM, were significantly older (49.9 vs 43.0, P < 0.01), had higher proportion of females (71.4 vs 56.3, P < 0.02), and showed higher fasting glucose (167.0 vs 102.9, P < 0.01), higher HbA1c (8.09 vs 5.83, P < 0.01), higher BMI (41.0 vs 35.9, P < 0.01), higher international normalized ratio (1.02 vs 0.97, P = 0.03), higher rates of hypertension (71.4 vs 37.3, P < 0.01), higher rate of dyslipidemia (83.8 vs 61.2, P < 0.01), increased rate of metabolic syndrome (88.7 vs 36.4, P < 0.01) and positive indirect markers of NASH such as aspartate transaminase (AST)/alanine transaminase (ALT) ratio (0.94 ± 0.4 vs 0.78 ± 0.4, P < 0.01) and Fib-4 index (1.65 vs 1.06, P < 0.01) compared to patients with NAFLD without T2DM. Conversely, patients with NAFLD without T2DM had significantly higher rates of abnormal liver function tests as a leading cause of liver biopsy (74 vs 54.5, P = 0.02), higher ALT (79.2 vs 59.2, P = 0.01) and showed higher serum iron levels (89.3 vs 71.3, P < 0.01) compared to patients with NAFLD and T2DM (Table 1).
Table 1.
Characteristics of patients with nonalcoholic fatty liver disease, divided according to nonalcoholic fatty liver disease without type 2 diabetes mellitus and nonalcoholic fatty liver disease with type 2 diabetes mellitus (mean ± SD)
| Parameter | All patients (n = 235) | NAFLD (without T2DM) (n = 158) | NAFLD (with T2DM) (n = 77) | P value |
| Age (yr) | 45.3 ± 11.9 | 43.0 ± 12.2 | 49.9 ± 9.8 | < 0.01 |
| Gender (male/female) | 91/144 | 69/89 | 22/55 | 0.02 |
| Race (% Caucasian) | 65.1 | 63.9 | 67.5 | 0.9 |
| BMI (kg/m2) | 37.6 ± 11.5 | 35.9 ± 10.6 | 41.0 ± 12.6 | < 0.01 |
| Hypertension, n (%) | 114 (48.1) | 59 (37.3) | 55 (71.4) | < 0.01 |
| Dyslipidemia, n (%) | 139 (68.8) | 82 (61.2) | 57 (83.8) | < 0.01 |
| Metabolic syndrome, n (%) | 115 (53.7) | 52 (36.4) | 63 (88.7) | < 0.01 |
| Abnormal LFT’s as an indication of liver biopsy, n (%) | 159 (67.6) | 117 (74.0) | 42 (54.5) | 0.02 |
| Platelets | 257.5 ± 73.5 | 262.6 ± 67.7 | 247.4 ± 83.3 | 0.15 |
| INR | 0.99 ± 0.16 | 0.97 ± 0.10 | 1.02 ± 0.23 | 0.03 |
| Protein (g/dL) | 7.4 ± 0.72 | 7.4 ± 0.73 | 7.3 ± 0.70 | 0.12 |
| Albumin (g/dL) | 4.3 ± 0.5 | 4.3 ± 0.5 | 4.2 ± 0.46 | 0.06 |
| AST (U/L) | 49.8 ± 34.5 | 50.4 ± 38.4 | 48.3 ± 24.8 | 0.6 |
| ALT (U/L) | 72.6 ± 58.5 | 79.2 ± 65.4 | 59.2 ± 38 | 0.01 |
| Bilirubin (mg/dL) | 0.67 ± 0.4 | 0.68 ± 0.4 | 0.62 ± 0.3 | 0.3 |
| Alkaline Phosphate | 96.7 ± 61.8 | 94.9 ± 68.2 | 100.1 ± 46.9 | 0.5 |
| Total cholesterol (mg/dL) | 200.2 ± 47.3 | 200.9 ± 45.4 | 198.8 ± 50.9 | 0.8 |
| HDL cholesterol (mg/dL) | 44.3 ± 11.8 | 44.2 ± 12.0 | 44.6 ± 11.7 | 0.9 |
| Triglycerides (mg/dL) | 211.4 ± 119.9 | 209.7 ± 127.7 | 214.4 ± 105.9 | 0.8 |
| LDL cholesterol (mg/dL) | 118.9 ± 41.0 | 120.5 ± 41.3 | 116.4 ± 40.8 | 0.6 |
| Fasting glucose (mg/dL) | 125.6 ± 55.5 | 102.9 ± 28.3 | 167.0 ± 67.8 | < 0.01 |
| HbA1c | 6.93 ± 2.09 | 5.83 ± 1.11 | 8.09 ± 2.26 | < 0.01 |
| Iron (μg/dL) | 82.6 ± 38.7 | 89.3 ± 40.8 | 71.3 ± 32.0 | < 0.01 |
| Total iron binding capacity (μg/dL) | 322.4 ± 72.3 | 317.16 ± 71.8 | 333.0 ± 72.9 | 0.2 |
| Ferritin (ng/mL) | 258.2 ± 306.0 | 281 ± 323.1 | 223.0 ± 276.7 | 0.3 |
| AST/ALT | 0.836 ± 0.432 | 0.781 ± 0.434 | 0.947 ± 0.41 | < 0.01 |
| FIB-4 index | 1.26 ± 1.01 | 1.06 ± 0.77 | 1.65 ± 1.28 | < 0.01 |
| APRI | 0.581 ± 0.471 | 0.549 ± 0.458 | 0.643 ± 0.493 | 0.16 |
NAFLD: Nonalcoholic fatty liver disease; T2DM: Type 2 diabetes mellitus; AST: Aspartate transaminase; ALT: Alanine transaminase; LDL: Low-density lipoprotein cholesterol; NAS: NAFLD activity score; HDL: High-density lipoprotein cholesterol; BMI: Body mass index; INR: International normalized ratio; APRI: AST to platelet ratio index; LFT: Liver function test.
Background data evaluating all patients with NAFLD showed higher rate of T2DM (38.1 vs 19.4, P < 0.01) and presence of cirrhosis (8.3 vs 0.0, P = 0.01) among patients with NASH compared to those with steatosis (fatty liver) alone. Moreover, there was significant higher values of AST (54.2 vs 38.3, P < 0.01), ALT (78.7 vs 57.0, P = 0.01) and Bilirubin (0.71 vs 0.56, P = 0.01) and lower values of high density lipoprotein (HDL) (42.7 vs 48.1, P = 0.01) and platelet count (247 vs 283, P < 0.01) among patients with NASH respectively. There was no difference in age, sex, ethnicity, BMI and various other risk factor variables between the groups (Table 2).
Table 2.
Characteristics of patients with nonalcoholic fatty liver disease, divided according to nonalcoholic steatohepatitis and steatosis (mean + SD)
| Parameter | All patients (n = 235) | NASH | Steatosis | P value |
| Age (yr) | 45.3 + 11.9 | 45.9 + 12.4 | 43.8 + 10.4 | 0.2 |
| Gender (male/female) | 91/144 | 67/101 | 24/43 | 0.6 |
| Race (% caucasian) | 65.1 | 64.9 | 65.7 | 0.1 |
| BMI (kg/m2) | 37.7 + 11.5 | 37 + 10.6 | 39.4 + 13.7 | 0.17 |
| Hypertension, n (%) | 114 (48.5) | 82 + 48.8 | 32 + 63.9 | 1 |
| Dyslipidemia, n (%) | 139 (68.8) | 100 (70.9) | 39 (63.9) | 0.3 |
| Metabolic syndrome, n (%) | 115 + 53.7 | 67 (42.7) | 32 (56.1) | 0.08 |
| Platelets | 258 + 73.5 | 247 + 72.1 | 283 + 71.1 | < 0.01 |
| INR | 0.996 + 0.162 | 1 + 0.178 | 0.98 + 0.112 | 0.3 |
| Protein (g/dL) | 7.39 + 0.762 | 7.39 + 0.733 | 7.37 + 0.714 | 0.8 |
| Albumin (g/dL) | 4.27 + 0.536 | 4.27 + 0.555 | 4.26 + 0.487 | 0.8 |
| AST (U/L) | 49.8 + 34.5 | 54.2 + 36.3 | 38.3 + 26.1 | < 0.01 |
| ALT (U/L) | 72.6 + 58.5 | 78.7 + 63.1 | 57 + 40.7 | 0.01 |
| Bilirubin (mg/dL) | 0.668 + 0.416 | 0.71 + 0.456 | 0.567 + 0.275 | 0.01 |
| Alkaline Phosphate | 96.7 + 61.8 | 100 + 70.9 | 88.3 + 28.3 | 0.1 |
| Total cholesterol (mg/dL) | 200 + 47.3 | 202 + 48.6 | 197 + 44.5 | 0.5 |
| HDL cholesterol (mg/dL) | 44.4 + 11.9 | 42.7 + 10.7 | 48.1 + 13.6 | 0.01 |
| Triglycerides (mg/dL) | 211 + 120 | 220 + 125 | 191 + 105 | 0.18 |
| LDL cholesterol (mg/dL) | 1119 + 41 | 121 + 41.6 | 114 + 39.8 | 0.33 |
| Fasting glucose (mg/dL) | 126 + 55.5 | 129 + 55.9 | 117 + 53.8 | 0.14 |
| HbA1c | 6.93 + 2.09 | 7.12 + 2.19 | 6.37 + 1.69 | 0.11 |
| Iron (μg/dL) | 82.6 + 38.7 | 83.8 + 40.5 | 78.6 + 31.9 | 0.52 |
| Total iron binding capacity (μg/dL) | 322 + 72.3 | 319 + 74.6 | 335 + 63.4 | 0.31 |
| Ferritin (ng/mL) | 258 + 306 | 276 + 324 | 201 + 232 | 0.21 |
| Mallory bodies | 50 + 21.3 | 50 + 29.8 | 0 | < 0.01 |
| CAD | 67 + 28.9 | 50 + 29.9 | 17 + 26.2 | 0.63 |
| Gastric bypass | 21 + 8.94 | 14 + 8.33 | 7 + 10.4 | 0.61 |
| Cirrhosis | 14 + 5.96 | 14 + 8.33 | 0 | 0.01 |
| Steatosis | 1.92 + 0.861 | 2.1 + 0.83 | 1.46 + 0.765 | < 0.01 |
| Inflammation | 0.44 + 0.62 | 0.625 + 0.663 | 0 | < 0.01 |
| Fibrosis | 0.753 + 1.11 | 1.05 + 1.21 | 0 | < 0.01 |
| Ballooning | 0.821 + 0.712 | 1.15 + 0.575 | 0 | < 0.01 |
| NAS score | 3.18 + 1.69 | 3.86 + 1.45 | 1.46 + 0.765 | < 0.01 |
AST: Aspartate transaminase; ALT: Alanine transaminase; HDL: High-density lipoprotein cholesterol; NAS: Nonalcoholic fatty liver disease activity score; NASH: Nonalcoholic steatohepatitis; LDL: Low-density lipoprotein cholesterol; BMI: Body mass index; INR: International normalized ratio; CAD: Coronary artery disease.
Histological features
Patients with NAFLD and T2DM had significantly higher mean NAS (3.5 vs 3.0, P = 0.03) and fibrosis scores (1.2 vs 0.52, P < 0.01) compared to patients with NAFLD without T2DM. In addition, a significantly higher percentage of patients with NAFLD and T2DM showed advanced fibrosis (27.3 vs 13.3, P < 0.01) and prominent ballooning (27.3 vs 13.3, P < 0.01) compared to patients with NAFLD without T2DM. However, there were no statistically significant differences between groups regarding NAS ≥ 5 (29.9 vs 20.9, P = 0.12), steatosis ≥ 2 (66.2 vs 55.1, P = 0.1) and inflammation ≥ 2 (9.1 vs 6.3, P = 0.4). On the contrary, there was a trend for presence of higher Mallory bodies in patients with NAFLD and T2DM compared to patients with NAFLD without T2DM (28.6 vs 17.7, P = 0.06) (Table 3).
Table 3.
Histological findings of patients with nonalcoholic fatty liver disease, divided according to nonalcoholic fatty liver disease without type 2 diabetes mellitus and nonalcoholic fatty liver disease with type 2 diabetes mellitus n (%)
| Parameter | All patients (n = 235) | NAFLD (without T2DM) (n = 158) | NAFLD (with T2DM) (n = 77) | P value |
| NAS score (mean ± SD) | 3.2 ± 1.7 | 3.0 ± 1.6 | 3.5 ± 1.7 | 0.03 |
| Fibrosis score (mean ± SD) | 0.75 ± 1.1 | 0.52 ± 0.9 | 1.2 ± 1.3 | < 0.01 |
| Fibrosis score | < 0.01 | |||
| < 2 | 191 (81.3) | 139 (88.0) | 52 (67.5) | |
| ≥ 2 | 44 (18.7) | 19 (12.0) | 25 (32.5) | |
| NAS score | 0.12 | |||
| < 5 | 179 (76.2) | 125 (79.1) | 54 (70.1) | |
| ≥ 5 | 56 (23.8) | 33 (20.9) | 23 (29.9) | |
| NAS score | 0.11 | |||
| < 3 | 90 (38.3) | 66 (41.7) | 24 (31.2) | |
| ≥ 3 | 145 (61.7) | 92 (58.3) | 53 (68.8) | |
| Steatosis | 0.1 | |||
| < 2 | 97 (41.3) | 71 (44.9) | 26 (33.8) | |
| ≥ 2 | 138 (58.7) | 87 (55.1) | 51 (66.2) | |
| Ballooning | < 0.01 | |||
| < 2 | 193 (82.1) | 137 (86.7) | 56 (72.7) | |
| 2 | 42 (17.9) | 21 (13.3) | 21 (27.3) | |
| Inflammation | 0.4 | |||
| < 2 | 218 (92.7) | 148 (93.7) | 70 (90.9) | |
| ≥ 2 | 17 (7.3) | 10 (6.3) | 7 (9.1) | |
| Mallory bodies | 50 (21.3) | 28 (17.7) | 22 (28.6) | 0.06 |
NAFLD: Nonalcoholic fatty liver disease; NAS: NAFLD activity score; T2DM: Type 2 diabetes mellitus.
NASH
After controlling for age, gender, ethnicity and BMI, a multivariate analysis showed significant association between NASH and presence of diabetes (OR = 3.27, 95%CI: 1.43-7.50, P < 0.01) and ALT ≥ 36 IU (OR = 3.88, 95%CI: 1.15-13.1, P = 0.029) in all patients with NAFLD, while only ALT ≥ 36 IU (OR = 4.21, 95%CI: 1.13-15.6, P = 0.03) was significantly associated with NASH in patients with NAFLD and T2DM (Table 4).
Table 4.
Variables associated with nonalcoholic steatohepatitis (nonalcoholic fatty liver disease activity score ≥ 5) (multivariate analysis)
| Variables | P value | Odds ratio | 95%CI |
| All patients | |||
| Presence of T2DM | < 0.01 | 3.27 | 1.43-7.50 |
| ALT ≥ 36 | 0.029 | 3.88 | 1.15-13.1 |
| Protein | 0.045 | 1.84 | 1.01-3.35 |
| Patients with NAFLD without T2DM | |||
| Protein | 0.048 | 3.06 | 1.01-9.25 |
| Patients with NAFLD with T2DM | |||
| ALT ≥ 36 | 0.03 | 4.21 | 1.13-15.6 |
ALT: Alanine transaminase; NAFLD: Nonalcoholic fatty liver disease; T2DM: Type 2 diabetes mellitus.
Fibrosis
After controlling for age, gender, ethnicity and BMI on multivariate analysis, only T2DM (OR = 3.45, 95%CI: 1.53-7.77, P < 0.01) and low platelet (OR = 0.99, 95%CI: 0.99-1.0, P = 0.025) showed an independent association for advanced fibrosis among all patients with NAFLD. No separate indicator for advanced fibrosis was noted among patients with NAFLD and T2DM (Table 5).
Table 5.
Variables associated with advanced fibrosis (multivariate analysis)
| Variables | P value | Odds ratio | 95%CI |
| All patients | |||
| Presence of T2DM | < 0.01 | 3.45 | 1.53-7.77 |
| Platelet | 0.025 | 0.99 | 0.99-1.0 |
| INR | < 0.01 | 4151 | 97.7-176426 |
| Patients with NAFLD without T2DM | |||
| Platelet | 0.03 | 0.99 | 0.98-0.99 |
| INR | 0.001 | 16950 | 62.4-4607293 |
| Patients with NAFLD with T2DM | |||
| Platelet | 0.045 | 0.99 | 0.98-0.99 |
NAFLD: Nonalcoholic fatty liver disease; T2DM: Type 2 diabetes mellitus; INR: International normalized ratio.
DISCUSSION
Research over the past few decades has dwelled upon identifying the potential risk factors associated with NASH and advanced fibrosis which indeed has significant prognostic value. T2DM is being increasingly recognized as an important risk factor in NAFLD progression, especially with NAFLD being regarded as an extension of metabolic syndrome[2-4]. However, there is a lack of definite predictors for NASH among NAFLD patients with T2DM[14,15], thereby highlighting the need to obtain liver histopathology for accurate determination of NASH and fibrosis. Recognizing unique histological features among NAFLD patients with T2DM would enhance our understanding of disease progression and aid development of potential remedies. Prior studies evaluating the histopathological features of NAFLD among T2DM patients utilized the Brunt system of histological evaluation[3,8,16-20] and were limited by lack of a comparative control group of patients with NAFLD without T2DM[3,16,17,20,21]. To date, only one multicenter study compared the histopathological differences among patients with NAFLD with and without T2DM[5] utilizing the NAS criteria among United States population with the limitation of potential inter-observer variability due to histology being read by multiple pathologists. Therefore, we conducted a detailed comparative histological evaluation using the NAS criteria in patients with NAFLD with and without T2DM at a single center using a single well-trained histopathologist to identify factors predicting NASH and advanced fibrosis in a multiethnic United States cohort.
Our study highlights the findings that patients with NAFLD and T2DM have higher ballooning (27.3 vs 13.3, P < 0.01) compared to patients with NAFLD without T2DM. Ballooning is considered to be the most important feature of steatohepatitis and correlates with features of insulin resistance very well[7]. Recently, Leite et al[21] examined the histological features of NAFLD in T2DM patients using the updated NASH CRN grading system and also assessed for interpathologist’s agreement on histological features. Although they demonstrated presence of significant higher ballooning (42%-55%) among patients with NAFLD and T2DM, they had several limitations. The biopsies were evaluated by two pathologists, and the kappa score for inter-observer agreement for assessment of the degree of ballooning was 0.45 indicating that there was more disagreement between the pathologists than agreement. Moreover, the study sample in Leite et al[21] lacked a comparative control group of patients with NAFLD without T2DM. Thus, our study overcomes the limitations of the Leite et al[21] study, but also demonstrates that patients with NAFLD and T2DM have higher rates of prominent ballooning compared to patients with NAFLD without T2DM using a larger and more diverse patient population including Asians, Hispanics, African Americans and Caucasians. Interestingly, in a large clinical trial involving 173 biopsy proven pediatric NAFLD patients, Lavine et al[22] demonstrated a significant improvement in hepatocyte ballooning at 96 wk of therapy with vitamin E -0.5 (95%CI: -0.8 to -0.3, P = 0.006) and metformin -0.3 (95%CI: -0.6 to -0.0, P = 0.04) compared to placebo. No other histological features of NAFLD had shown any significant improvement otherwise. Furthermore, Chen et al[23] in a large case-control study demonstrated that diabetes is associated with higher risk of HCC (OR = 2.29, 95%CI: 2.25-2.35, P < 0.001) and use of metformin resulted in 7% reduction in the risk of HCC in diabetic patients incrementally (adjusted OR = 0.93, 95%CI: 0.91-0.94, P < 0.0001) by inhibiting hepatoma proliferation and inducing cell cycle arrest. Thus, hepatocyte ballooning is not only a distinct histological feature of NASH but may play an essential role in the management and prognosis of NASH. Further studies are needed to explore the exact role of hepatocyte ballooning in the progression onto cirrhosis and HCC.
While ballooning is one component of the overall NAS, our results showed that the mean NAS was significantly higher among patients with NAFLD and T2DM (3.5 vs 3.0, P = 0.03) compared to patients with NAFLD without T2DM. However, there was no significant difference in NAS ≥ 5 (20.9 vs 29.9, P = 0.12) between the two groups in our study (Table 2) which could likely be attributed to relatively smaller sample size of the patients with NAFLD and T2DM (n = 77). NAS is different from Brunt scoring system, as it includes the active and potentially reversible features of NAFLD, such as steatosis, inflammation and ballooning, and is separate from the potentially less reversible features like fibrosis[6]. Moreover, the numeric value of the composite NAS is distinct from the qualitative histopathologic diagnosis utilized by Brunt scoring system (i.e., definite NASH, kappa score = 0.57)[7]. Furthermore, Miyaaki et al[20] and Adams et al[24] have pointed that patients with NAFLD and T2DM tend to have advanced fibrosis and less steatosis, which could be attributed to the natural progression of disease or advanced age of the population with T2DM at the time of evaluation. Thus variation among individual histological features of NASH among patients with NAFLD and T2DM could lower the overall composite NAS. Therefore, it is important to quantify individual histological features separately along with NAS while interpreting liver biopsies among NAFLD patients. Further large population-based studies are needed to explore the additional utility of NAS in patients with NAFLD and T2DM.
In our study, patients with NAFLD and T2DM demonstrated significantly higher mean fibrosis score and advanced fibrosis compared to patients with NAFLD without T2DM which corroborates with prior studies[5,8,20]. On adjusting for age, gender, BMI and ethnicity, T2DM was noted to be an independent predictor for NASH and advanced fibrosis among all patients with NAFLD in our study which is not surprising. Recently, Loomba et al[5] conducted a cross-sectional analysis of 1069 patients with NAFLD from the NAFLD Database study and PIVENS trial to examine the association of personal and family history of T2DM to histological features of NASH and advanced fibrosis. On sub-analysis using a comparative group of patients with NAFLD without T2DM and adjusting for age, sex, ethnicity and BMI, patients with NAFLD and T2DM had increased risk of NASH (OR = 2.48, 95%CI: 1.31-4.72, P = 0.01), any fibrosis (OR = 2.94, 95%CI: 1.49-5.81, P < 0.01) and advanced fibrosis (OR = 6.03, 95%CI: 3.16-11.52, P < 0.0001) which is similar to our study results. However, the diagnosis of NASH was based on modified Brunt criteria as opposed to composite NAS criteria in our study and comparative histological analysis for differences in ballooning or NAS ≥ 5 was lacking. Younossi et al[8], in a cohort study of 132 patients with NAFLD, demonstrated that patients with NAFLD and T2DM (n = 44) had significantly higher rates of cirrhosis (25 vs 10.2, P = 0.04), liver related mortality (RR = 22.83, 95%CI: 2.97-175.03, P = 0.003) and overall mortality (RR = 3.30, 95%CI: 1.76-6.18, P = 0.002), compared to patients with NAFLD without T2DM (n = 88). However, there was no statistical difference for fibrosis ≥ 2 between the two groups (17 vs 32, P = 0.07) in their study, which is contrary to our study and could be related to their smaller sample size.
Early diagnosis of NASH and advanced fibrosis in patients with NAFLD and T2DM has important clinical significance. In a long term follow up study of biopsy proven NAFLD patients, Ekstedt et al[25] reported higher mortality particularly among patients with NASH compared to reference population with respect to liver-related causes (2.8 vs 0.2, P = 0.04) and from cardiovascular disease (15.5 vs 7.5, P = 0.04). Younossi et al[8] in their study reported higher occurrence of cirrhosis (25 vs 10.2, P = 0.04), overall mortality (RR = 3.30, 95%CI: 1.76-6.18, P = 0.002) and mortality related to liver disease (RR = 22.83, 95%CI: 2.97-175.03, P = 0.003) among diabetic patients with NAFLD as opposed to non-diabetic patients with NAFLD. Moreover, Welzel et al[26], found a 2.9 fold risk for development of HCC among diabetic patients in a recent analysis within the SEER-database. Thus, it is imperative to identify and evaluate patients with NAFLD and T2DM early on for presence of NASH and advanced fibrosis. Currently, no single non-invasive panel has been proven to be a valid substitute for a liver biopsy[27]. In our study, the indirect markers for NASH such as AST/ALT ratio (0.94 ± 0.4 vs 0.78 ± 0.4, P < 0.01) and Fib-4 index (1.65 vs 1.06, P < 0.01) showed higher significance among patients with NAFLD with T2DM compared to patients without T2DM while APRI score (0.64 vs 0.54, P = 0.16) did not. The utility of most non-invasive tests is limited as a screening test to exclude advanced fibrosis due to high negative predictive value (76.9%-90.5%) and a modest positive predictive value (36.1%-61.1%)[28]. Moreover, patients with NAFLD and BMI > 30 (OR = 8.4, 95%CI: 6.6-10.8, P < 0.0001) and T2DM (OR = 2.0, 95%CI: 1.5-2.6, P < 0.0001) pose unique challenge for accurate liver stiffness measurement using Fibroscan, mainly due to attenuation of elastic waves by subcutaneous and pre-hepatic fat thickness[29]. Therefore, a selective liver biopsy could be helpful among obese, elderly patients especially with NAFLD and T2DM to accurately stage NAFLD. Furthermore, our results (Table 5) comparing the background features among patients with NASH and steatosis showed higher number of patients with NASH to have T2DM and cirrhosis in addition to indirect evidence for advanced disease and fibrosis reflected by significantly higher Bilirubin, AST and ALT and lower platelet and lower HDL values. Therefore optimization of risk factors of NAFLD such as metabolic syndrome, T2DM and dyslipidemia to prevent further progression of liver disease to cirrhosis, HCC and other related complications is warranted.
Our study has some limitations. Data on duration of T2DM and insulin resistance were lacking, hence we could not assess for the association of these variables with NAFLD progression. However, the mean HbA1c among patients with NAFLD and T2DM was 8.09 which reflects uncontrolled T2DM thereby minimizing the confounding effect of diabetic medication and highlighting the role of uncontrolled T2DM in NAFLD progression. Furthermore, the data being obtained at a major tertiary care referral center, could have led to selection bias. Lastly, the smaller sample size of patients with NAFLD and T2DM limited our ability to show statistically significant rate of cirrhosis. Therefore, larger prospective studies on patients with NAFLD with and without T2DM are needed to further our understanding of the relationship between T2DM and NAFLD progression.
In conclusion, patients with NAFLD and T2DM tend to have higher ballooning and advanced fibrosis compared to patients with NAFLD without T2DM. Our study also demonstrated that T2DM is an independent predictor for both NASH and advanced fibrosis, while utilizing the more current NAS criteria. Patients with NAFLD and T2DM in general should be advised and educated regarding optimal diabetes control, hyperlipidemia management and vascular disease screening which may not only prevent cardiovascular complications but also could prevent further progression to NASH and/or advanced fibrosis/cirrhosis. Future prospective studies are needed to explore the role of NASH (especially hepatocyte ballooning) among patients with and without T2DM and development of advanced fibrosis, HCC and cardiovascular disease burden and mortality.
COMMENTS
Background
Identifying the unique histopathological features of nonalcoholic fatty liver disease (NAFLD) among patients with type 2 diabetes mellitus (T2DM) is not only important to stage the disease progression but also help us understand the impact of diabetes in progression of NAFLD. Presently, only few studies have described the histopathological differences among NAFLD patients with and without T2DM in a multiethnic United States population cohort.
Research frontiers
While screening for diabetes and other risk factors for NAFLD is easily performed, management of NAFLD is still evolving. Current interest is in understanding the pathogenesis of NAFLD progression, which will pave way to potential novel treatments.
Innovations and breakthroughs
The authors’ study highlights the findings that patients with NAFLD and T2DM have higher ballooning compared to patients with NAFLD without T2DM. Ballooning is considered to be the most important feature of steatohepatitis and correlates with features of insulin resistance very well. Similarly, significantly higher mean fibrosis score and advanced fibrosis is demonstrated among patients with NAFLD and T2DM. It is important to quantify individual histological features separately along with NAS while interpreting liver biopsies among NAFLD patients.
Applications
T2DM is an independent predictor for both NASH and advanced fibrosis. Hepatocyte ballooning is not only a distinct histological feature of NASH but may play an essential role in the management and prognosis of NASH. Patients with NAFLD and T2DM in general should be advised and educated regarding optimal diabetes control, hyperlipidemia management and vascular disease screening thereby preventing progression of liver disease burden and mortality.
Terminology
NAFLD includes a histological spectrum of liver diseases ranging from simple steatosis to NASH and the latter histological entity can progress to cirrhosis. T2DM is shown to be an independent risk factor for the development of NASH. The components that make up NASH include fat in liver cells (steatosis), inflammation, scar tissue (fibrosis) and degeneration of liver cell (ballooning).
Peer-review
This is an excellent work dealing with a very interesting topic, the histopathological alterations in diabetic and non-diabetic patients with NAFLD. There are few studies describing the mentioned differences and this work constitutes a novel approach. The authors showed clearly the impact of T2DM on NAFLD. Comorbidity seems to be very important in converting the disease and determining the severity of the disease either ways.
Footnotes
P- Reviewer: Lee YY, Ramos S, Sazci A, Takaki A, Zhang XC S- Editor: Tian YL L- Editor: A E- Editor: Liu SQ
Institutional review board statement: Noted under methods section of manuscript. This study was approved by the Institutional Review Board at University of Chicago Medical Center.
Informed consent statement: Noted under methods section of manuscript. There was no identifiable data in this study and informed consent was waived by University of Chicago Medical Center Institutional Review Board.
Conflict-of-interest statement: We do not have any commercial relationships (i.e., consultancies, patent-licensing agreements) or any conflicting interests (including but not limited to commercial, personal, political, intellectual, or religious interests) that might pose a conflict of interest in connection with the submitted manuscript.
Data sharing statement: Technical appendix, statistical code, and dataset available from the corresponding author at srm9006@nyp.org. Consent from participants was not obtained as the presented data is anonymized and risk of identification is nil.
Open-Access: This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Peer-review started: April 18, 2015
First decision: June 3, 2015
Article in press: October 27, 2015
References
- 1.Choudhury J, Sanyal AJ. Clinical aspects of fatty liver disease. Semin Liver Dis. 2004;24:349–362. doi: 10.1055/s-2004-860864. [DOI] [PubMed] [Google Scholar]
- 2.Marchesini G, Bugianesi E, Forlani G, Cerrelli F, Lenzi M, Manini R, Natale S, Vanni E, Villanova N, Melchionda N, et al. Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome. Hepatology. 2003;37:917–923. doi: 10.1053/jhep.2003.50161. [DOI] [PubMed] [Google Scholar]
- 3.Gupte P, Amarapurkar D, Agal S, Baijal R, Kulshrestha P, Pramanik S, Patel N, Madan A, Amarapurkar A. Non-alcoholic steatohepatitis in type 2 diabetes mellitus. J Gastroenterol Hepatol. 2004;19:854–858. doi: 10.1111/j.1440-1746.2004.03312.x. [DOI] [PubMed] [Google Scholar]
- 4.Fan JG. Impact of non-alcoholic fatty liver disease on accelerated metabolic complications. J Dig Dis. 2008;9:63–67. doi: 10.1111/j.1751-2980.2008.00323.x. [DOI] [PubMed] [Google Scholar]
- 5.Loomba R, Abraham M, Unalp A, Wilson L, Lavine J, Doo E, Bass NM. Association between diabetes, family history of diabetes, and risk of nonalcoholic steatohepatitis and fibrosis. Hepatology. 2012;56:943–951. doi: 10.1002/hep.25772. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ, Cummings OW, Ferrell LD, Liu YC, Torbenson MS, Unalp-Arida A, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology. 2005;41:1313–1321. doi: 10.1002/hep.20701. [DOI] [PubMed] [Google Scholar]
- 7.Brunt EM, Kleiner DE, Wilson LA, Belt P, Neuschwander-Tetri BA. Nonalcoholic fatty liver disease (NAFLD) activity score and the histopathologic diagnosis in NAFLD: distinct clinicopathologic meanings. Hepatology. 2011;53:810–820. doi: 10.1002/hep.24127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Younossi ZM, Gramlich T, Matteoni CA, Boparai N, McCullough AJ. Nonalcoholic fatty liver disease in patients with type 2 diabetes. Clin Gastroenterol Hepatol. 2004;2:262–265. doi: 10.1016/s1542-3565(04)00014-x. [DOI] [PubMed] [Google Scholar]
- 9.Brunt EM, Janney CG, Di Bisceglie AM, Neuschwander-Tetri BA, Bacon BR. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol. 1999;94:2467–2474. doi: 10.1111/j.1572-0241.1999.01377.x. [DOI] [PubMed] [Google Scholar]
- 10.American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2014;37 Suppl 1:S81–S90. doi: 10.2337/dc14-S081. [DOI] [PubMed] [Google Scholar]
- 11.Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. 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;285:2486–2497. doi: 10.1001/jama.285.19.2486. [DOI] [PubMed] [Google Scholar]
- 12.World Health Organization. Definition, Diagnosis and Classification of Diabetes Mellitus and its Complications; Part 1: Diagnosis and Classification of Diabetes Mellitus. Geneva: WHO Department of Noncommunicable Disease Surveillance; 1999. [Google Scholar]
- 13.Mohanty SR, Troy TN, Huo D, O’Brien BL, Jensen DM, Hart J. Influence of ethnicity on histological differences in non-alcoholic fatty liver disease. J Hepatol. 2009;50:797–804. doi: 10.1016/j.jhep.2008.11.017. [DOI] [PubMed] [Google Scholar]
- 14.Angulo P, Keach JC, Batts KP, Lindor KD. Independent predictors of liver fibrosis in patients with nonalcoholic steatohepatitis. Hepatology. 1999;30:1356–1362. doi: 10.1002/hep.510300604. [DOI] [PubMed] [Google Scholar]
- 15.Angulo P, Hui JM, Marchesini G, Bugianesi E, George J, Farrell GC, Enders F, Saksena S, Burt AD, Bida JP, et al. The NAFLD fibrosis score: a noninvasive system that identifies liver fibrosis in patients with NAFLD. Hepatology. 2007;45:846–854. doi: 10.1002/hep.21496. [DOI] [PubMed] [Google Scholar]
- 16.Amarapurka DN, Amarapurkar AD, Patel ND, Agal S, Baigal R, Gupte P, Pramanik S. Nonalcoholic steatohepatitis (NASH) with diabetes: predictors of liver fibrosis. Ann Hepatol. 2006;5:30–33. [PubMed] [Google Scholar]
- 17.Prashanth M, Ganesh HK, Vima MV, John M, Bandgar T, Joshi SR, Shah SR, Rathi PM, Joshi AS, Thakkar H, et al. Prevalence of nonalcoholic fatty liver disease in patients with type 2 diabetes mellitus. J Assoc Physicians India. 2009;57:205–210. [PubMed] [Google Scholar]
- 18.Shima T, Uto H, Ueki K, Takamura T, Kohgo Y, Kawata S, Yasui K, Park H, Nakamura N, Nakatou T, et al. Clinicopathological features of liver injury in patients with type 2 diabetes mellitus and comparative study of histologically proven nonalcoholic fatty liver diseases with or without type 2 diabetes mellitus. J Gastroenterol. 2013;48:515–525. doi: 10.1007/s00535-012-0653-5. [DOI] [PubMed] [Google Scholar]
- 19.Jaskiewicz K, Rzepko R, Sledzinski Z. Fibrogenesis in fatty liver associated with obesity and diabetes mellitus type 2. Dig Dis Sci. 2008;53:785–788. doi: 10.1007/s10620-007-9942-x. [DOI] [PubMed] [Google Scholar]
- 20.Miyaaki H, Ichikawa T, Nakao K, Yatsuhashi H, Furukawa R, Ohba K, Omagari K, Kusumoto Y, Yanagi K, Inoue O, et al. Clinicopathological study of nonalcoholic fatty liver disease in Japan: the risk factors for fibrosis. Liver Int. 2008;28:519–524. doi: 10.1111/j.1478-3231.2007.01614.x. [DOI] [PubMed] [Google Scholar]
- 21.Leite NC, Villela-Nogueira CA, Pannain VL, Bottino AC, Rezende GF, Cardoso CR, Salles GF. Histopathological stages of nonalcoholic fatty liver disease in type 2 diabetes: prevalences and correlated factors. Liver Int. 2011;31:700–706. doi: 10.1111/j.1478-3231.2011.02482.x. [DOI] [PubMed] [Google Scholar]
- 22.Lavine JE, Schwimmer JB, Van Natta ML, Molleston JP, Murray KF, Rosenthal P, Abrams SH, Scheimann AO, Sanyal AJ, Chalasani N, et al. Effect of vitamin E or metformin for treatment of nonalcoholic fatty liver disease in children and adolescents: the TONIC randomized controlled trial. JAMA. 2011;305:1659–1668. doi: 10.1001/jama.2011.520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Chen HP, Shieh JJ, Chang CC, Chen TT, Lin JT, Wu MS, Lin JH, Wu CY. Metformin decreases hepatocellular carcinoma risk in a dose-dependent manner: population-based and in vitro studies. Gut. 2013;62:606–615. doi: 10.1136/gutjnl-2011-301708. [DOI] [PubMed] [Google Scholar]
- 24.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. doi: 10.1016/j.jhep.2004.09.012. [DOI] [PubMed] [Google Scholar]
- 25.Ekstedt M, Franzén LE, Mathiesen UL, Thorelius L, Holmqvist M, Bodemar G, Kechagias S. Long-term follow-up of patients with NAFLD and elevated liver enzymes. Hepatology. 2006;44:865–873. doi: 10.1002/hep.21327. [DOI] [PubMed] [Google Scholar]
- 26.Welzel TM, Graubard BI, Zeuzem S, El-Serag HB, Davila JA, McGlynn KA. Metabolic syndrome increases the risk of primary liver cancer in the United States: a study in the SEER-Medicare database. Hepatology. 2011;54:463–471. doi: 10.1002/hep.24397. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Musso G, Gambino R, Cassader M, Pagano G. Meta-analysis: natural history of non-alcoholic fatty liver disease (NAFLD) and diagnostic accuracy of non-invasive tests for liver disease severity. Ann Med. 2011;43:617–649. doi: 10.3109/07853890.2010.518623. [DOI] [PubMed] [Google Scholar]
- 28.Wong VW, Vergniol J, Wong GL, Foucher J, Chan HL, Le Bail B, Choi PC, Kowo M, Chan AW, Merrouche W, et al. Diagnosis of fibrosis and cirrhosis using liver stiffness measurement in nonalcoholic fatty liver disease. Hepatology. 2010;51:454–462. doi: 10.1002/hep.23312. [DOI] [PubMed] [Google Scholar]
- 29.Castéra L, Foucher J, Bernard PH, Carvalho F, Allaix D, Merrouche W, Couzigou P, de Lédinghen V. Pitfalls of liver stiffness measurement: a 5-year prospective study of 13,369 examinations. Hepatology. 2010;51:828–835. doi: 10.1002/hep.23425. [DOI] [PubMed] [Google Scholar]