Patatin-like phospholipase domain-containing 3 I148M affects liver steatosis in patients with chronic hepatitis B

Authors

  • Mauro Viganò,

    1. Hepatology Unit, Ospedale San Giuseppe, Università degli Studi di Milano, Milano, Italy
    Search for more papers by this author
    • These authors contributed equally.

  • Luca Valenti,

    Corresponding author
    • Internal Medicine 1B, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author
    • These authors contributed equally.

  • Pietro Lampertico,

    1. Centro A.M. e A. Migliavacca, First Division of Gastroenterology, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author
  • Floriana Facchetti,

    1. Centro A.M. e A. Migliavacca, First Division of Gastroenterology, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author
  • Benedetta Maria Motta,

    1. Internal Medicine 1B, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author
  • Roberta D'Ambrosio,

    1. Centro A.M. e A. Migliavacca, First Division of Gastroenterology, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author
  • Solange Romagnoli,

    1. Pathology Unit, Department of Health Sciences, San Paolo Hospital, Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author
  • Paola Dongiovanni,

    1. Internal Medicine 1B, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author
  • Benedetta Donati,

    1. Internal Medicine 1B, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author
  • Silvia Fargion,

    1. Internal Medicine 1B, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author
  • Massimo Colombo

    1. Centro A.M. e A. Migliavacca, First Division of Gastroenterology, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy
    Search for more papers by this author

  • Potential conflict of interest: Mauro Viganò has worked in a speaking and teaching capacity for Bristol-Myers Squibb, Roche, and Gilead Sciences. Massimo Colombo has been awarded grant and research support from MSD, Roche, Bristol-Myers Squibb, and Gilead Sciences, has served on the advisory committees of MSD, Roche, Novartis, Bayer, Bristol-Myers Squibb, Gilead Sciences, Tibotec, and Vertex, and has worked in a speaking and teaching capacity for Tibotec, MSD, Roche, Novartis, Bayer, Bristol-Myers Squibb, Gilead Sciences, and Vertex. Pietro Lampertico has served on the advisory board/speaker bureau for Bristol-Myers Squibb, Roche, Gilead Sciences, and GlaxoSmithKline.

Address reprint requests to: Luca Valenti, M.D., Ph.D., Fondazione IRCCS Ca' Granda–Ospedale Maggiore Policlinico, Università degli Studi di Milano, Via F. Sforza 35, 20122 Milan, Italy. E-mail: luca.valenti@unimi.it; fax: 00390250320296.

Abstract

Steatosis is a common histopathological feature of chronic hepatitis B (CHB) and has been associated with severity of liver disease. Recently, the rs738409 I148M patatin-like phospholipase domain-containing 3 (PNPLA3) polymorphism has been demonstrated to influence steatosis susceptibility and fibrosis progression in patients with different liver diseases, but no data are yet available for CHB. The aim of this study was to evaluate whether PNPLA3 I148M influences steatosis susceptibility in a large series of patients with CHB. We enrolled 235 treatment-naïve CHB patients consecutively examined by percutaneous liver biopsy. In ≥2-cm-long liver tissue cores, steatosis and fibrosis were staged by Kleiner and METAVIR scores, respectively. The I148M polymorphism was determined by Taqman assays. Steatosis was present in 146 (62%) patients, of whom 24 (10%) had severe (>33% of hepatocytes) steatosis. Steatosis was independently associated with age (odds ratio [OR]: 2.67; confidence interval [CI]: 1.50-4.92; for age ≥50 years), body mass index (BMI; OR, 2.84; CI, 1.30-6.76; for BMI ≥27.5 kg/m2), diabetes or impaired fasting glucose (OR, 4.45; CI, 1.10-30.0), and PNPLA3 148M allele (OR, 1.62; CI, 1.00-7.00; for each 148M allele). Independent predictors of severe steatosis were BMI (OR, 3.60; CI, 1.39-9.22; for BMI ≥27.5 kg/m2) and PNPLA3 148M allele (OR, 6.03; CI, 1.23-5.0; for each 148M allele). PNPLA3 148M alleles were associated with a progressive increase in severe steatosis in patients with acquired cofactors, such severe overweight and a history of alcohol intake (P = 0.005). Conclusion: In CHB patients, the PNPLA3 I148M polymorphism influences susceptibility to steatosis and, in particular, when associated with severe overweight and alcohol intake, severe steatosis. (Hepatology 2013;58:1245–1252)

Abbreviations
Ab

antibody

ALT

alanine aminotransferase

AST

aspartate aminotransferase

BCP

basal core promoter

BMI

body mass index

CHB

chronic hepatitis B

CHC

chronic hepatitis C

CI

confidence interval

GGT

gamma-glutamyl transferase

HBeAg

hepatitis B e antigen

HbsAg

hepatitis B surface antigen

HBV

hepatitis B virus

HCC

hepatocellular carcinoma

HCV

hepatitis C virus

HDV

hepatitis delta virus

HIV

human immunodeficiency virus

IFG

impaired fasting glucose

INR

international normalized ratio

IQR

interquartile range

IR

insulin resistance

MetS

metabolic syndrome

NAS

nonalcoholic fatty liver disease activity score

OR

odds ratio

PC

precore

PCR

polymerase chain reaction

PNPLA3

patatin-like phospholipase domain-containing 3

TG

triglyceride.

Chronic hepatitis B virus (HBV) infection is estimated to affect more than 350 million people worldwide and is one of the leading causes of cirrhosis, hepatocellular carcinoma (HCC), and anticipated liver-related mortality.[1] Liver steatosis, strongly associated with obesity and metabolic syndrome (MetS), is very prevalent and a common cause of chronic liver disease in the general population. On the other hand, steatosis also represents a common histopathological feature of chronic hepatitis B (CHB) patients,[2-4] being observed in nearly 30% of cases (ranging from 14% to 73%).[5-17] Steatosis in CHB seems to be favored by risk factors defining MetS, such as increased body mass index (BMI), central adiposity, dyslipidemia, insulin resistance (IR), and diabetes.[5-17] Differently from patients with chronic hepatitis C (CHC), most reports failed to demonstrate any association between steatosis and viral factors.[6, 12, 13, 18] Although MetS and steatosis have been negatively associated with hepatitis B surface antigen (HBsAg) positivity in Asian subjects,[19-21] overall evidence suggests that they contribute to CHB progression.[17, 22-24] However, the role of host genetic factors in the pathogenesis of steatosis in CHB patients has never been assessed before.

Recently, patatin-like phospholipase domain-containing 3 (PNPLA3), also known as adiponutrin, rs738409 C>G single-nucleotide polymorphism, encoding for the I148M protein variant, has been recognized as a genetic determinant of liver fat content, independently of IR and serum lipids.[25-33] It is believed that the 148M allele alters the enzymatic activity shifting the balance from predominantly lipase activity toward de novo lipogenesis.[34-36] PNPLA3 I148M is also a genetic determinant of liver damage progression associated with steatohepatitis.[29, 31, 37-41] Therefore, we investigated whether the PNPLA3 I148M polymorphism influences liver fat accumulation in a large series of Italian patients with biopsy-proven CHB.

Patients and Methods

From an initial cohort of 306 treatment-naïve patients with CHB consecutively referred for a liver biopsy and concomitant transient elastography evaluation at the Liver Center, Fondazione IRCCS Ca' Granda Ospedale Policlinico (Milan, Italy) between January 2007 and March 2012, we considered 235 (77%) patients with stored DNA for genetic analysis, genetic testing consent, and without regular use of steatosis-inducing drugs that were retrospectively enrolled in this study. Part of this cohort has been reported on in a previous publication.[42] In all cases, diagnosis of CHB was carried out in the presence of serum HBsAg, persistently or intermittently abnormal alanine aminotransferase (ALT) values, and serum HBV DNA >2,000 IU/mL lasting for >6 months. Patients with hepatitis C virus (HCV), hepatitis delta virus (HDV), and human immunodeficiency (HIV) virus coinfections, other concomitant liver diseases, current or previous hepatic decompensation, current or previous antiviral treatment, and/or an absolute contraindications to liver biopsy (platelets, <60 × 109/L; INR, >1.35) were excluded from the original study. None of the patients reported consumption of cannabinoids.

All patients were investigated to assess clinical features, anthropometric parameters, liver enzymes (aspartate aminotransferase [AST], ALT, and gamma-glutamyl transferases [GGTs]), liver function tests (bilirubin and international normalized ratio [INR]), indices of portal hypertension (serum platelets), and serologic markers of HBV replication. Clinical, biochemical, and anthropometric data as well as daily alcohol intake during the previous 5 years and/or current history of alcohol intake were assessed at the time of liver biopsy. Positive alcohol intake was defined in the presence of reported regular consumption of any amount of alcohol. BMI was calculated on the basis of weight in kilograms and height, and subjects were classified as with or without severe overweight (BMI, ≥27.5 kg/m2). Diagnosis of diabetes was based on detection of fasting blood glucose of ≥126 mg/dL on at least two occasions.[43] In patients with a previous diagnosis of diabetes, current therapy with oral hypoglycemic agents was documented.

Demographic, clinical, and genetic features of subjects included are presented in Table 1. Informed written consent was obtained from each subject. The study conforms to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the institutional review board of the Fondazione IRCCS Ca' Granda.

Table 1. Demographic and Clinical Features of 235 Italian Patients With CHB in the Overall Series and in Patients Subdivided According to PNPLA3 I148M Polymorphism Status
Demographic and Clinical FeaturesOverallPNPLA3 I148MP Valuea
 I/II/MM/M
  1. a

    For trend for increasing number of 148M alleles.

  2. b

    Four patients were being treated with statins (P = not significant according to PNPLA3 I148M polymorphism).

  3. c

    Of the patients affected by diabetes, 7 were being treated with oral hypoglycemic agents and 1 with insulin (P = not significant according to PNPLA3 I148M polymorphism).

  4. d

    Median and IQR for non-normally distributed variables.

  5. e

    Available in 217 patients.

  6. f

    Available in 210 patients. Severe steatosis: grade 2-3 Kleiner.

  7. g

    Available in 213 patients. Advanced fibrosis: METAVIR stage 3-4; cirrhosis: METAVIR 4.

  8. Abbreviation: IGT, impaired glucose tolerance.

Number (%)235136 (58)87 (37)12 (5) 
Female (%)65 (28)42 (31)20 (23)3 (25)0.430
Age, years48 ± 1248 ± 1247 ± 1250 ± 130.780
BMI, kg/m224.6±3.524.6±3.624.4±3.525.7±3.30.490
Alcohol intake (%)49 (21)28 (21)18 (21)3 (25)0.940
Cholesterol, mg/dLb190 ± 40191 ± 42189 ± 38183 ± 360.870
TGs, mg/dL104 ± 46100 ± 42108 ± 50122 ± 580.130
Glucose, mg/dL88 ± 1690 ± 1887 ± 1383 ± 70.100
Diabetes or IFGc (%)16 (7)13 (10)3 (3)00.046
ALT, IU/Ld75 (49-133)79 (46-151)72 (51-118)75 (62-131)0.190
AST, IU/Ld50 (37-81)49 (37-81)50 (37-84)61 (35-88)0.800
GGT, IU/Ld33 (24-63)34 (23-60)32 (24-76)30 (25-41)0.790
Bilirubin, mg/dLd0.6 (0.4-0.9)0.7 (0.5-0.8)0.6 (0.5-0.9)0.7 (0.7-1.0)0.990
Platelets, ×103/mm3181 ± 48183 ± 55182 ± 48183 ± 550.740
Prothrombin time, INR1.08 ± 0.11.09 ± 0.131.07 ± 0.101.05 ± 0.040.140
HBV genotype: A/D/othere44/148/25 (20/68/12%)19/91/16 (15/72/13%)24/49/9 (29/60/11%)1/8/0 (11/89/0%)0.420
HBV DNA, log10, IU/mLd6.3 (4.9-7.6)6.3 (5.0-7.6)6.0 (4.8-7.8)6.8 (5.4-7.3)0.770
HBeAg (%)47 (26)26 (1929 (33)4 (33)0.046
Basal core promoter mutationsf (%)     
A1762T136 (65)81 (67)49 (60)6 (67)0.440
G1764A140 (67)82 (68)52 (64)6 (67)0.610
Precore mutationf (%)     
G1896A137 (65)88 (73)42 (52)77 (78)0.032
Steatosis146 (62)80 (59)55 (63)11 (92)0.045
Severe steatosis24 (10)8 (6)13 (15)3 (25)0.005
NAS >2g109 (51)71 (52)50 (57)10 (83)0.070
NAS fibrosis 3-4g85 (40)   0.500
Advanced fibrosis94 (40)53 (39)35 (40)6 (50)0.550
Cirrhosis48 (20)28 (21)18 (21)2 (17)0.940

Serum Assays

Laboratory data of all patients, including AST and ALT, GGT, alkaline phosphatase, blood cell count, fasting plasma glucose, cholesterol and triglyceride (TG) serum levels, were measured in all patients by standard laboratory procedures. Serum HBsAg, hepatitis B e antigen (HBeAg), and antibody (Ab) to hepatitis B e antigen were detected by microparticle enzyme immunoassay (AXSYM; Abbot Laboratories, North Chicago, IL). Serum HBV DNA was assessed by a real-time polymerase chain reaction (PCR) assay (COBAS TaqMan HBV; Roche Molecular Systems, Inc., Branchburg, NJ), with a lower limit of quantification of 12 IU/mL. HBV genotypes were determined using the INNO-LiPA HBV Genotyping assay (Innogenetics NV, Ghent, Belgium). This kit is a line probe assay designed to identify HBV genotypes A-H by detection of type-specific sequences in the HBV polymerase gene domain B-C. Purified DNA was amplified over two rounds of PCR using biotinylated PCR primers, according to the instructions of the manufacturer. Mutations in the HBV precore (PC) and basal core promoter (BCP) region were detected by INNO-LiPA HBV preCore (Innogenetics NV). Except for primers and reaction strips, the procedure was similar to that for HBV genotyping. Probes were designed to determine nucleotide sequences at position 1896 in the PC region (G versus A) and positions 1762 (A versus T) and 1764 (G versus A and G versus T) in the BCP region. Commercially available enzyme immunoassays were used to determine Abs to HCV, HDV, and HIV.

Histological Assessment

All patients underwent an ultrasound-guided liver biopsy with a semiautomatic modified Menghini system (16 G, BioMol; Hospital Service, Pomezia, Italy; and iU22; Philips, Bothell, WA). Examinations were carried out by two highly experienced pathologists (with experience in liver disease). Liver specimens were considered of adequate size if longer than 2 cm, and patients with a smaller specimen underwent repeated procedures during the same session. Five-micron-thick sections of formalin-fixed, paraffin-embedded liver tissue were stained with hematoxylin and eosin and Masson trichrome and were read by a liver pathologist (R.D.) who was blind to clinical data. Staging was evaluated according to METAVIR score (staging F0 = fibrosis absent; F1 = portal fibrosis without septa; F2 = portal fibrosis with few septa; F3 = severe fibrosis; F4 = cirrhosis).[44] Advanced fibrosis was defined in the presence of bridging fibrosis or cirrhosis (METAVIR stage 3-4). Steatosis was quantified as follows: grade 0: absent or <5% of hepatocytes involved; grade 1: 5%-33%; grade 2: 34%-66%; and grade 3: >66% of hepatocytes affected, according to the nonalcoholic fatty liver disease activity score (NAS).[45] Henceforth, we refer to mild steatosis as grade 1 steatosis and to severe steatosis as grade 2-3 steatosis. Lobular necroinflammation, ballooning, and fibrosis were also scored according to the NAS in 213 patients (91%), for whom histological samples were still available for a further reevaluation by an expert pathologist (S.R.).[45]

Genetic Analysis

The rs738409 I148M PNPLA3 polymorphism was genotyped in a blinded fashion on DNA specimens obtained by peripheral blood by a 5' nuclease Taqman assay (assay on demand for rs738409; Applied Biosystems, Foster City, CA), as previously described.[37]

Statistical Analysis

Results are expressed as means ± standard deviation or median (interquartile range; IQR) according to data distribution. Mean values were compared by analysis of variance and frequencies by chi-square test, according to data distribution, and differences were considered significant when P ≤ 0.05 (two-tailed). Non-normally distributed variables were log-transformed before analysis. Our sample had >95% power of detecting an OR of 1.5 for steatosis, of the 148M PNPLA3 allele, with a significance of 5%. The association between the PNPLA3 I148M polymorphism and steatosis (dependent variable) was evaluated by logistic regression analysis adjusted for confounding variables, which included those selected a priori for their biological relevance plus those that were found to be associated with the outcome of interest at univariate analysis (specified below). Analyses were carried out with JMP 9.0 statistical analysis software (SAS Institute Inc., Cary, NC).

Results

Clinical Features of Patients, Prevalence, and Clinical Determinants of Steatosis

Clinical characteristics of CHB patients are summarized in Table 1. Most patients were HBeAg-negative men with normal body weight and no significant alcohol consumption. Mild steatosis (5%-33% of hepatocytes involved) was present in 146 (62%) patients, whereas severe steatosis (≥33% of hepatocytes) was present in 24 (10%) patients. Advanced fibrosis (METAVIR stage 3-4) was detected in 94 patients (40%).

Variables significantly associated with steatosis severity are presented in Table 2. As expected, severity of steatosis was significantly associated with older age, male sex, and higher BMI, whereas it was not significantly associated with regular consumption of any amount of alcohol. A higher prevalence of hyperglycemia was observed in patients with mild steatosis, whereas TGs increased progressively with steatosis severity. There was also an increase in fibrosis stage associated with lower platelets in patients with steatosis.

Table 2. Variables Associated With Steatosis Severity at Univariate Analysis in 235 Patients With CHBa
VariablesNo Steatosis (Grade 0; <5%)Mild Steatosis (Grade 1; 5%-32%)Severe Steatosis (Grade 2-3; >33%)P Value
  1. Values in parentheses represent median and IQR. Advanced fibrosis: METAVIR stage 3-4.

  2. a

    Shown for P < 0.1.

Number (%)89 (38)122 (52)24 (10) 
Female (%)30 (34)31 (25)4 (17)0.0650
Age, years43 ± 1250 ± 1153 ± 12<0.0001
BMI, kg/m222.9 ± 3.025.2 ± 3.427.3 ± 3.6<0.0001
BMI >27.5 kg/m2 (%)9 (11)26 (22)10 (42)0.0007
Advanced fibrosis (%)29 (33)57 (47)8 (33)0.0900
Cirrhosis (%)11 (12)33 (27)4 (17)0.0290
Platelets, ×103/mm3191 ± 51172 ± 45182 ± 460.0110
Glucose, mg/dL85 ± 1391 ± 1888 ± 130.0150
Diabetes or IFG (%)2 (3)13 (11)1 (4)0.0430
TGs, mg/dL96 ± 32105 ± 49128 ± 650.0440
PNPLA3 I148M (%)   0.0200
I/I56 (63)72 (59)8 (33) 
I/M32 (36)42 (34)13 (54) 
M/M1 (1)8 (7)3 (13) 

Liver Steatosis and PNPLA3 Polymorphism

Prevalence of the 148M PNPLA3 allele increased progressively with severity of steatosis (P = 0.020; Table 2).

Clinical features of patients subdivided according to I148M PNPLA3 polymorphism are reported in Table 1. The 148M PNPLA3 allele was significantly associated with steatosis (P = 0.045), but, in particular, with severe steatosis (P = 0.005), whereas a trend was observed for association between the 148M allele and a NAS >2, consistent with the presence of steatohepatitis (P = 0.07). The 148M allele was not associated with fibrosis in the whole series of patients. There was a negative association between the 148M PNPLA3 allele and diabetes or impaired fasting glucose (IFG; P = 0.046) as well as between the 148M allele and HBeAg positivity (P = 0.046) and the precore mutation (P = 0.032).

Independent predictors of steatosis, severe steatosis, and NAS >2 at multivariate logistic regression analysis are presented in Table 3. Steatosis of any degree was independently associated with older age (OR, 2.67; CI, 1.50-4.92; for age ≥50 years, median value), higher BMI (OR, 2.84; CI, 1.30-6.76; for BMI ≥27.5 kg/m2), presence of diabetes or IFG (OR, 4.45; CI, 1.10-30.00), and the PNPLA3 148M allele (OR, 1.62; CI, 1.00-7.00; per each 148M allele). The only independent predictors of advanced steatosis were higher BMI (OR, 3.60; CI, 1.39-9.22;for BMI ≥27.5 kg/m2) and the 148M PNPLA3 allele (OR, 6.03; CI, 1.23-5.00; per each 148M allele). Similarly, higher BMI (OR, 2.38; CI, 1.22-4.82; for BMI ≥27.5 kg/m2) and the 148M PNPLA3 allele (OR, 1.70; CI, 1.07-2.74; per each 148M allele) were independently associated with NAS >2.

Table 3. Clinical and Genetic Factors Independently Associateda With the Presence of Steatosis (Left Panel), Severe Steatosis (Middle Panel) in 235 Italian Patients With CHB, and of NAS >2, Consistent With Steatohepatitis, in 210 Patients (Right Panel)
Clinical and Genetic FactorsSteatosis of Any DegreeSevere SteatosisNAS >2
OR (95% CI)P ValueOR (95% CI)P ValueOR (95% CI)P Value
  1. a

    At multivariate logistic regression analysis considering clinical risk factors and the I148M PNPLA3 polymorphism.

Age (≥50 years)2.67 (1.50-4.92)0.0011.82 (0.74-4.62)0.1901.67 (0.95-2.95)0.070
Female gender0.58 (0.30-1.12)0.1100.44 (0.11-1.35)0.1800.58 (0.31-1.04)0.070
BMI (≥27.5 kg/m2)2.84 (1.30-6.76)0.0083.60 (1.39-9.22)0.0072.38 (1.22-4.82)0.011
Diabetes or IFG4.45 (1.10-30.0)0.0341.81 (0.30-55)0.5901.36 (0.62-3.03)0.420
Alcohol intake1.56 (0.62-4.19)0.3502.11 (0.65-6.55)0.7401.21 (0.64-2.32)0.550
148M PNPLA3 allele1.62 (1.00-7.0)0.0506.03 (1.23-5.0)0.0111.70 (1.07-2.74)0.023

Interaction Between PNPLA3 and Acquired Risk Factors in the Pathogenesis of Steatosis

Because the phenotypic expression of the I148M PNPLA3 polymorphism has been reported to be dependent on the presence of acquired cofactors triggering steatosis, including obesity and alcohol, we next evaluated whether the association of the 148M allele and severe steatosis was dependent on the presence of severe overweight (BMI, ≥27.5 kg/m2) and a positive history of alcohol intake. Either one of these acquired risk factors was present in 82 (35%) of patients, and this condition was associated with a higher prevalence of steatosis (60 of 82 [73%] versus 86 of 153 [56%]; P = 0.01) and severe steatosis (13 of 82 [16%] versus 11 of 153 [7%]; P = 0.04). The PNPLA3 148M allele was associated with a progressive increase in the prevalence of severe steatosis in patients with, but not in those without, acquired cofactors, that is, severe overweight and regular consumption of any amount of alcohol (Fig. 1; P = 0.001 in patients with cofactors).

Figure 1.

Interaction between PNPLA3 genotype and alcohol intake and severe overweight (BMI, ≥27.5 kg/m2), defined as acquired cofactors (either one present = cofactors yes), on the prevalence of severe steatosis (>33% of hepatocytes involved). P = 0.001 and P = 0.38 for increasing prevalence of severe steatosis with PNPLA3 148M alleles in patients with and without cofactors, respectively.

Association of Liver Steatosis and Metabolic Risk Factors With Advanced Fibrosis

Independent predictors of advanced fibrosis at multivariate logistic regression analysis are presented in Table 4. Advanced fibrosis was associated with older age (OR, 4.17; CI, 2.21-8.13; for age >50 years), HBeAg positivity (OR, 2.53; CI, 1.16-5.72), but not with gender and viral load. Interestingly, advanced fibrosis was also independently associated with a positive history of any degree of alcohol consumption (OR, 2.09; CI, 1.02-4.32) and higher BMI (OR, 1.11; CI, 1.02-1.22; per g/m2), that is, two known risk factors for steatosis, whereas the association of advanced fibrosis with severe steatosis was not independent of these variables, although a nonsignificant trend was observed (OR, 2.56; CI, 0.98-7.60). Similarly, there was a trend for an independent association of NAS with advanced fibrosis, when this variable was introduced in the model in substitution of severe steatosis (OR, 1.15; CI, 0.98-1.35; P = 0.08).

Table 4. Clinical and Genetic Factors Independently Associateda With the Presence of Advanced Fibrosis (METAVIR Stage 3-4) in 235 Italian Patients With CHB
Clinical and Genetic FactorsMETAVIR Stage 3-4
OR (95% CI)P Value
  1. a

    At logistic regression analysis considering clinical risk factors and the I148M PNPLA3 polymorphism.

Age >50 years4.17 (2.21-8.13)<0.0001
Female gender0.89 (0.45-1.77)0.7400
BMI (kg/m2)1.11 (1.02-1.22)0.0150
Diabetes or IFG1.65 (0.54-5.51)0.3800
Alcohol intake2.09 (1.02-4.32)0.0440
Severe steatosis2.56 (0.98-7.60)0.0600
148M PNPLA3 allele1.08 (0.65-1.76)0.7600
HBeAg positive2.53 (1.16-5.72)0.0180
HBV DNA (log10 IU/mL)1.10 (0.90-1.33)0.3100

Discussion

This is the first study demonstrating an association between the 148M PNPLA3 allele and an increased risk of both steatosis of any degree and severe steatosis in CHB patients. The association with severe steatosis was particularly evident in patients with comorbidities, such as increased body mass and abnormal alcohol intake.

The PNPLA3 I148M polymorphism has recently been identified as a major determinant of liver fat accumulation and progressive steatohepatitis[25, 29, 41] in the presence of triggering factors for steatosis, such as increased visceral adipose tissue,[32, 46] high carbohydrate intake,[47] alcohol,[48, 49] and CHC.[40, 50-52]

In line with previous results,[53, 54] the presence of steatosis, which was observed in 62% of patients, was independently associated with older age, increased BMI, and hyperglycemia, but not with viral features, such as HBeAg status, and viral load, thus suggesting that metabolic alterations are the leading cause of steatosis in CHB, as in the general population and in CHC,[55] whereas differently from hepatitis C, the virus itself does not play a role.[54] The high prevalence of steatosis in the present series[54] can be explained by the high prevalence of metabolic risk factors and the inclusion criteria (e.g., allowance of excessive alcohol consumption).

The major finding of the present study is the I148M polymorphism representing a genetic determinant of steatosis susceptibility in CHB. Similarly to what was observed in CHC,[40, 50] the 148M allele was an independent predictor of steatosis of any degree, but it was even more strongly associated, together with increased BMI, with the presence of severe steatosis, increasing the risk by approximately 6-fold. Interestingly, the effect was particularly evident in the 35% of patients with acquired cofactors, such as a positive history of alcohol intake and/or severe overweight, whereas it was negligible in low-risk teetotalers with normal weight, which is consistent with the hypothesis that severe steatosis results from the interaction of different predisposing conditions, including the 148M PNPLA3 allele.[41] Recently, an interaction between the PNPLA3 I148M polymorphism and tea drinking in the pathogenesis of steatosis have been reported in an epidemiological study conducted in Asia.[56] Although a limitation of the present study is that tea and coffee drinking was not quantitatively assessed, tea drinking was not frequent in Italian patients, and both coffee and tea consumption were not associated with steatosis (not shown).

Of note, increased BMI and active alcohol consumption were also independently associated with advanced fibrosis, and a nonsignificant trend for an association between advanced disease and severe steatosis (or the NAS) was also observed, thus leaving open the possibility that altered hepatic lipid metabolism is a risk factor for fibrosis progression also in CHB,[15, 17] although prospective studies are required for confirmation.

As a result of the many confounders influencing disease history, the PNPLA3 I148M polymorphism was not associated with fibrosis severity, but, despite the relatively large number of well-characterized biopsied patients included, the power of the study was not sufficient to formally test the interaction between genetic and acquired risk factors in the pathogenesis of liver fibrosis.

In addition, whether the I148M PNPLA3 polymorphism influences clinical outcomes associated with steatosis, such as response to therapy[53, 57] and HCC,[58-61] should be further assessed. Indeed, in 267 treatment-naïve Asian patients with CHB under entecavir treatment, steatosis has recently been reported to represent an independent predictor of viral response, which, if confirmed by independent studies, would advise for a specific antiviral strategy in CHB patients with steatosis.[53]

Despite the limitations related to the cross-sectional design and the limited number of subjects considered with coexistent genetic and acquired risk factors for steatosis, strenghts of our study consist in the possibility to analyze one of the largest series of well-characterized biopsied CHB patients of Western countries with systematic assessment of liver steatosis and fibrosis as well as to evaluate, for the first time, the effect of the I148M PNPLA3 polymorphism on steatosis in CHB.

In conclusion, the PNPLA3 I148M polymorphism is an independent predictor of steatosis and, especially, of severe steatosis in patients with CHB. The study also suggests that steatosis is highly prevalent in Italian CHB patients with indications for liver biopsy and is related to genetic and metabolic, but not to viral, factors.

Ancillary

Advertisement