Effect of PNPLA3, TM6SF2 and MBOAT7 Genetic Variants on Non-alcoholic Fatty Liver Disease Therapeutic Outcome.

The aim of this study was to evaluate if the PNPLA3 rs738409, TM6SF2 rs58542926 and MBOAT7 rs641738 could influence the response of NAFLD patients regarding metabolic parameters, liver damage and hepatic fat accumulation ones, to a treatment with 303 mg of silybin-phospholipids complex, 10 mg of vitamin D, and 15 mg of vitamin E twice a day for six months.

The investigators performed a baseline comparison of Weight, waist-to-height ratio (WHtR), blood pressure measurement, body mass index (BMI), Blood glucose and insulin, the homeostatic model for insulin resistance assessment (HOMA-IR), aspartate and alanine aminotransferase (AST, ALT), gamma-glutamyl transferase (GGT), blood count, C reactive protein (CRP), the Thiobarbituric Acid Reactive Substance (TBARS), liver stiffness and controlled attenuation parameter (CAP) among the three study groups: NAFLD wild type control group (n. 30), NAFLD treated wild type group (n. 30), NAFLD treated mutated group (n. 32). The block randomization method was used to randomize the 60 not mutated patients in the NAFLD wild-type control group and NAFLD treated wild type one, using the online randomization software http://www.graphpad.com/quickcalcs/index.cfm.

The wild-type control group was composed by NAFLD patients without PNPLA3, TM6SF2 and MBOAT7 mutations that did not receive any type of treatment during the study period.

The patients inserted in the NAFLD treated wild type and NAFLD treated mutated groups were undergone to an oral administration of 303mg of silybin-phospholipid complex, 10mg of vitamin D, and 15mg of vitamin E, twice a day for six months. None of the enrolled patients dropped-out the study.

At the end of treatment, the assessment of WHtR, blood pressure measurement, BMI, Blood glucose and insulin, HOMA-IR, AST, ALT, GGT, blood count, CRP, TBARS, liver stiffness and CAP was re-performed. During the experimental observation, patients were on free diet on the basis of dietary habits before the enrollment and any type of physical exercise was recommended during the study period. Food intake was evaluated both at baseline and end of treatment using a computerized software. The investigators recorded, with a diet diary, the food intake of a complete week, including working days and the weekend. On the basis of the quantities and qualities of food consumed, the soft elaborates the daily energy intake and the percentage/caloric amount of macronutrients. For the physical exercise assessment, the investigators submitted a specific questionnaire at baseline and end of treatment with some simple questions: Are the participant doing or have the participant ever done (in the last two years) sport in a continuative and regular way? Have the participant changed his/her daily physical activity in the last six months? 1: no, 2: yes; if yes, has the physical activity enhanced or worsened? Alcohol consumption was assessed at the beginning and at the end of the treatment. One hundred two patients with histological diagnosis of NAFLD followed by the Hepatogastroenterology Division of the University of Campania "Luigi Vanvitelli, between January and October 2017 were screened, after signing an informed consent, for the PNPLA3 rs738409, TM6SF2 rs58542926 and MBOAT7 rs641738 genetic variants and thirty-two met the inclusion criteria for the study and showed at least one among PNPLA3 I148I/M, I148M/M, TM6SF2 167E/K, 167K/K and MBOAT7 TMC4C/T or TMC4T/T genetic variants, were enrolled together with sixty patients without the mutations. Ten patients were excluded from the enrollment due to the coexistence of several comorbidities and/or advanced stages liver disease such as cirrhosis and/or hepatocellular carcinoma (HCC).

The definition of the presence/absence of NAFLD and the staging were assessed by performing a liver biopsy, serological tests and collecting clinical data. Medical history, alcohol consumption (AUDIT-C), medications, drug abuse, smoking habits, were also investigated. Blood pressure, weight, height were directly measured and WHtR was calculated. BMI was also calculated by dividing the weight (kg) by the square of height (m). The patients were undergone after 12 hours fast to peripheral venous blood samples collection in order to evaluate some biochemical parameters and to perform the genetic analysis.

Insulin, GGT, CRP, levels were measured enzymatically using commercially available kits, AST, ALT and glucose using colorimetric assay kit (Amplite 13801/13803 and Thermo Fisher Scientific EIAGLUC). HOMA-IR was also calculated using the formula: fasting insulin (μU/mL) × plasma glucose (mmol/L)/22.5.

FibroScan® transient elastography (TE) was performed using the FibroScan® version 502 (Echosens, Paris, France) with standard probes (M and XL probes. The extra large (XL) probe was used when the distance from the skin to the liver capsule, assessed by ultrasonography, exceeded 2.5 cm and/or when BMI was >30. FibroScan® was performed by an expert physician obtaining ten acceptable measurements, defined as a successful liver stiffness (LS) measurement, with the maximum number of attempts set at 20. The criteria proposed by Boursier et al. were used to consider the measurement "very reliable" (IQR/M ≤ 0:1), "reliable" (0:1 < IQR/M ≤ 0:3 or IQR/M > 0:3 with LS median < 7:1 kPa), or "poorly reliable" (IQR/M > 0:3 with LS median ≥ 7:1 kPa). On the basis of CAP these scores, the investigators classified the enrolled patients in S0, no steatosis (0%-10% fat; 0-237 dB/m); S1, mild steatosis (11%-33% fat; 238-259 dB/m); S2, moderate steatosis (34%-66% fat; 260-292 dB/m); and S3, severe steatosis (>67% fat; ≥293 dB/m) in accordance with calculation of the attenuation of ultrasonic signals used for TE.

TBARS assay was performed using 10 μl of serum. The cromogen TBARS was quantified using a spectrophotometer at a wavelength of 532nm with 1,1,3,3-tetramethoxyprophane as a standard. The amount of TBARS was expressed as nmol/μg of protein. Presented data are the mean ± standard deviation, resulting from three independent experiments.

The genomic DNA extraction from peripheral blood samples was performed by using the extraction kit PureLink Genomic DNA Kit (Invitrogen by Life Technologies, USA). The DNA amount of each sample was assessed by spectrophotometer (NanoDrop Thermo Fisher Scientific) using a wave length of 260 nm. The presence of sample contamination was assessed by using a 280 nm wave length evaluation and all the sample showed good degree of purity because of the 260/280 ratio was between 1.8 and 2. The DNA extracted was then stored in -20°C freezer until the analysis of the polymorphisms.

The single nucleotide polymorphisms (SNPs) analysis, using the access code to the data bank TM6SF2 rs58542926 (SNP 1), MBOAT7 rs641738 (SNP2), PNPLA3 rs738409 (SNP3), was performed using the DNA genotyping RealTime PCR with TaqMan (Applied Biosystem) c_89463510_10 for SNP 1, c_8716820_10 for SNP 2, c_7241_10 for SNP 3 probes. All the evaluations were done in three phases: DNA PCR amplification, allelic identification and end point analysis with melting curve.

The first phase was done by using the AmpliTaq Gold® DNA polymerase contained in the TaqMan Universal PCR Master Mix amplifying the target sequence by specific primers contained in the SNP Genotyping Assay 40X together with TaqMan® MGB probes: one marked with fluorochrome VIC® that recognize the sequence of the allele 1 and another probe marked with fluorochrome FAM™ that recognize the sequence of the allele 2. For each experiment a 48 well plate was used, in which, a part of the biologic samples of unknown genotype for the SNP analysis, three different negative controls were analyzed in order to avoid possible errors due to the contamination of samples. Each experiment was done in triplicate. The amplifications were performed using the StepOne™ Real Time PCR System (Applied Biosystems). The genotyping assessment is based on the allelic discrimination thanks to a different fluorescence of the specific gene primer, using two MGB TaqMan® probes, marked at 5' with a different fluorochrome: SNP1 in reverse G in VIC® and A in FAM™; SNIP2 in forward C in VIC® and T in FAM™; SNIP3 in forward C in VIC® e G in FAM™. For the results analysis the software StepOne™2.0 was used. After the differentiation of the fluorescence made from the probes to the background in each well, it measures the normalized signal intensities (Rn) projecting the results in allelic discrimination plot. The software gives to the samples a specific genotype based on the fluorescence signal position: horizontal axis (allele one), vertical axis (allele 2) diagonal axis (both allele one and two). The prevalence of a specific fluorescence on the other one identified the homozygosis genotype, on the contrary the presence of both the heterozygosis.

Statistical Analysis The number of patients (30 in the NAFLD wild-type control group, 30 in the NAFLD treated wild type and 32 in NAFLD treated mutated ones) was calculated using the power and sample size calculation function of STATA-14® for mac-OS, on the basis of an expected difference among the study groups in the response of the HOMA-IR to the therapy, assuming a double amount of patient responder to the therapy in wild type group in comparison to the mutated one. Specifically, the investigators considered the patients responder if at least one of the following criteria was addressed: normalization of the HOMA-IR (<2.5) at the end of treatment starting from baseline values greater than 2.5; reduction of the HOMA-IR ≥ 2 points at the end of treatment in comparison to baseline. On the basis of this difference, the investigators estimated 29 patients per arm as the correct sample size of subjects to be investigate maintaining an 0.01 alpha error and a 90% statistical power in a two-sided test with a 95% Confidence Interval. A Kolgoromov-Smirnov for normality was performed to evaluate if parametric or non-parametric analysis should be applied. Wilcoxon signed ranks test, Mann-Whitney U test and t-test for dependent or independent groups were performed in order to compare continuous variables. The Kruskal-Wallis test or ANOVA test with post-hoc Bonferroni analysis, in case of non-normal or normal distribution respectively, were performed to compare the continuous variables among three groups. Pearson's or Kendall Tau-b correlations as well as linear regression were applied to test the associations among variables. Multiple logistic regression analysis was performed to assess the relationship between the genotype of patients (NAFLD treated wild type vs mutated patients) and the therapeutic outcome on: insulin, HOMA-IR, ALT, CRP and TBARS. The investigators identified in the following values the specific cut-offs to consider the parameter improved: insulin normalization (< 24 micro-IU/ml), normalization of the HOMA-IR (<2.5) at the end of treatment starting from baseline values greater than 2.5 and/or reduction of the HOMA-IR ≥ 2 points at the end of treatment in comparison to baseline, ALT normalization (<45 IU/L), CRP normalization (<0.6 mg/dL) or reduction of at least 1 mg/dL, TBARS reduction of at least 10 nmol/μg. The abovementioned parameters were chosen in relation to the main therapeutic effect of silybin in this context, considering NAFLD as a systemic disease. The relative risk (RR) of a useful therapeutic outcome considering the genotype of the patients was calculated considering the confounding variables (age, sex, comorbidities, medications, liver stiffness and CAP). Statistical significance was defined as p<0.05 in a two-tailed test with a 95% Confidence Interval.

Statistical analyses were performed using Statistical Program for Social Sciences (SPSS®) vs.18.0.