SEARCH

SEARCH BY CITATION

Keywords:

  • fibrosis progression;
  • hepatitis C;
  • VDR bAt-haplotype;
  • vitamin D;
  • vitamin D receptor polymorphisms

Abstract

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Background

Decreased vitamin D levels have been described in various forms of chronic liver disease and associated with advanced fibrosis. Whether this association is a cause or consequence of advanced fibrosis remains unclear to date.

Aims

To analyse combined effects of 25-OH vitamin D plasma levels and vitamin D receptor gene (VDR;NR1I1) polymorphisms on fibrosis progression rate in HCV patients.

Methods

251 HCV patients underwent VDR genotyping (bat-haplotype: BsmI rs1544410 C, ApaI rs7975232 A and TaqI rs731236 A). Plasma 25-OH vitamin D levels were quantified in a subgroup of 97 patients without advanced fibrosis. The VDR haplotype and genotypes as well as plasma 25-OH vitamin D levels were associated with fibrosis progression.

Results

The bAt[CCA]-haplotype was significantly associated with fibrosis progression >0.101 U/year (P = 0.007; OR = 2.02) and with cirrhosis (P = 0.022; OR = 1.84). Forty-five percent of bAt[CCA]-haplotype patients were rapid fibrosers, 21.1% were cirrhotic. Likewise, ApaI rs7975232 CC genotype was significantly associated with fibrosis progression and cirrhosis. Lower plasma 25-OH vitamin D levels were significantly associated with fibrosis progression >0.101 U/year in F0–2 patients (P = 0.013). Combined analysis of both variables revealed a highly significant additive effect on fibrosis progression with 45.5% rapid fibrosers for bAt[CCA]-haplotype and 25-OH vitamin D < 20 μg/L compared with only 9.1% for the most favourable combination (P = 0.006). In multivariate analysis, the bAt-haplotype was an independent risk factor for fibrosis progression (P = 0.001; OR = 2.83).

Conclusion

Low 25-OH vitamin D plasma levels and the unfavourable VDR bAt[CCA]-haplotype are associated with rapid fibrosis progression in chronic HCV patients. In combination, both variables exert significant additive effects on fibrosis progression.

Progressive liver fibrosis and development of cirrhosis are the leading causes of death in patients with chronic hepatitis C virus (HCV) infection [1]. However, there are pronounced individual differences with regard to the pace of fibrosis progression.

Vitamin D and its receptor have increasingly been recognized for a multitude of biological effects besides regulating calcium homeostasis and bone metabolism [2]. Decreased 25-OH vitamin D levels have been described in various forms of chronic liver disease and associated with advanced fibrosis [3, 4]. Irrespective of the underlying disease, serum concentrations of 1,25-(OH)2 vitamin D are decreased in patients with cirrhosis vs noncirrhotic patients [5, 6] and a gradual decline has been observed in cirrhotic patients according to increasing Child-Pugh class [5] and clinical decompensation [5]. Recently, significant inverse correlations of 25-OH vitamin D with stages of fibrosis and severity of necroinflammatory activity were observed in different populations with chronic hepatitis C [7, 8] and HCV/HIV-coinfection [9]. Also, patients with non-alcoholic fatty liver disease were found to present with reduced 25-OH vitamin D levels, which were closely associated with the histological severity of hepatic steatosis, necroinflammation and fibrosis [10]. Whether this observed vitamin D reduction in patients with advanced fibrosis or cirrhosis is the effect of impaired liver function or rather the cause of more advanced liver disease remains unclear to date.

The vitamin D receptor (VDR) serves as the physiological target to mediate vitamin D effects. Common genetic variations of the vitamin D receptor (NR1I1) gene such as the bat-haplotype consisting of rs1544410 (formerly BsmI), rs7975232 (ApaI) and rs731236 (TaqI) have been described [11]. VDR (NR1I1) gene variants modulate biological effects of vitamin D without influencing vitamin D plasma levels [11, 12]. In fact, genetic variations in the VDR (NR1I1) gene have been described as important modulator of multiple diseases including hepatic disorders such as primary biliary cirrhosis and autoimmune hepatitis [2, 13].

Hepatocytes express only low levels of VDR mRNA [14, 15], so that vitamin D effects on the liver are most probably not conferred by direct signalling in parenchymal liver cells. In contrast, non-parenchymal hepatic cells such as sinusoidal endothelial cells, Kupffer cells and hepatic stellate cells do express VDR mRNA and functionally active VDR protein [16]. More recent studies showed marked anti-inflammatory and antifibrotic effects of VDR-signalling in HSC. During inflammatory liver injury after endotoxin injection, the activation of VDR signalling by vitamin D attenuated liver damage in vivo[17]. From these data, we hypothesized that common VDR (NR1I1) polymorphisms may lead to increased hepatic fibrogenesis by abolishing protective vitamin D effects in chronic viral hepatitis C infection.

To examine the association between vitamin D and liver fibrosis in chronic hepatitis C patients in more detail, it appears attractive to study the combined effects of plasma 25-OH vitamin D levels and genetic alterations in vitamin D signalling on the individual fibrosis progression rate.

Patients and methods

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Patients and study protocol

All patients (n = 251) with chronic hepatitis C infection were recruited from the Swiss Hepatitis C Cohort Study (SCCS) (n = 224) [18] and another cohort from Erlangen University (n = 27). Overall, 223 patients with available peripheral blood mononuclear cell (PBMC) pellets and available liver biopsy results for histological fibrosis staging were included for the assessment of NR1I1 genotype and haplotype distribution. Haplotype reconstruction was performed using the Haploview software (http://www.broadinstitute.org/mpg/haploview).

All patients were 18 years and over and tested HCV antibody positive by immunoblot. Chronic HCV infection was confirmed by detectable HCV RNA. Exclusion criteria were coinfection with HBV or HIV (positive HBs antigen or anti-HIV antibody), alcohol consumption >40 g/day (as assessed by standardized questionnaire) and morbid obesity (BMI > 40 kg/m2). Alanine-aminotransferase (ALT) levels were available from standard autoanalyzer determinations. The METAVIR system was used to score fibrosis stage and necroinflammatory activity [19]. Fibrosis progression rates were estimated using stage-constant methods according to Bochud et al. [20] by division of disease duration by METAVIR F stage.

The study was approved by the local ethics committee (EK-695) following the guidelines of the Declaration of Helsinki. A second independent cohort of 105 hepatitis C negative patients with liver resection for focal liver lesions in otherwise healthy liver underwent genotyping of NR1I1 polymorphisms as a genetic control group as previously described [21, 22].

DNA Isolation and genotyping of NR1I1 polymorphisms

DNA was isolated from whole blood/PBMC using the QIAamp DNA Minikit (Qiagen, Hilden, Germany) Genotyping of NR1I1 polymorphisms was performed using standard TaqMan technology. Fluorogenic 5’-nuclease (TaqMan) assays on a ABIPRISM 7700 sequence detection system (Applied Biosystems, Darmstadt, Germany) to determine human VDR SNPs (rs731236, rs1544410, rs7975232) as described [23-26]. All genotyping data were analysed with the SDS-Software 2.3; Applied Biosystems, Darmstadt, Germany.

Quantification of hepatic mRNA expression levels

In 35 HCV patients without cirrhosis liver biopsy and serum samples were further analysed for mRNA expression levels of metalloproteinase-1 (TIMP-1) and matrixmetalloproteinase-9 (MMP-9) and VDR. Total RNA was extracted according to the manufacturer's recommendations and cDNA was obtained by reverse transcription of 1μg of total RNA. Relative transcript levels were quantified on a ABIPRISM 7700 using TaqMan probes and primer sets obtained from Applied Biosystems (Hs00957MM_m1 MMP-9, Hs00171558_m1 TIMP-1 and Hs01045840_m1 VDR). For normalization, 18s mRNA was amplified in a parallel reaction.

Quantification of 25-hydroxyvitamin D serum levels

In a subgroup of 97 patients with early stage fibrosis (F0-2) and adequately stored serum samples, 25-hydroxy-vitamin D levels were measured by radioimmunosorbent assay (DiaSorin, Saluggia, Italy) according to the manufacturer's protocol.

Statistical analysis

Statistical analysis was performed using SPSS 17.0 (SPSS Inc., Chicago, Illinois, USA). Statistical associations were calculated for liver function tests (ALT), Metavir fibrosis stage (F0 no fibrosis – F3 bridging fibrosis vs. F4 cirrhosis) and in a subset of patients hepatic gene expression (MMP-9, TIMP-1 and VDR). For statistical analyses to compare categorical measures between the different groups, Fisher's exact test was used. Effects of age, gender, body mass index (BMI), ALT and the CCA NR1I1 haplotype on fibrosis progression rate (<0.101 vs. >0.101 U/year) and fibrosis stage (F0–3 vs. F4) were analysed using univariate and multiple logistic regression. Odds ratios with 95% confidence intervals (CI) were computed using logistic regression. A correction for multiple comparison according to Bonferroni has been performed where appropriate. Genotype distribution was checked for Hardy-Weinberg equilibrium [27]. For the association of serum markers vs VDR genotypes, the Mann-Whitney or Kruskal–Wallis test was used as appropriate. For the correlation of hepatic gene expression vs VDR (NR1I1) genotypes, the Spearman rank correlation was used. The patient number within the two different fibrosis progression groups was estimated to yield potential significance (power 0.80 for CCA haplotype, α = 0.016).

Results

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Patient characteristics

A total of 251 patients with chronic hepatitis C infection were included for VDR (NR 1I1) genotyping. Characteristics of the cohort are shown in Table 1. For 223 patients of the HCV cohort, liver biopsy results including Metavir fibrosis were available. In 134 patients with a known HCV risk factor, the duration of infection and the fibrosis progression rate have been calculated. The frequency of different reported HCV risk factors was similar to Bochud et al. [20].

Table 1. Patient characteristics
 HCV Cohort (n = 251)
Gender (male/female)163/88
Median Age (range)52 (22–78)
Ethnicity (%)
 Caucasian93
 Black2
 Asian3
 Others2
HCV Genotype (%)
 Genotype 155.7
 Genotype 2 & 344.3
Liver histology
Fibrosis score/F n (%)223 (100)
 016 (7.2)
 160 (26.9)
 278 (35)
 331 (13.9)
 438 (17)
Activity score/A n (%)203 (100)
 07 (3.4)
 171 (35)
 290 (44.3)
 335 (17.3)
Fibrosis progression rate (n=134)0.111 ± 0.11
HCV reported risks (%)
 Drug use43.9
 Invasive procedures/needle stick31.7
 Transfusion23.6
 HCV pos. partner0.8
ALT (U/L) mean value (n = 133)126.4 ± 94.49
BMI (kg/m²) (n = 195)24.61 ± 4.02
25 OH-Vitd (ng/L) mean value F0–2 (n = 97)
 25.8% below 12ng/ml22.35 ± 12.47
 48.5% below 20 ng/ml
 72.2% below 30 ng/ml

VDR (NR 1I1) genotype and haplotype distribution in association to fibrosis progression rate

Firstly, HCV patients were compared with a control cohort of 105 HCV negative patients undergoing liver resection with regard to allelic frequency without significant difference in the distribution of geno- and haplotypes (supplement Table S1). Genotype distribution was found to be in Hardy-Weinberg equilibrium for HCV patients and controls.

The VDR (NR 1I1) genotypes were associated to a retrospectively calculated fibrosis progression rate whenever respective data were available. For this purpose, HCV patients were divided into rapid (>0.101 U/year) and slow fibrosers (<0.101 U/year). As shown in Fig. 1A, the bAt [CCA]-haplotype consisting of rs1544410 (formerly BsmI) C, rs7975232 (ApaI) C and rs731236 (TaqI) A alleles showed a significant association with fibrosis progression rate. 45% of the HCV-infected patients with the bAt [CCA]-haplotype were rapid fibrosers as compared with 28.8% in the other common haplotypes CAA (bat) and TAG (BaT) resulting in an odds ratio of 2.02 (95% CI 1.21–3.37, P = 0.007). In line with the haplotype data, also the ApaI rs7975232 CC genotype was associated with rapid fibrosis progression (CC vs. CA vs. AA: P = 0.033 and for CC vs. CA & AA, OR 2.32, CI 1.05–5.10, P = 0.042) (Fig. 1A). For the TaqI rs731236 and BsmI rs1544410 polymorphisms, no significant associations with fibrosis progression were detected. Multivariate analysis confirmed the bAt [CCA]-haplotype (OR 2.83, 95% CI 1.81–5.32, P = 0.001) as an independent risk factor for fibrosis progression (Table 2).

image

Figure 1. Fibrosis progression and presence of cirrhosis according to VDR genotype. (A) VDR bAt [CCA]-haplotype and ApaI rs7975232 CC were associated with fibrosis progression rate in chronic hepatitis C patients using the Fisher's exact test. (B) For the presence of cirrhosis, the bAt [CCA]-haplotype and ApaI rs7975232 CC were similarly associated.

Download figure to PowerPoint

Table 2. Univariate and multivariate analysis
 ParameterOR (95% CI)P
Univariate analysis fibrosisprogression >0.101Age (>45 vs. <45 years)0.69 (0.43–1.17)0.17
Gender (male vs. female)1.43 (0.82–2.49)0.19
BMI0.94 (0.88–1.01)0.13
ALT1.002 (0.999–1.005)0.31
Haplotype (CCA vs. TAG & CAA)2.02 (1.21–3.37)0.007
Multivariate analysis fibrosisprogression >0.101Age (>45 vs. <45 years)0.69 (0.35–1.36)0.29
Gender (male vs. female)1.72 (0.83–3.57)0.14
BMI0.96 (0.89–1.05)0.46
ALT1.001 (0.998–1.005)0.45
Haplotype (CCA vs. TAG & CAA)2.83 (1.51–5.32)0.001
Univariate analysis cirrhosis F4Age (>45 vs. <45 years)3.09 (1.53–6.22)0.002
Gender (male vs. female)0.54 (0.32–0.89)0.017
BMI1.02 (0.95–1.09)0.47
ALT1.005 (1.001–1.009)0.007
Haplotype (CCA vs. TAG & CAA)1.84 (1.098–3.08)0.022
Multivariate analysis cirrhosis F4Age (>45 vs. <45 years)2.06 (0.75–5.66)0.157
Gender (male vs. female)0.38 (0.15–0.94)0.038
BMI1.02 (0.92–1.13)0.65
ALT1.005 (1.001–1.009)0.008
Haplotype (CCA vs. TAG & CAA)2.54 (1.07–6.01)0.034

In a subgroup analysis, the bAt [CCA]-haplotype was significantly associated with accelerate fibrosis progression rate in HCV-genotype 2 and 3 patients (P = 0.046), whereas in genotype 1 patients, no significant association between any VDR haplotype and fibrosis progression has been observed (Supplement Fig. 1A).

VDR (NR 1I1) genotype and haplotype distribution in association to fibrosis stage

In addition to the analysis of the fibrosis progression rate, VDR (NR 1I1) genotypes were associated with fibrosis stage. The distribution of individual VDR (NR 1I1) genotypes across the different stages of fibrosis is shown in supplement Table S2. Patients were subsequently categorized into ‘cirrhosis’/F4 vs. ‘no cirrhosis’ defined as F0–3.

As shown in Fig. 1B, the bAt [CCA]-haplotype was associated with the presence of cirrhosis. 21.1% of the HCV-infected patients with the bAt [CCA]-haplotype were cirrhotic as compared with 12.7% in the other common haplotypes CAA (bat) and TAG (BaT) resulting in an odds ratio of 1.84 (95% CI 1.09–3.08; P = 0.022). 28.3% of the patients carrying the corresponding ApaI rs7975232 CC genotype were cirrhotic and a significant association for the cirrhotic patients between CC vs. CA vs. AA (P = 0.025) and between CC vs. CA/AA (P = 0 0.009, OR 2.67, 95% CI 1.29–5.51) could be identified. The TaqI rs731236 and BsmI rs1544410 polymorphisms were not significantly associated with the presence of cirrhosis.

25-OH Vitamin D serum levels in association to progression rate and combined effect together with the VDR (NR 1I1) haplotype

25-OH vitamin D serum levels were measured in serum samples of a subset of 97 patients with early stage fibrosis (F0–F2). Patients with higher stages of fibrosis >F2 were excluded from this subgroup analysis because of known interference of advanced liver fibrosis and cirrhosis per se with 25-hydroxylation and progressively decreased 25-OH vitamin D levels as shown by Petta et al. [7]. This effect could also be observed in early stages of fibrosis in the present study and led to exclude patients with higher fibrosis stages. As shown in Fig. 2A, 25-OH vitamin D levels were significantly lower in patients with Metavir fibrosis stage 2 compared with stage 0 or 1 (median 15.5 vs. 22.2 μg/L; P = 0.005). 25-OH vitamin D levels were significantly decreased in patients with accelerated fibrosis progression rate (>0.101 U/year; P = 0.013). The median of 25-OH vitamin D was 12.3 μg/L among patients with a fibrosis progression rate >0.101 U/year compared with a median of 21.1 μg/L in patients with slower fibrosis progression (Fig. 2B). In a subgroup analysis, the 25-OH vitamin D levels were significantly correlated with fibrosis progression rate in HCV-genotype 1 patients (P = 0.019). In the genotype 2 and 3 population, no significant correlation was detected (Supplement Fig. 1B).

image

Figure 2. Combined effects of 25-OH vitamin D and VDR haplotype on fibrosis progression rate. 25-OH vitamin D serum levels were measured in a subset of 97 patients with early stage fibrosis (F0–F2). (A) A significant inverse correlation between early stages of fibrosis and the 25-OH vitamin D serum levels was observed using the Spearman Rho correlation. (B) Similarly, 25-OH vitamin D serum levels were also correlated with fibrosis progression rate. (C) In patients with a higher progression rate, 25-OH vitamin D serum levels were significantly lower than patients with less advanced fibrosis progression. A Combined analysis of VDR (NR1I1) haplotype together with 25-OH vitamin D plasma levels using the Fisher's exact test, showed a highly significant additive effect of both variables on fibrosis progression rate with 5-fold increased proportion of rapid fibrosers for bAt [CCA]-haplotype and 25-OH vitamin D < 20 μg/L compared with only 9.1% for the most favourable combination.

Download figure to PowerPoint

A combined analysis of VDR (NR1I1) haplotype together with 25-OH vitamin D plasma levels revealed a highly significant additive effect of both variables on fibrosis progression rate. Depending on VDR (NR1I1) haplotype and 25-OH vitamin D level, a gradual decrease over the different combinations was detectable with 45.5% rapid fibrosers for bAt (CCA)-haplotype and 25-OH vitamin D < 20 μg/L, 35.0% for other haplotypes and 25-OH vitamin D < 20 μg/L and 21.9% for bAt (CCA)-haplotype with normal 25-OH vitamin D compared with only 9.1% for the most favourable combination of both (P = 0.006) (Fig. 2C). No significant correlation between any VDR (NR1I1) polymorphism and 25-OH vitamin D levels could be observed (Supplement Fig. 2).

Liver function tests according to VDR (NR1I1) genotype and haplotype

Alanine-aminotransferase levels as serum markers reflecting liver cell damage and inflammation were accordingly analysed (n = 133 with available serum parameters) and were associated with the VDR (NR1I1) haplotype (bAt-haplotype P = 0.021 vs. bat and BaT) (Fig. 3). In addition, a significant association of the corresponding ApaI rs7975232 CC genotype (P = 0.034 vs. CA vs. AA) was observed. No significant association was detectable for BsmI rs1544410 CC (P = 0.08 vs. CT vs. TT) and TaqI rs731236 AA (P = 0.07 vs. AG vs. GG).

image

Figure 3. Alanine aminotransferase levels according to VDR genotype and haplotype. To assess liver cell damage and inflammation by clinical chemistry ALT serum levels were analysed (n = 132) and associated to VDR polymorphisms using the Mann–Whitney and the Kruskal–Wallis test. Elevated ALT levels were significantly associated with the bAt [CCA]-haplotype and ApaI rs7975232 CC.

Download figure to PowerPoint

Hepatic gene expression levels according to NR1I1 genotype in non-cirrhotic patients

The mRNA levels of fibrosis marker genes TIMP-1 and MMP-9 as well as VDR expression were quantified in liver tissue of n = 35 HCV-infected patients. To obtain reliable results, expression of these genes in the context of VDR (NR1I1) polymorphisms was quantified only in patients without development of advanced fibrosis or even cirrhosis.

The MMP-9 expression was increased in trend for the bAt [CCA]-haplotype compared with the other common haplotypes CAA (bat) and TAG (BaT) (P = 0.058), but significantly elevated for the ApaI rs7975232 CC genotype (median: 263.35; range: 9.3–1244.7) vs. CA (86.9; 14.8–566.5) and AA (63.0; 20.9–210.5; P = 0.038) (Fig. 4). In contrast, no changes in TIMP-1 expression could be observed for any VDR (NR1I1) genotype and haplotype (Fig. 4). VDR gene expression levels were not affected by VDR (NR1I1) genotype and haplotype (data not shown).

image

Figure 4. Hepatic expression of fibrosis-associated genes. To determine functional effects of NR1I1 polymorphisms mRNA levels of TIMP-1 and MMP-9 as markers of fibrogenesis and VDR expression were quantified in liver tissue (n = 35; Metavir stages F0-3). The bAt [CCA]-haplotype showed elevated MMP-9 expression compared with the other common haplotypes with a clear statistical trend in the Spearman Rho correlation. MMP-9 mRNA expression showed a significant elevation in the Apa CC genotype. VDR expression was not affected by respective NR1I1 genotype and the common haplotypes.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Vitamin D deficiency is a common phenomenon in chronic liver disease, particularly in advanced fibrosis and cirrhosis. Whether this association reflects the cause of accelerated fibrosis progression or the consequence of impaired liver function in advanced disease is still unclear.

To address this main purpose of the present study, it is important to discriminate different requirements for the genetic study and the analysis of 25-OH vitamin D levels. For the association of 25-OH vitamin D levels with the presence of cirrhosis and fibrosis progression rate, it appeared mandatory to exclude patients with advanced fibrosis (F3–F4) because of their significantly decreased 25-OH vitamin D levels compared with patients with earlier fibrosis stages (F1–F2) [7]. In line with these findings by Petta et al. [7], in the present study, this effect could even be observed in fibrosis stage F2 compared with stages F0–F1. To address the basic question on cause and consequence of fibrosis progression without vitamin D bias in advanced disease, it appeared mandatory to exclude those patients with advanced fibrosis (F3–F4) but to keep a statistically sufficient number of patients (n = 97 including F2) in the analysis. This decision to include F2 patients was based on statistical power calculations and the need for at least three different stages of fibrosis for a meaningful calculation of disease progression. Requirements for the genetic analysis were less strict as the VDR (NR1I1) genotype is obviously present as a putative modifier before the onset of fibrosis development in all patients. For this reason, the entire available cohort was used for the genetic associations with fibrosis stage (n = 223) and fibrosis progression rate (n = 134).

The present study now demonstrates a significant contribution of both plasma 25-OH vitamin D levels and a genetic predisposition within the VDR (NR1I1) gene to the individual fibrosis progression rate in patients with chronic hepatitis C infection. As a novel finding of this study, the common bAt [CCA]-haplotype and the contributing ApaI rs7975232 CC genotype are both significantly associated with a fibrosis progression rate >0.101 U/year and with the presence of cirrhosis. Lower plasma 25-OH vitamin D levels were significantly associated with rapid fibrosis progression in a subgroup of patients without advanced fibrosis. Combined analysis of VDR (NR1I1) haplotype together with 25-OH vitamin D plasma levels revealed a highly significant additive effect of both variables on fibrosis progression rate with 5-fold increased proportion of rapid fibrosers for bAt (CCA)-haplotype and 25-OH vitamin D < 20 μg/L compared with only 9% for the most favourable combination.

Hepatic stellate cells as the pivotal cell type of hepatic fibrogenesis have been identified to express functional VDR and vitamin D attenuates its activation during inflammatory liver injury in vivo[16, 17]. In mice, administration of vitamin D also resulted in a reduced expression of chemokines, cytokines and fibrosis-related matrix genes [17]. Following the hypothesis of increased cell injury, inflammation and fibrogenesis in certain genetic VDR (NR1I1) variants, analysis of serum cell injury markers and hepatic mRNA expression of fibrosis-associated genes has been performed in our study. In line with potential anti-inflammatory effects of vitamin D signalling, elevated serum ALT levels, as marker of cell injury were significantly associated with the bAt [CCA]-haplotype and the contributing ApaI rs7975232 CC genotype. In support of an activation of fibrogenesis in the carriers of these VDR (NR1I1) variants, increased hepatic MMP-9 mRNA expression was equally correlated. In fact, humans with chronic HCV infection demonstrate elevated MMP-9 mRNA levels [28] and MMP-9 activity in liver tissue [29]. Interestingly, vitamin D insufficiency has also been linked to increased circulating MMP-9 levels in healthy adults [30] and the tumour necrosis factor-α-mediated MMP-9 up-regulation in HaCaT keratinocyte cells can be attenuated by vitamin D [31]. Hepatic MMP-9 expression, which is generally observed during the chronic stage of fibrogenesis and scar formation, may not only be associated with the degree of fibrosis but also linked to above discussed tissue inflammation [32].

The current study clearly demonstrates that enhanced progression of fibrosis is significantly and independently associated with both genetic VDR (NR1I1) variants and 25-OH vitamin D plasma levels <20 μg/L. This suggests vitamin D substitution as a preventive measure for patients with liver fibrosis. In fact, oral vitamin D replacement is effective to correct reduced 25-OH vitamin D levels which decrease to baseline after termination [33]. Of note, 25-OH vitamin D plasma levels >20 μg/L are associated with a decreased frequency of accelerated fibrosis progression in all VDR (NR1I1) haplotype groups and warrant substitution, irrespective of the genotype. Currently, vitamin D supplementation is recommended in cirrhosis patients with bone disease [34]. From our data, it can be assumed that vitamin D substitution may be beneficial for all patients with chronic liver disease to modulate fibrosis progression, irrespective of the presence of bone disease. To further support this concept and quantify the potential benefit of vitamin D substitution, future interventional studies need to be conducted.

Two methodological aspects are relevant for this study. Fibrosis progression was estimated using the stage-constant fibrosis progression rate based on the so far unverified assumption that fibrosis progression is constant over time [20]. Another limitation resulted from the fact that the estimated duration of infection was based on the first event at risk (such as the first year of intravenous drug use), which may not be completely accurate and may have led to an overestimated, rather than an underestimated duration. Furthermore, categorization of fibrosis stages into cirrhosis vs no cirrhosis was carried out in this study. Several studies in the past demonstrated the variability of histological scoring based on specimen size, location and quality of the liver biopsy and on the experience of the pathologist [35-37]. This variability is mostly limited to single stages and has been successfully overcome by categorization of all patients in only two groups, namely cirrhosis vs no cirrhosis, to minimize the effect of this sampling error [22].

In summary, the present study demonstrates strong, significant and independent associations of low 25-OH vitamin D plasma levels and the unfavourable VDR bAt [CCA]-haplotype with accelerated fibrosis progression in chronic HCV patients. Possible beneficial antifibrotic effects of vitamin D substitution in patients with chronic liver disease in general and HCV patients in particular remain to be investigated in controlled and randomized clinical studies in the future.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The authors thank Sarah Mertens and Claudia Gottier for their excellent technical assistance and Dr Franziska Schöni-Affolter (University of Lausanne, CHUV, SCCS Cohort) for her epidemiological advice and logistic support.

Conflict of interest and funding source: This study was funded in part by Hoffmann-La-Roche Pharmaceuticals (to A.G. and B.M.). Hoffmann-La-Roche had no role in study design, analysis and interpretation of the data.

References

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information
FilenameFormatSizeDescription
liv2674-sup-0001-FigS1.pptxapplication/mspowerpoint104KFig. S1. Fibrosis progression and 25-OH vitamin D levels according to VDR genotype and haplotype in subgroups with HCV genotype 1 vs. non-1.
liv2674-sup-0002-FigS2.pptxapplication/mspowerpoint57KFig. S2. 25-OH vitamin D levels according to VDR (NR1I1) genotype and haplotype.
liv2674-sup-0003-TableS1.docxWord document16KTable S1. Comparison of distribution of geno- and haplotypes between HCV infected patients and controls.
liv2674-sup-0004-TableS2.docxWord document17KTable S2. Distribution of individual VDR (NR 1I1) genotypes across the different stages of fibrosis.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.