Genes and Hepatitis C
Susceptibility, Fibrosis Progression and Response to Treatment
Posted: 04/28/2011; Liver International. 2011;31(4):443-460. © 2011 Blackwell Publishing
Abstract
Hepatitis C virus contact and infection show three different phenotypes: spontaneous viral clearance (SVC), chronic hepatitis C (CHC) and sustained virological response (SVR) following antiviral treatment. Many factors, including genetics, influence the evolution of these three phenotypes. We performed a literature search (PubMed) up to 31 January 2010 without language restriction to identify relevant studies on genes and hepatitis C. Additional studies were sought by reviewing the reference lists of the identified articles. Meta-analysis (using Meta-disk 1.4) was conducted to evaluate the association of single nucleotide polymorphism (SNP) in the IL28B region and SVR. The candidate gene approach showed strong relationships between human leucocyte antigen class II (DQB1*0301 and DRB1*1101) and SVC. A cirrhosis risk score involving 7 SNPs has been validated recently. The set of odds ratios of studies demonstrated an association between SNP (rs12987960/rs8099917) in the IL28B and SVR in CHC treated with peginterferon plus ribavirin (OR: 4.6; 95% CI: 2.9–7.3). The overall distribution of protective allele correlated with ethnic differences in SVR (Asians, Europeans, Hispanic and Afro-Americans) together with SVC, but not with fibrosis stage or viral load. These polymorphisms did not influence SVR in very-easy-to-treat patients such as genotype 2/3, rapid virological responders or patients with acute hepatitis C. While the genetic fingerprint for fibrosis progression remains elusive, IL28b polymorphism predicts SVC and SVR. However, nearly half of patients achieving SVR did not show favourable genotype. Further genetic signals are warranted to complete the puzzle of factors influencing hepatitis C.Introduction
Over the past 15 years, several studies have addressed the role of genetic factors in spontaneous clearance, fibrosis progression and response to combined antiviral therapy. Documented evidence indicates that interindividual genome variability contributes considerably to the observed differences in natural resistance or susceptibility to specific micro-organism, to the observed phenotype once infection is established, or to the therapeutic response when the infectious disease is pharmacologically treated. Recently, genome-wide association studies (GWAS) including a large number of single nucleotide polymorphisms (SNP) confirmed the influence of some polymorphisms in the interleukin 28B (IL28B) region on the possibility of achieving viral clearance, either spontaneously or after pegylated interferon alpha+ribavirin (Peg-IFN/RBV) treatment. This genetic marker is a stronger predictor of sustained response than other well-documented factors such as viral genotype, viral load or fibrosis. All these data supported the emerging interest of hepatologists in the influence of genetic factors in hepatitis C virus (HCV). In this review, we analyse the impact of genes on susceptibility, fibrosis progression and sustained virological response (SVR).
Methods
Relevant studies were identified by searching Medline (PubMed) up to 31 January 2010. We searched the literature without language restriction using a combination of the following terms: ('Hepatitis C'[Mesh] OR 'Hepacivirus'[Mesh] OR 'Hepatitis C, Chronic'[Mesh]) AND ('Databases, Genetic'[Mesh] OR 'Genetic Testing [Mesh] OR 'Genetic Association Studies'[Mesh] OR 'Genetic Loci'[Mesh]). To guarantee no loss of specific articles relating genetic factors and susceptibility to HCV infection, or hepatic fibrosis progression, or SVR following IFN treatment, a search combining the terms 'Hepatitis C'[Mesh]' AND 'Genetic Predisposition to Disease'[Mesh] OR 'Liver Cirrhosis'[Mesh] OR 'sustained virological response' was also performed. Additional studies were sought by reviewing the reference lists of the identified articles. Two co-authors separately examined the titles and abstracts. All original human studies concerning the topics of this review were selected for further full-article analysis, provided phenotypes and genotypes were correctly defined. Because the field of hepatitis C genetics remains a topic on which few studies have used prospective designs or had included large sample sizes, it was our intent to include as much information as possible despite some of these data requiring confirmation in better-designed studies. We conducted a meta-analysis that included all published studies reporting odds ratio (OR) or risk ratio that had been calculated by comparing the prevalence of genotype CC (rs12968760) in patients with SVR vs non-SVR. Data were combined using fixed effect (Mantel–Haenszel) as well as random effect (DerSimonian and Laird) models. Random effects were selected when heterogeneity was present.
The phenotypes analysed in the current review were: (a) spontaneous viral clearance (SVC), defined as patients showing positive anti-HCV by EIA 3.0 and repeatedly negative HCVRNA by polymerase chain reaction (PCR); (b) histological fibrosis assessed by Metavir, Scheuer or Ishack index. Fibrosis progression rate was calculated in some studies when the date of infection was known. An index was derived as the fibrosis stage divided by the period of time between infection and liver biopsy. Patients could be classified in cross-sectional analysis as advanced vs non-advanced fibrosis or according fibrosis progression rate as 'faster' or 'slower' fibrosis; (c) SVR was defined as negative HCVRNA at 6 months from the end of therapy.
Genetic Factors and Susceptibility to Hepatitis C
Following acute HCV infection, spontaneous resolution of HCV infection has been reported as being between 10 and 80% of those infected; approximately a quarter of patients with acute HCV spontaneously clear the virus.[1] Race and gender are the main factors involved in spontaneous clearance, together with clinical presentation (clearance is more often seen after icteric hepatitis), absence of human immunodeficiency virus (HIV) co-infection, rapid decline of HCV RNA and the strength and pattern of hepatitis C-specific CD4 cell responses.[2] Host factors have been involved in infectious diseases under the paradigm of gene–environmental interaction. To control infectious agents, multicellular organisms have developed an extremely sophisticated defense network, i.e. the immune system, which serves to repel viral and bacterial infections and to avoid pervasive distribution of microorganisms within tissues that may compromise overall hemostasis and survival. Genome-wide RNA interference experiments have revealed a panoply of host genes involved in HCV propagation.[3] Host genetic factors contribute towards explaining differences in the natural history of several infectious diseases such as leprosy,[4,5] AIDS,[6,7] dengue,[8] malaria[9] or chronic hepatitis B.[10] Further, there are specific genomic variants related to the immune system, such as those observed within MAL gene, that can confer on the carriers of these variants a degree of protection against a variety of microorganisms.[11] HCV infection is not an exception, and the discovery of host factor contributing to HCV pathogenesis will help us to understand host–virus relationships, and to improve our management of the HCV pandemic.
An Update on the Search for Hepatitis C Virus-related Genetic Host Factors
The candidate gene approach has been the most widely used in identifying genetic factors involved in HCV susceptibility. This implies that almost all studies performed to-date are hypothesis driven and that the selections of candidate genes have been based on the knowledge accumulated on HCV lifecycle, or genes involved in the functioning of the immune system. An analysis of the selected candidate genes reveals that investigators have scanned the human leucocyte antigen (HLA) region in detail, as well as cytokines related to natural and acquired immunity such as IFN, tumour growth factor (TGFβ1) and IL pathways. In contrast, other genes related to viral lifecycle have received less attention (see details later).An Update on the Search for Hepatitis C Virus-related Genetic Host Factors
Candidate Genes from the Human Leucocyte Antigen Region
The HLA region, located at chromosome 6p21, comprises several of genes involved in antigen recognition and host immunity. Alric et al.[12] originally proposed the contribution of HLA region to HCV risk. Antigen class II DQB1*0301 and DRB1*1101 alleles have been associated with spontaneous clearance of HCV in several independent studies with individuals of different ethnic backgrounds, and supported by meta-analyses.[13] Indeed, high-resolution genotyping of major histocompatibility (MHC) antigens class I and II of a large multiracial cohort of women from the USA infected by HCV demonstrated that several class I and class II alleles are associated with HCV viraemia.[14] Although the association between HLA region and SVC seems incontrovertible, the specific variants involved in this genetic association, the exact proteins related to SVC and/or the molecular mechanisms of such variants remain largely unclear. Further, the strong (and long range) linkage disequilibrium (LD) within the HLA region and the co-existence of multiple optimal candidate genes within this DNA segment complicate the positional cloning of true functional variants underlying the observed genetic associations with SVC. For example, it was proposed that one of the contributing genes for this association could be tumour necrosis factor (TNF). TNF locus is located at 6p21.33 within the HLA region between lymphotoxin alpha and beta (LTA and LTB) loci. Overall, TNF is located closer to HLA-G and HLA-B regions than to the HLA-DQ locus and it was originally proposed to play a critical role in immune response to HCV infection.[15] However, a meta-analysis of 12 independent studies comprising 1395 cases and 1288 controls for the −308A/G marker revealed a slight, and statistically non-significant, effect of this marker for the risk of HCV infection (OR=1.18; P=0.096).[16] However, because of this weak effect, combined with the observed trend towards association, it cannot be ruled out that TNF variants might contribute to HLA region-SVC phenotype association by acting in concert (additively or epistatically) with other polymorphisms within the HLA region.[17] Indeed, MHC classes I and II loci are >1.2 megabases away, and the genetic signals observed are probably tracking different functional variants. Individual differences in spontaneous resolution of HCV infection could be explained by the existence of epistasis between HLA region and other unlinked loci. It was hypothesized that the existence of epistasis between MHC class I and killer cell immunoglobulin-like receptor (KIR) genes might modulate the spontaneous clearance of HCV. Notably, homozygous carriers of genotypic combinations of KIR (2DL3/2DL3) and its natural ligand, HLA-C (C1/C1 allotype), are protected against HCV infection (OR for spontaneous resolution=1.71, CI: 1.2–2.42; P=0.003).[18] Further, this original study also suggested that a therapeutic reduction of inhibitory signals to natural killer might help the host in the spontaneous clearance of HCV. However, studies to replicate the effect of this genotype pair have produced contradictory results, i.e. positive association in some series,[19,20] and negative in others.[21] Analyses using larger series and exhaustive meta-analyses are necessary for a definitive view of these interesting findings.
Other Candidate Genes Analysed
Other Candidate Genes Analysed
Early-on, the existence of genes outside of the HLA region had been hypothesized in relation to SVC.[22] Several investigators selected markers within/near IL genes such IL1, IL6, IL4, IL10, IL12B, IL18, IL19/IL20, IL22 or IL10RA and their natural receptors, as candidate genes in exploring HCV infection risk.[23–31] Overall, these studies lack the consistence observed for the HLA region; the findings have not been systematically confirmed[22,24,26,32,33] and remain controversial. Most of these studies need to be considered preliminary and require corroboration in larger series. IL10 is located at 1q32 chromosomal region within a cluster of paralog genes comprising IL10, IL19, IL20 and IL24 cytokines. The protein encoded by this gene is a cytokine produced primarily by monocytes, and to a lesser extent, by lymphocytes. This cytokine has pleiotropic effects in immunoregulation and inflammation. It down-regulates the expression of T helper 1 (Th1) cytokines, MHC class II antigens and co-stimulatory molecules on macrophages. It also enhances B cell survival, proliferation and antibody production. High IL10 levels have been related to disease progression[34] and IL10 polymorphisms have been associated with SVC in at least four independent studies.[23,25,28,29]
Some genetic signals observed in loci such as TGFβ1,[35] CTLA4,[36] interferon regulatory factor-1 (IRF-1),[37,38] C–C motif chemokine receptor gene cluster,[39,40] LTA (a gene of the HLA region),[41] STAT1 or interferon alpha2 (INFA2)[42] and interferon gamma[43] require further validation. Lastly, genetic elements not directly related to the immune system, most of them HCV-interacting proteins, are involved in different stages of HCV lifecycle. The role of claudin 1 coreceptor, involved in late-stage HCV binding to the cell,[44] or polymorphisms in the low-density lipoprotein receptor (LDL-R) gene involved in viral endocytosis[45,46] together with apolipoproteins (apoE and apoB) are used by HCV as blood vehicles that facilitate interaction with the LDL receptor during HCV endocytosis.[47,48] ApoE can also be recruited by HCV machinery during late stages of lifecycle for viral assembly.[49] APOE¢4 alleles might protect against severe liver disease[50] but APOE¢3 allele has been associated with HCV infection persistence.[51] In addition, variability in the promoter region of the APOB gene might also modify HCV susceptibility.[49] Hence, the apolipoprotein variability hypothesis for HCV infection seems attractive, and further research is warranted. Factors influencing the interaction between host and the virus, such as the paraoxonase-1 (PON1)-192 polymorphism[52] could contribute, together with other polymorphisms, to the variations in the host response to HCV infection.
Genome-wide Association Studies in Susceptibility Analysis
Genome-wide Association Studies in Susceptibility Analysis
In addition to information obtained using hypothesis-driven studies (the candidate gene approach), we now have information from early GWAS analyses conducted to identify the mechanisms underlying spontaneous HCV clearance. Overall, these hypothesis-free investigations using massive, parallel, genotypic research technologies (DNA arrays) combined with the emerging independent studies that reproduce these findings have provided incontrovertible evidence that genomic variation near the IL28B locus is directly related to SVC. IL28B encodes a cytokine distantly related to type I IFNs and the IL10 family. Together with IL28A and IL29, there are three closely related cytokine genes that form a cytokine gene cluster on a chromosomal region mapped to 19q13. Expression of the cytokines encoded by the three genes can be induced by viral infection. All three cytokines have been shown to interact with a heterodimeric class II cytokine receptor that consists of interleukin 10-receptor beta and interleukin 28-receptor alpha. Both receptors are now strong candidates for further genetic analysis. In the initial study, Ge et al.[53] found a higher prevalence of the CC genotype in healthy individuals compared with those with chronic hepatitis C (CHC) (73 vs 63%; P<2.5 × 10−6). However, these differences have not been confirmed in subsequent studies. Thomas et al.[54] evaluated a cohort of 388 patients who spontaneously resolved the infection and compared them with 620 patients with CHC. Genotype CC from the rs12979860 in the IL28B gene was found in 58.5% (227/388) of patients with spontaneous clearance and in 32.6% (202/620) of patients with CHC (P=3 × 10−13). These results have been confirmed in two further studies. Montes-Cano et al.[55] confirmed that the prevalence of genotype CC in patients with SVC was two-fold that of chronic carriers. Rauch and colleagues conducted a GWAS and identified a genetic signal in chromosome 19 that was strongly related to spontaneous clearance, irrespective of HIV co-infection or route of infection. Moreover, gender remained an independent variable associated with spontaneous clearance (OR: 1.7; 13–2.4).[56] Indeed, the distribution of IL28B rs12979860 genotype CC was similar in males and females (72.5 vs72.4%; P=NS) with spontaneous clearance. Lastly, spontaneous clearance rate was lower in males and in Afro-Americans (AAs). In AA, the prevalence of the polymorphism was higher in patients with SVC [genotype CC: 33% (32/97) vs. 13.5% (26/193) than in patients with CHC; P=1 × 10−4].
In summary, as hypothesized previously, genetic factors play a key role in clearing the virus post-infection. HLA class II and some genetic signals located in the 19q13 region, including IL28B, KIR2DL3, TGFβ1, LDLR and APOE, are the strongest predictors of spontaneous clearance (Table 1) Table 1b, 1c.
Genetic Factors Related to Fibrosis Progression
Genetic Factors Related to Fibrosis Progression
Fibrosis progression is the main prognosis determinant of liver disease outcome in CHC.[57] The natural history of the infection clearly shows a subgroup of patients chronically infected with HCV (a range from 5 to 20% in several cohort studies[58]) who gradually could progress to cirrhosis and to end-stage liver disease. Patients may be classified according to Metavir score from F0 to F4. In genetic studies, the fibrosis phenotype could be analysed in two different ways: (a) patients who reached advanced fibrosis (F3–F4 vs F0–F2); (b) fibrosis progression rate, i.e. dividing fibrosis detected in liver biopsy (0–4) by duration of infection (in years). Variables associated with fibrosis progression include: (a) age at infection; (b) alcohol intake (>50 g/day); (c) male gender; (d) hepatitis B co-infection; (e) immunodeficiency because of HIV or the use of immunosuppressant drugs such of those used after liver transplantation; (f) excess weight; (g) liver steatosis; (h) presence of metabolic syndrome and/or type II diabetes; (i) iron overload.[59] Host genetic factors might have a relevant influence on the natural history of CHC. Many studies have implicated several SNPs (single nucleotide variations at specific positions of the genome detected in more than 1% of population) in order to analyze their possible impact on fibrosis progression, and the risk of hepatocellular carcinoma.[60]
Candidate Genes Implicated in the Immune and Inflammatory Response: Human Leucocyte Antigen and Interleukins
Candidate Genes Implicated in the Immune and Inflammatory Response: Human Leucocyte Antigen and Interleukins
Human leucocyte antigen class II haplotypes, mainly of DRB1, were associated with persistently normal alanine transaminase (ALT) and mild fibrosis[61,62] in a cohort of 83 patients with normal ALT levels over a 6-month period and 233 patients with elevated ALT. HLADRB1*11 was overrepresented in those with normal ALT levels (43 vs 24%, OR: 2.36) and mild fibrosis.[63] Moreover, DRB1*11 allele was associated with a lower progression rate (1.58 vs 2.14) and a lower probability of developing cirrhosis.[64] However, other studies did not find this association between HLA class and fibrosis progression.[65,66] Association studies of fibrosis progression with polymorphisms in chemokine receptor 5 (CCR-5), monocyte chemotactic protein 2 (MCP-2) and monocyte chemotactic protein-1 (MCP-1) are conflicting in their findings. Hellier et al.[67] included 337 patients in their study and found that the Δ–32 deletion was associated with more advanced fibrosis (OR=1.97; P=0.015), but with reduced portal inflammation. These results have not been confirmed in other studies but, instead, have been contradicted.[68–70] Despite an initial association observed between MCP-2 and MCP-1[71] with fibrosis, subsequent studies have not confirmed this association.[72,73] Kato and colleagues[74] studied nine SNPs of the interferon regulatory factor-7 (IRF-7) gene (four of these SNPs in the promoter region) in 406 patients (178 with cirrhosis). Two non-synonymous SNPs at positions 1047 and 2157 (A-to-G in both cases) resulting in amino acid changes (Lys/Glu and Gln/Arg respectively) were reported. The polymorphisms 1047AG and 2157AG were in complete LD, and they were more frequently seen in cirrhotic patients (5.6%) than in non-cirrhotic individuals (1.7%) (OR: 3.27; P=0.03). However, there was no association between SNPs in the promoter region and the presence of cirrhosis. In multivariate analysis, 1047AG and 2157AG were independently associated with cirrhosis (AA vs AG-adjusted OR: 2.5; 95% CI: 1.2–5.6; P=0.02). IRF-7 has been found to affect immune responses mainly by regulating the transcription of IFN-stimulated genes.[75] Additional possible mechanisms for the effect of IRF-7 gene polymorphisms on the progression of liver fibrosis include the IRF-7-induced activation of IFN-β and regulated on activation normal T cell expressed and secreted (RANTES). RANTES serves as a key ligand for CCR5 and plays a significant role in attracting T cells to the portal area of the liver infected with HCV. The activation of RANTES has been suggested to be involved in the progression of CHC to advanced forms of liver disease.[76,77] IL10 is known to influence the Th1/Th2 cytokine profile, affecting both the innate and adaptive immune responses to infection. HCV has been shown to induce the activation of IL10 secretion, and increased IL10 production has been observed to correlate with persistent HCV infection, higher inflammation grade and an increased risk of liver cancer. Two receptor chains, IL10RA and IL10RB, are known to mediate the functions of IL10. Although SNP in the minor allele in G330R IL10R1[78] and the homozygosity for two IL10 haplotypes:[79,80] −819 (Cto-A); and −1082 (AA genotype, ATA/ATA and ACC/ACC) have been reported to be associated with faster fibrosis rates, contradictory findings have been observed in different studies evaluating the significance of this last-mentioned polymorphism. Hennig et al.[30] examined 631 HCV patients and found a variation in IL10RA, IL10RA-rs9610 (3'-UTR), which appeared to be correlated with reduced inflammation, individuals carrying an A allele being less likely to present with severe inflammation (OR=0.29; 95% CI: 0.10–0.83; P<0.05). IL18, also called IFN-c–inducing factor, is an obligatory cytokine for IFN-c production, and plays a key role in the induction of Th1 responses and viral clearance as well as in the development of liver fibrosis. Single-nucleotide promoter polymorphisms influence the transcription of IL18 mRNA. Pro-IL18 is a biologically inactive precursor that is produced by monocytes, macrophages and immature dendritic cells during an acute immune response. It is activated intracellularly by caspase-1 to induce IFN-c, iNOS (inducible nitric oxide synthetase), TNF-α secretion, as well as induction of other inflammatory cytokines such as IL6, IL8, IL2, MCP-1, MIP-a and MIP-b. Increased IL18 production is neutralized by IL18BP in CHC infection, and this neutralization is crucial for the regulation of inflammation and fibrosis development. Manohar et al.[81] have investigated the association between the −607 polymorphism and severity of HCV infection. They evaluated 204 patients with CHC and 350 matched healthy controls. The −607 A/A allele was more common in patients with mild disease than in patients with severe diseases (38.6% vs 21%, OR=0.424; P=0.05). IL18 promoter −607 A/A allele is a potential protective marker in patients with CHC. These results have been confirmed by another study.[82] IL12 is a cytokine that induces production of the IFNγ. Suneetha et al.[83] showed that homozygosity for the minor allele of this SNP, 1188C/C, was more common in patients who had mild fibrosis compared with those with severe fibrosis (23.7 vs 6.25%; P=0.004). Several studies have shown that TNF may play a role in the pathogenesis of CHC by influencing fibrosis progression rate. Dai et al.[84] investigated the biallelic polymorphism G vs A in the promoter region at positions −308 (TNF308.2) and −238 (TNF238.2). In 250 biopsy-proven CHC individuals, the TNF 308.2 allele copy numbers were significantly associated with more severe fibrosis stage (F3–F4; P=0.006) and higher mean fibrosis score (P=0.007). Logistic regression analysis showed that a higher fibrosis score was independently related to the TNF308.2 allele (OR: 1.38). However, conflicting results have been reported on this association.[85,86] ApoE binds to some cellular receptors including proteoglycans, heparan sulphate and LDL-R. The HCV competes with the apoE in the process of entry into the cell via the LDL-R. Specifically, apoE4 is the protein subtype with more affinity for these receptors while the apoE-2 has the lowest affinity. Patients carrying ¢4 allele seem to be protected against cirrhosis (4.3 vs 19.1% in ¢2/¢3 allele carriers). In a cohort of 111 patients, the ¢4 allele was associated with a lower risk of cirrhosis when evaluated in the multivariate analysis that took into account confounding factors such as gender, alcohol intake, time of evolution of the infection and the age at biopsy[87] However, APOB SNPs did not influence liver fibrosis rate.[47] SNPs at positions 1874 (T-to-A) of the IFN-γ gene has also been described as accelerating the rate of fibrosis progression in HCV patients.[88] The enzyme 2'-5'-oligoadenylate synthetase 1 (OAS-1), an important component of the innate immune system, has an antiviral function.[89] Li et al.[90] reported the association between six SNPs of OAS-1 and fibrosis in 409 patients with HCV. Patients with rs3741981 genotypes A/A, A/G and G/G of an SNP of OAS-1 at the exon 3 were at a gradient increased risk of fibrosis progression and suffering from cirrhosis (P=0.001). Multivariate logistic regression analysis indicated that genotype G/G was an independent factor associated with cirrhosis (OR: 3.11; P=0.013). Mannan-binding lectin (MBL), encoded by the MBL2 gene, can have an important role as an opsonin and complement-activating molecule with an important function in the initiation, regulation and amplification of immune response. A small number of studies have examined the relationship between MBL polymorphisms and HCV infection. These studies differ with respect to the cohorts used, categorization of subjects and the MBL gene mutations investigated. Brown et al.[91] demonstrated an association of higher MBL/MASP-1 complex activity (related to polymorphisms in the promoter and structural regions of MBL gene) with severity of fibrosis in HCV. In 102 Euro-Brazilian patients,[92] MBL2 polymorphism was associated with moderate and severe fibrosis in CHC, after controlling for gender and age. Six common SNPs, three in the promoter (H/L, X/Y and P/Q) and three in exon 1 (A, the wild-type, and B, C or D also known as O) were evaluated using real-time polymerase chain reaction (RT-PCR) with fluorescent hybridization probes. The frequency of the YA/YO genotype was significantly higher in H patients vs the controls (P=0.022). Genotypes associated with low levels of MBL (XA/XA, XA/YO and YO/YO) were decreased significantly in the patients with severe fibrosis (stage F4) compared with patients with moderate fibrosis (stage F2) (P=0.04) and to the control group (P=0.011). Genes involved in the transduction of myxovirus (MxA) and protein kinase R (PKR) together with pro-inflammatory cytokines involved in IFN resistance such as IL10, TNF, IL6, GH and IL1 have been selected for candidate gene analyses. Yee and colleagues investigated the association between myxovirus resistance-1 (Mx1), PKR, liver fibrosis in 374 treatment-naive patients with genotype-1 chronic HCV infection (194 Caucasian Americans CAs and 180 Aas), using a genetic haplotype approach. He reported independent association between Mx1-CAGT and PKR-TGATT and less severe hepatic fibrosis even after control for the confounding factors such as race and gender. However, the associations were not statistically significant in a second, independent validation cohort.[93]
Candidate Genes Implicated in Iron Metabolism as Fibrogenic Factor
Candidate Genes Implicated in Iron Metabolism as Fibrogenic Factor
Iron overload seems to induce a deleterious effect on fibrosis progression rate. However, the relationship between the presence of HFE gene mutations and disease progression remains controversial. Some authors have reported a strong association between mutations and fibrosis progression[94] while others have not observed these relationships.[95–97] Thorburn et al.[98] in a large cohort did not detect association, whereas Erhardt et al.[99] did, and highlighted that their systematic study demonstrated that the association between HFE gene mutations and progression was independent of other confounding variables. Tung et al.[100] studied 316 CHC patients and highlighted that heterozygous mutations in both exon 2 and exon 4 were associated with increased histological damage, and a faster fibrosis progression rate. No associations between histological progression of liver disease and TfR1 SNPs[101–103] have been reported. Further, Nramp1 protein, now termed solute carrier family 11 member 1 (SLC11A1) protein is located in the late endosomal compartment of resting macrophages, and is recruited to the phagosome by phagocytosis and seems to be an intracellular transporter of iron, while playing an important role in immune response against intracellular microorganisms such as HCV. Several mutations resulting in polymorphic mRNA expression have been identified in the SLC11A1 gene (2q35). Four alleles have been found in different populations with the absence of allele 3 exerting a protective effect on progression to cirrhosis in patients with HCV. In a Spanish cohort that included 242 patients with biopsy-proven CHC and 194 healthy control subjects, allele 3 carriers showed faster fibrosis progression rate of 0.16 ± 0.20 vs 0.09 ± 0.08 units of fibrosis/year. Hence, the 2/2 genotype of the promoter region of the SLC11A1 gene was found to be associated with mild portal inflammation and a lack of advanced fibrosis (OR: 8.85; P=0.002). Indeed, the interaction between allele 3 of SLC11A1 and the -238 A/G mutations in the promoter region of TNF gene appears to promote accelerated progression of fibrosis (OR: 2.53; P=0.039).[104]
Candidate Genes Implicated in Hepatic Stellate Cell Activation
Candidate Genes Implicated in Hepatic Stellate Cell Activation
TGFβ1 plays a significant role in hepatic stellate cell activation, and increased levels of this chemokine have been related to a faster rate of fibrosis progression.[105] Several polymorphisms have been described: Arg/Pro mutation at codon 25 (G/C carriers) induces a rapid progression to cirrhosis with a progression rate of 0.23 vs 0.008 of the Arg/Arg group (non-mutated TGFβ1). These results proceeded from studies with small series of subjects but have been confirmed in further cohorts.[106–109] Thus, SNPs from these 2 genes (TGFβ1 and AT) are confirmed as being implicated in fibrosis progression. Several factors such as hyperhomocysteinaemia and C protein deficiency as well as increased factor VIII:C have been associated with fibrosis progression.[110] Factor V Leiden mutation A560G has been associated with an increased risk of cirrhosis and faster fibrosis progression (0.37 vs 0.18 units of fibrosis/year).[111,112] Inhibition of thrombin receptor protease-activated receptor 1 polymorphism, 1426 (C/T instead of T/T);[113] or myeloperoxidase gene −463A have been shown to be associated with greater fibrosis.[114,115] With the haplotype approach, Mx1-CAGT and PKR-TGATT haplotypes of antiviral genes were independently associated with mild liver fibrosis, following adjustment for potential confounders (for Mx1-CAGT haplotype: OR: 0.33; P=0.0027; for PKR-TGATT haplotype: OR: 0.56; P=0.0405). These findings were validated using an independent cohort in which a protective trend for the PKR-TGATT and Mx1-CAGT haplotypes was confirmed, albeit the association with the latter haplotype was not statistically significant.
Genome-wide Scan and Fibrosis Progression
Genome-wide Scan and Fibrosis Progression
A genome-wide scan including 24 823 candidate SNPs from 12 248 covering genes in 433 biopsy-proven CHC individuals identified 100 SNPs associated with an increased risk of advanced fibrosis.[116] In 483 patients from the validation cohort, only two out of 100 were found to be associated with advanced fibrosis. A missense SNP in the DEAD (Asp–Glu–Ala–Asp) box polypeptide 5 gene causing an amino acid replacement at position 480 (S480A) in exon 13 was associated with an increased risk of advanced fibrosis (OR: 1.8 and 2.3 in the 2 cohorts), while a missense SNP in the carnitine palmitoyltransferase 1A gene causing amino acid change at position 275 (A275T) in exon 8 was associated with a decreased risk for advanced fibrosis (OR: 0.3 and 0.6 in the 2 cohorts). A cirrhosis risk score (CRS) based on genetic markers identified from two Caucasian cohorts was derived from 361 SNPs showing associations with fibrosis. Seven SNPs (one SNP in the antizyme-inhibitor-1 gene, one SNP in the Toll-like receptor-4 gene and 5 SNPs in five other genes of unclear function) showed the highest predictability for cirrhosis.[117] CRS offered a better prediction of cirrhosis compared with clinical factors (age, gender and alcohol abuse). Two CRS cut-off values were eventually suggested to identify the majority of patients at low risk (<50) and those at high risk (>70) of developing cirrhosis. Genetic CRS has been recently confirmed by Li et al.[118] in 420 Caucasian individuals and by Marcolongo et al.[119] in 271 patients with mild fibrosis followed-up over 60 months without therapy. The best prediction accuracy of CRS was in males with no fibrosis at baseline.
Genetic Variations and Sustained Response to Peginterferon+ribavirin Treatment
Genetic Variations and Sustained Response to Peginterferon+ribavirin Treatment
Sustained virological response rate variability following a course of Peg-IFN/RBV treatment is extremely high. Genotype, viral load, fibrosis and metabolic disturbances including obesity, insulin resistance and steatosis were the factors most influential in SVR.[120] Several SNPs from candidate genes have been associated with achieving SVR in patients receiving Peg-IFN/RBV treatment. Regulatory genes of IFN antiviral activity, immune-response genes and genes implicated in obesity and in insulin resistance have been analysed. The mechanism of action of IFN has been characterized, and the key points identified are: (a) interaction with IFN alpha receptor; (b) Janus-kinase and tyrosin-kinase activity; (c) STATs phosphorylation; (d) synthesis of antiviral proteins such as 2'–5' OAS, MxA protein induced by IFN. Intracellular MxA, which works like GTPase to achieve its antiviral effect, seems to be the most specific marker of antiviral activity of IFN. In patients receiving induction doses of IFN, MxA levels increased in parallel with antiviral activity.[121] SNP −88T in the MxA gene was found to be associated with lower MxA protein activity. In patients with a low viral load, SVR was statistically significantly higher (62%) in −88T patients than in patients bearing the −88A allele (32%).[122] The 2'–5' OAS enzyme plays a major role in the clearance of the virus. However, some studies that had included analysis of the GG genotype (in the 3'UTR region) do not predict SVR.[89] Lastly, a tandem repeat of three nucleotides in the PKR gene classified as 'large' when containing >9 repeats has been found to be associated with SVR. Large/large polymorphism was more often seen in patients achieving SVR than in non-responders (89.4 vs 71.8%; P=0.017).[13] ApoE has been implicated in the mechanism of entry of the HCV into the cell via the LDL-R. In a cohort of 506 patients treated with Peg-IFN/RBV, the ¢4 allele was found to be associated with poorer response in patients with genotype 1 (30 vs 42%; P<0.05).[123] Several polymorphisms from proinflammatory cytokines have been included as candidate genes in the prediction of SVR. The biallelic polymorphism in TNF (−238 and −308) seems not to be associated with SVR.[124] TGFβ1 and interleucin-10 polymorphisms have been strongly related to achieving SVR. Genotype −29 C/C in TGFβ1 promotes resistance to Peg-IFN/RBV treatment.[125] Further, genotypes −592 A/A and −819 T/T in the IL10 gene have been found to be linked to higher SVR.[126] Lastly, in a multivariate analysis of 105 patients treated with IFN/RBV, HLA class I B44 was seen to be independently associated with improved SVR to combined IFN/RBV, together with viral non-1 genotype. However, no association between this allele and SVR was detected in patients receiving IFN alone.[127] In spite of these well-selected candidate genes, and some associations being confirmed in multivariate analysis, the majority of them showed minor impact on clinical practice, and they have not been included in the daily management of patients with CHC.
Several pharmacogenetic studies using GWAS for HCV treatment response assessment have demonstrated relationships between several polymorphisms in the 19q13 region and SVR (Table 2). Ge and colleagues conducted a GWAS analysis in 1137 patients of a cohort from the IDEAL study; a trial comparing Peg-IFN α-2b in two different doses (1.0 μg/kg/week vs. 1.5 μg/kg/week) versus standard doses of Peg-IFN α-2a. Using the Illumina Human 610® quad bead chip, the authors demonstrated that the probability of achieving SVR in patients bearing CC in the position rs12979860 in 19q13 region was double that of those with CT/TT (OR: 2; 95% CI: 1.8–2.3; P=1.37 × 10−28). Moreover, the distributions of this CC genotype in several World populations were strongly related to SVR whether in Asians, Europeans, Hispanic or AAs. Tanaka et al.[128] conducted a GWAS analysis in 154 Japanese patients; 82 non-responders and 72 with SVR using Affimetrix SNP 6.0® genome wide SNP typing array testing 621 220 SNPs. Several genetic signals in the 19q13 region were observed to be strongly related to SVR (rs12980275; P=1.93 × 10−13 and rs8099917; P=3.11 × 10−15). Sequencing a 40 kb region in the 19q13 region indicated that 7 SNPs were strongly related to each other, suggesting that the association with SVR was primarily driven by one or other of these SNPs. Further, using quantitative RT-PCR, IL28B mRNA was found to be higher in patients who were homozygous carriers of the major allele. Lastly, multivariate analysis indicated that rs8099917 (G allele) and female gender were independently associated with SVR while fibrosis, markers of liver dysfunction (such as platelets) or viral load were excluded from the final multivariate model. Suppiah et al.[129] conducted a GWAS of SVR to Peg-IFN/RBV in 293 Australian patients with genotype 1, and a validation cohort of 555 individuals. An association between rs8099917 in the IL28B gene and SVR was observed (OR: 1.98, 95% CI: 1.57–2.52; P<0.05) confirming the previous data from Ge and colleagues. In a recent GWAS analysis, all these markers in the 19q13 region were found to be associated with SVR but, after multivariate analysis, rs12979860 was found to be independently associated with the chance of achieving a cure, and as such, supporting a major role for this genetic signal.[130] These results have since been confirmed in different series such as those of McCarthy et al.,[131] Del Campo et al.,[132] Montes-Cano et al.[55] and Rauch et al..[56] Further, this polymorphism has also been strongly associated with the possibility of achieving SVR in patients infected by genotype 1, without rapid virological response (RVR) (clearance of the virus after 4 weeks of treatment).[133] Conversely, in patients with CHC non-1 genotype, IL28B polymorphism rs8099917 G was not associated with higher SVR. In 230 patients with genotype 2 or 3 receiving Peg-IFN/RBV, SVR was 79.5% in patients bearing the G allele vs 86.4% in patients with the T allele (OR: 1.62; 95% CI: 0.8–3.3; P=NS).[58] Recently, Mangia et al.[134] confirmed the usefulness of genotype CC in predicting SVR in patients with genotype 2/3 without RVR, but not in the overall cohort (Table 3).
Further, favourable polymorphism rs12979860 was more often seen in genotype 2, 3 than genotype 1 (Fig. 1) suggesting that IL28B polymorphism not only strongly influences SVR but also appears to explain much of the difference in response observed between population groups representing different viral genotypes and host ethnicity.[57] A meta-analysis including all these studies was conducted (Fig. 2) and all studies confirmed the association between genotype CC rs12987960 and SVR (OR: 4.5; 95% CI: 2.8–7.3). In a multivariate analysis of HCV-4 patients, baseline viral load, fibrosis and the IL28 T allele (OR: 0.124, 95% CI: 0.030–0.505) were significantly associated with SVR.[106] The strongest predictor for the final outcome was RVR (OR: 26.00; 95% CI: 7.148–94.545, P<0.0001). If RVR was included in the multivariate model, only the RVR and the fibrosis score remained significant. Thus, determination of IL28 polymorphism may not be useful to select patients with HCV-4 for abbreviated treatment schedules.[135] IFN-λ is transduced from IL28B gene (19q13). It binds to the heterodimer receptor complex composed of 2 subunits IFN-λ-R1 and IL10-R2. After receptor binding, IFN-λ has been observed to promote the JAK-STAT antiviral pathway, inducing phosphorylation of Janus-kinase1 and tyrosine-kinase2, promoting signal transducer and activating transcription (STAT) phosphorylation. After coupling with IFN, regulatory factor 9 translocates to the nucleus and binds the IFN-stimulated response element in DNA to initiate gene t
ranscription, mainly IFN-stimulated genes including 5'2'OAS, protein-tyrosine-kinase, IL8 and IFN-regulated factor-7. Thus, IFN-l, seems to be able to: (a) inhibit HCV replication; (b) down-regulate HLA-I presentation; (c) inhibit the entry of HCV particles into the ribosome; (d) have better haematological tolerance than IFN alpha, mainly because of reduced expression of IFN-λ-R1 receptor in blood cells. Nevertheless, the specific genetic variant involved remains to be determined, together with the amino acid change that could promote the protein. Indeed, antiviral activity of IFN-λ3 varies depending on the final amino acid change: Val97 to Ala increases 68-fold the antiviral activity, while Arg51 to Ala does not modify it.[136]
Figure 1.
Genotype CC (rs12979860) distribution in spontaneous viral clearance (n=69), healthy individuals (n=1169) and chronic hepatitis C (n=524) in Spain.
Figure 2.
Estimated impact of rs12979860 genotype on the possibility of achieving sustained virological response in patients infected by genotype 1. Note: Half of the responders are non-CC genotype.
.
.
Overall, these hypothesis-free investigations using massive, and parallel, genotypic research technologies such as DNA arrays, together with independently reproduced studies, have provided incontrovertible evidence that genomic variation in or around the IL28B locus is directly related to SVR induced by Peg-IFN/RBV therapy. The genetic variant (or variants) needs to be described, as are the functional analyses to demonstrate the impact of these changes on the host's ability to clear viral infection. Further, the role of the genetic alteration in patients infected by non-1 genotypes should be demonstrated; IL28 polymorphism may not be useful in selecting patients with HCV-4 for abbreviated treatment schedules. However, these data need further confirmation before final conclusions can be formulated. Lastly, from a clinical point of view, approximately a third of patients bearing the CC genotype (rs12979860) did not achieve SVR, while on the other hand, nearly a half of CT heterozygous and a third of TT homozygous individuals could achieve SVR when treated with Peg-IFN/RBV. Thus, genetic factors that could modulate (positively or negatively) the effect of this polymorphism on SVR warrant further exploration. Perhaps other genetic markers could explain this gap between IL28B genotype and SVR[137] (Fig. 3).
Click Figure To Enlarge
Figure 3.
Meta-analysis showing a strong association between sustained virological response and genotype CC (rs 12987960).
/
/
In conclusion, the lessons from the review of these studies are: (a) there are innate host characteristics that model HCV lifecycle and infection resolution; (b) the investigation of SVR and SVC are fully complementary. This means that host genetic factors might behave similarly by simultaneously affecting both phenotypes; (c) together with HLA region studies and meta-analyses related to them, GWAS have provided overwhelming evidence indicating that the host immunological hypothesis for virus clearance is plausible; (d) the isolation of this unanticipated factor (IL28B) will provide new research opportunities and will have a considerable clinical impact on HCV diagnosis, prognosis and therapy in these patients; (e) based on genetic studies, spontaneous clearance or fibrosis progression can be considered as complex phenotypes. This assumption implies that many other genetic and non-genetic factors need to be identified, using candidate gene or hypothesis-free approaches. To achieve these goals, genetic markers need to be studied in large cohorts of patients, keeping in mind their interaction with environmental and viral factors, which could affect the natural history of CHC. We feel that genetic studies are opening up a new era in HCV investigation. However, despite recent successes, there is still a considerable gap between genetic discoveries in the laboratory and application of the findings to innovative clinical practice; fortunately, it is a gap that is continuing to close.
No comments:
Post a Comment