SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Background  Hepatitis C virus (HCV) co-opts very-low-density lipoprotein (VLDL) pathways for replication, secretion and entry into hepatocytes and associates with apolipoprotein B (apoB) in plasma. Each VLDL contains apoB-100 and variable amounts of apolipoproteins E and C, cholesterol and triglycerides.

Aim  To determine whether baseline lipid levels predicted treatment outcome.

Methods  Retrospective analysis was performed of 250 chronic hepatitis C (CHC) patients who had received anti-viral agents interferon-alpha and ribavirin; 165 had a sustained virological response (SVR). Pre- and post-treatment nonfasting lipid profiles were measured and non-high-density lipoprotein (non-HDL) cholesterol (i.e. apoB-associated) was calculated. Binary logistic regression analysis assessed factors independently associated with treatment outcome.

Results  There was an independent association between higher apoB-associated cholesterol (non-HDL-C) and increased odds of SVR (odds ratio 2.09, P = 0.042). In multivariate analysis, non-HDL-C was significantly lower in HCV genotype 3 (g3) than genotype 1 (P = 0.007); this was reversible upon eradication of HCVg3 (pre-treatment non-HDL-C = 2.8 mmol/L, SVR = 3.6 mmol/L, P < 0.001).

Conclusions  Higher apoB-associated cholesterol is positively associated with treatment outcome in CHC patients receiving anti-viral therapy, possibly due to competition between apoB-containing lipoproteins and infectious low-density HCV lipo-viral particles for hepatocyte entry via shared lipoprotein receptors.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The liver synthesizes and exports cholesterol and triglycerides through secretion of very-low-density lipoprotein (VLDL).1, 2 Each VLDL particle consists of a neutral core of lipids [triglycerides (TG) and cholesterol esters] surrounded by a single nontransferable molecule of apolipoprotein B (apoB)-100, transferable lipoproteins including apolipoprotein E (apoE) and phospholipids. VLDL is the precursor of low-density lipoprotein (LDL). The conversion of VLDL to LDL is a complex process;3 briefly, in the circulation, TG are hydrolysed by lipoprotein lipase; the resulting TG depleted particle (IDL – also known as VLDL remnants) can go on to become LDL or be rapidly cleared from the circulation by the liver. The clearance of VLDL remnants is primarily dependent upon apoE, a ligand for LDL receptor (LDLr).

There is increasing evidence that the lifecycle of HCV utilizes host lipoprotein pathways for replication, secretion, binding and entry of infectious particles.4, 5In vitro HCV production is blocked by two agents that block VLDL assembly6 and in vivo studies suggest that the association between HCV and VLDL occurs during VLDL assembly in the endoplasmic reticulum.7 HCV is found associated with apoB in plasma,8–10 but the importance of host lipid metabolism in influencing the natural history of HCV infection and response to anti-viral therapy is only beginning to be appreciated.

Lipid metabolism may be important in the minority who spontaneously clear HCV. Cohort studies have suggested that elevated TG facilitate spontaneous viral clearance11 and we have reported an association between the Apo E3 allele (encoding wild-type isoform with normal LDLr binding) and viral persistence.12 Chronic HCV infection is an important global health problem affecting 2% of the worlds population13 and may progress to liver cirrhosis, liver failure and hepatocellular carcinoma.14 Treatment with combination pegylated interferon-alpha (IFNα) and ribavirin15, 16 improves the prognosis in those who have a sustained virological response (SVR). Factors that influence response to therapy include male gender, older age, HCV genotype, insulin resistance17–19 and advanced fibrosis.20 Recently, there have been reports that higher baseline LDL cholesterol levels may be associated with SVR, both in mono-infected and HCV/HIV co-infected patients.21–23 Conversely, low cholesterol is a significant predictor of decreased likelihood of HCV genotype 1 patients having an SVR.24

Chronic HCV itself has been reported to influence serum lipid profiles. A community-based study of 11,239 individuals reported that lower serum cholesterol is associated with HCV infection.25 This finding may be more pronounced in HCV genotype 3 infection,26, 27 but this is controversial.28

The primary aim of this study was to determine whether outcome of therapy with pegylated IFNα and ribavirin in chronic HCV is independently associated with host lipid levels; the secondary aim was to examine some viral and host factors that may alter these lipid levels.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Study subjects

Data regarding chronic hepatitis C (CHC) patients treated with standard combination anti-viral therapy (pegylated IFNα and ribavirin) by the viral hepatitis service at Freeman Hospital, Newcastle upon Tyne has been collected in a clinical database which is used routinely for audit. The database records date of treatment, HCV genotype and treatment response. At 24 weeks post-therapy, individuals who are HCV RNA positive are considered to be non-responders (NRs), whilst those who achieve an SVR (HCV RNA negative) have durable viral clearance.16 From a search of the database, we retrospectively identified 250 consecutive CHC patients in whom treatment outcome was known.

Biochemical characteristics

Nonfasting serum lipid profiles [total cholesterol, TG, high-density lipoprotein (HDL)] were measured pre- and 24-week post-treatment, thereby allowing each patient to act as their own control. As these were nonfasting samples, LDL-cholesterol, estimated by the Friedewald calculation29 was not assessed. Therefore, apoB-associated cholesterol was calculated on each patient by subtracting HDL cholesterol (HDL-C) from total cholesterol level (non-HDL-C). Paired pre- and post-treatment cholesterol levels were available in 100 patients achieving an SVR (90 HCV genotypes 1 and 3) and 66 NRs (63 HCV genotypes 1 and 3). The remaining patients had incomplete or unpaired data.

ApoE genotyping

Apolipoprotein E genotyping was performed in 129 CHC patients who had undergone anti-viral treatment, 72 of whom achieved an SVR and 57 were NRs. This finding was used in the subsequent binary logistic regression model of treatment response. A standard PCR–RFLP method was used to determine ApoE genotype.12

Statistical analysis

The distribution of continuous data was assessed by normality tests. Age, total cholesterol and non-HDL-C conformed to a normal distribution. TG and HDL-C levels were positively skewed and therefore log10-transformed to normal distributions before parametric tests were applied. The F-test was applied to test the assumption of equal variances and then paired t-tests were used to compare paired total cholesterol, log10 TG, log10 HDL-C and non-HDL-C levels pre- and post-treatment. A two-sample t-test was used to compare the same lipid parameters between SVRs and NRs. The influence of HCV genotype and apoE genotype on lipid levels were assessed using one-way anova analysis. Categorical variables included HCV genotype, apoE genotype and treatment outcome (SVR or NR). These were compared using chi-squared and Fisher’s exact test was used where expected frequencies were <5. The extent to which total cholesterol, non-HDL-C, log10 TG and log10 HDL-C are influenced by HCV genotype was investigated using a multiple linear regression model. Predictor factors in the model included age, gender, apoE genotype and HCV genotype. Factors associated with achieving an SVR were assessed by a binary logistic regression model. The response was treatment outcome (SVR = 1, NR = 0). Continuous predictor factors in the model were total cholesterol, non-HDL-C, log10 TG, log10 HDL-C and age. Categorical factors in the model were gender, HCV genotype and apoE genotype. All statistical analysis was performed in Minitab Version 15 (Minitab limited, Coventry, UK). Statistical significance was defined at the 5% level based on two-tailed test of the null hypothesis.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The demographics of the 250 CHC patients studied are shown in Table 1. A minority (<12%) were not HCV genotype 1 or 3 and were excluded from subsequent analysis. The primary aim was to examine whether pre-treatment lipid levels were associated with treatment outcome. However, HCV genotype, age and gender are known to influence treatment outcome and may also influence lipid levels. Host APOE genotype is also known to influence nonfasting lipid levels in the general population. Therefore, we attempted to control for these interactions and confounders by performing a binary logistic regression analysis in 129 patients in whom complete data including APOE genotype were available (Table 2). This confirms the negative association of male gender [odds ratio (OR) 0.09, 95% CI 0.02–0.37, P = 0.001] and increasing age (OR 0.93, 95% CI 0.87–0.99, P = 0.021) with SVR. There was also an independent association between higher apoB-associated cholesterol (i.e. non-HDL-C) and increased odds of achieving SVR (OR 2.09, 95% CI 1.03–4.26, P = 0.042). We found no significant association of total cholesterol with SVR (OR 1.2, 95% CI 0.74–1.97, P = 0.459). There was also no association between SVR and log10 TG and log10 HDL-C. Overall, APOE genotype was not significantly associated with SVR. However, patients with APOE*E2/E3 had an increased OR of 4.93 of achieving SVR, but this was not statistically significant (95% CI 0.66–36.6, P = 0.119) because of the low frequency of this APOE genotype. Overall SVR rate in HCV genotype 1 was 45% and in genotype 3 was 63%. For those with wild-type APOE*E3/E3, the SVR rate was 44% for HCV genotype 1 and 68% for HCV genotype 3. Five patients with HCV genotype 3 were APOE*E2/E3 and all 100% (five of five) achieved an SVR compared to only 38% (five of 13) in those with APOE*E3/E4 (Fisher’s exact test P = 0.0359).

Table 1.   Baseline characteristics of age, gender, HCV genotypes and APOE genotype of chronic hepatitis C patients
 Total study populationSubgroup for binary logistic regression analysis
SVRNon-respondersSVRNon-responders
n165857257
Gender (%)
 Male58736365
 Female42273735
Age (years) ± s.d.
 Male42.5 ± 10.149.4 ± 9.842.5 ± 9.548.7 ± 10.2
 Female42.6 ± 11.256.8 ± 9.844.0 ± 11.752.3 ± 11.9
HCV genotype 1 (%)34634058
HCV genotype 3 (%)54314532
Other HCV genotypes (2, 4, 5, 6 and unknown) (%)1261510
APOE*E3/E3 (%)62.564.9
APOE*E3/E4 (%)23.629.8
APOE*E2/E3 (%)9.73.5
APOE*E2/E4 (%)2.80
APOE*E4/E4 (%)1.41.8
APOE*E2/E2 (%)00
Table 2.   Multivariate binary logistic regression analysis for factors associated with treatment outcome
VariableORP-value95% confidence interval of OR
  1. Results are the OR of the baseline event of a sustained virological response (SVR).

  2. OR, odds ratio; N.S., not significant; HDL, high-density lipoprotein.

  3. *P < 0.05.

Age0.930.021*0.87–0.99
Male (vs. female)0.090.001*0.02–0.37
HCV genotype 3 (vs. genotype 1)2.080.286 (N.S.)0.54–7.96
HCV genotype other (vs. genotype 1)1.640.699 (N.S.)0.13–20.40
APOE*E3/E4 (vs. E3/E3)0.720.611 (N.S.)0.20–2.60
APOE*E2/E3 (vs. E3/E3)4.930.119 (N.S.)0.66–36.61
Total cholesterol1.20.459 (N.S.)0.74–1.97
Non-HDL cholesterol2.090.042*1.03–4.26
Log10 triglyceride0.180.261 (N.S.)0.01–3.66
Log10 HDL cholesterol0.900.860 (N.S.)0.29–2.85

We also examined whether lipid profiles changed pre- and post-treatment in SVRs (Table 3a) and NRs (Table 3b). The sample sizes shown in Table 3 vary according to the availability of paired pre- and post-treatment data. We found significant increases in total cholesterol, non-HDL-C and TG in individuals infected with HCV genotype 3 who achieved an SVR, but not in those infected by HCV genotype 1. In contrast, HDL-C level remained unchanged pre- and post-treatment in both SVRs and NRs and in both HCV genotypes 3 and 1. In those patients who were NRs to therapy, there was no significant change in total cholesterol, non-HDL-C, log10 TG and log10 HDL levels pre- and post-treatment in either HCV genotype 1 or 3 (Table 3b).

Table 3.   Paired pre- and post-treatment lipid levels in chronic hepatitis C patients who achieved a sustained virological response (a) and in non-responders (b) between HCV genotypes 1 and 3
 nPre-treatmentPost-treatmentP-value
  1. Note that the sample size varies according to available paired data. Data are in mmol/L and expressed as mean ± standard deviation. Paired t-test was used to compare pre- and post-treatment results.

  2. SVR, sustained virological response; HDL, high-density lipoprotein; non-HDL calculated from total cholesterol – HDL.

  3. * P < 0.05, ** P < 0.001.

(a) SVRs
HCV genotype 1
 Total cholesterol354.70 ± 1.064.73 ± 0.980.907
 Non-HDL cholesterol243.63 ± 1.063.65 ± 0.980.933
 Log10 triglycerides290.16 ± 0.190.18 ± 0.220.771
 Log10 HDL290.10 ± 0.110.09 ± 0.100.782
HCV genotype 3
 Total cholesterol554.20 ± 0.855.09 ± 0.86<0.001**
 Non-HDL cholesterol422.80 ± 0.773.6 ± 0.75<0.001**
 Log10 triglycerides440.06 ± 0.230.19 ± 0.230.010*
 Log10 HDL430.11 ± 0.140.13 ± 0.140.538
(b) Non-responders
HCV genotype 1
 Total cholesterol424.66 ± 0.824.63 ± 1.040.871
 Non-HDL cholesterol333.31 ± 0.803.19 ± 0.930.562
 Log10 triglycerides420.17 ± 0.230.20 ± 0.240.630
 Log10 HDL380.09 ± 0.160.09 ± 0.150.986
HCV genotype 3
 Total cholesterol214.00 ± 1.203.63 ± 0.890.269
 Non-HDL cholesterol162.52 ± 0.992.31 ± 0.660.481
 Log10 triglycerides220.18 ± 0.260.20 ± 0.260.769
 Log10 HDL190.06 ± 0.140.01 ± 0.190.274

Finally, we attempted to assess the extent to which lipid levels are independently influenced by HCV replication and genotype, taking into consideration confounding factors of age, gender and APOE genotype. To consider the effect of HCV replication on lipid levels, a Box–Cox transformation was applied to pre-treatment viral load. However, even after transformation, viral load did not satisfy normality tests and hence we were unable to demonstrate any significant correlation in univariate regression between pre-treatment viral load and total cholesterol (r2 = 2.8%), non-HDL-C, (r2 = 4.8%), log10 TG (r2 = 0.5%) and HDL-C (r2 = 1.2%). Total HCV RNA viral load was therefore not included in the subsequent multivariate model. Considering the effect of HCV genotype, in HCV genotype 1, the variation in serum total cholesterol and non-HDL-C followed the expected trends with APOE genotype, APOE*E2/E3 being 9% lower than APOE*E3/E4 (Figure 1). However, the expected physiological increase in serum cholesterol (APOE*E2/E3 < E3/E3 < E3/E4) was not observed in HCV genotype 3. Total cholesterol and non-HDL-C was significantly lower in HCV genotype 3 compared with HCV genotypes 1 for those with APOE*E3/E3 (P = 0.008) and APOE*E3/E4 (P = 0.035), but not for APOE*E2/E3 probably because of the low frequency of this genotype in the study population. Each lipid parameter was used in turn as the outcome response in the multivariate analysis and predictor variables included age, gender, APOE genotype and HCV genotype (Table 4). We confirmed that lower total cholesterol (P = 0.015) and non-HDL-C (P = 0.007) were independently associated with HCV genotype 3. In summary, chronic infection with HCV genotype 3 lowers total cholesterol and non-HDL-C; this effect appears to be virally mediated because it is reversible upon eradication of the virus by successful anti-viral therapy. In HCV genotype 3, the expected variability in apoB-associated cholesterol levels according to host APOE genotype was not seen.

image

Figure 1.  Effect of host APOE genotype on pre-treatment total cholesterol in patients with chronic HCV genotype 1 and 3 infection (data on 13% of patients with HCV genotypes other than 1 or 3 is not shown). The expected pattern of serum cholesterol level is APOE*E3/E4 > E3/E3 > E2/E3. In CHC patients infected with genotype 3 and APOE*E3/E3 and E3/E4, cholesterol levels were significantly lower compared with HCV genotype 1 (*P = 0.008; **P = 0.035).

Download figure to PowerPoint

Table 4.   The influence of HCV genotype on lipid levels
VariableHCV genotype 3HCV genotype 1Two-sample t-testMultivariate analysis
nMean ± s.d.nMean ± s.d.P-value95% CIP-value
  1. Mean (mmol/L ± standard deviation) pre-treatment total cholesterol, non-HDL cholesterol, log10 triglycerides and log10 HDL were compared by two-sample t-tests. A multivariate regression analysis was performed to control for age, gender and APOE genotypes (E3/E3, E3/E4 and E2/E3). The multivariate analysis shows that HCV genotype is strongly associated with total cholesterol and non-HDL cholesterol (* P < 0.001).

  2. HDL, high-density lipoprotein.

Total cholesterol974.134 ± 0.903934.704 ± 0.932<0.001*−0.833, −0.3080.015
Non-HDL cholesterol782.714 ± 0.791723.432 ± 0.934<0.001*−0.996, −0.4390.007
Log10 triglycerides630.116 ± 0.269560.175 ± 0.2240.204−0.149, 0.0320.975
Log10 HDL680.124 ± 0.119610.074 ± 0.1470.0330.004, 0.0970.160

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Our data show that non-HDL (i.e. apoB-associated) cholesterol is positively associated with increased odds of achieving an SVR (OR = 2.09, P < 0.05). As this was a retrospective study, we were unable to assess the relative importance of non-HDL-C to other factors known to influence SVR such as insulin resistance17–19, 30, 31 and advanced fibrosis.20 However, similar observations have been reported in previous smaller studies of HCV mono-infected patients.21, 22 A recent study in co-infected patients has also shown independent association with LDL-cholesterol greater that 100 mg/dL and SVR (OR 2.51).23 One large study of US veterans showed that low cholesterol was a significant predictor of decreased likelihood of an SVR in HCV genotype 1 patients, but this was not found in genotype 2 or 3.24 Although these studies support our findings, LDL cholesterol was calculated by the Friedewald equation,29 which may not be valid in CHC because it assumes normal VLDL secretion. However, VLDL assembly and secretion appear to be directly influenced by hepatitis C, possibly by inhibition of microsomal TG transfer protein (MTP).32, 33 Hence, the calculation used in this study may be more appropriate in CHC because we measured all apoB-associated cholesterol including VLDL and LDL.

We have also evaluated some viral and host factors that could alter apoB-associated cholesterol. Our data show that HCV genotype 3 lowers total serum cholesterol and non-HDL-C, but not HDL-C and this is reversed after SVR. This effect of HCV genotype 3 on cholesterol is more profound than functional APOE polymorphisms that normally account for 7% of the variation in total cholesterol in healthy Caucasians. This confirms reports that HCV genotype 3 specifically decreases serum cholesterol.27, 34 However, other studies show that HCV viraemia per se is associated with clinically significant lower cholesterol levels when compared with normal subjects.25

It seems paradoxical that HCV genotype 3 lowers apoB-associated cholesterol levels, but responds better to anti-viral therapy, when higher apoB-associated cholesterol levels are a determinant of SVR. This may be related to the complex interaction of HCV with lipids, involving not only viral assembly and secretion but also HCV entry.4 Evidence suggests that the effect of HCV genotype 3 on serum cholesterol is exerted intracellularly, at the stage of virus maturation in, and secretion from the hepatocyte. HCV genotype 3 inhibits MTP transcription and activity,33 reducing VLDL secretion and promoting liver steatosis. In vitro studies indicate that HCV secretion is dependent on VLDL assembly6 and silencing apoB messenger RNA in infected liver cells causes a 70% reduction in the secretion of both apoB-100 and HCV.35 Recent reports indicate that a liver-specific microRNA (miR), miR-122, is a key regulator not only in cholesterol and fatty acid metabolism in adult liver36 but also in hepatitis C viral replication.37 miR-122 inhibition has been shown to decrease plasma cholesterol levels in a diet-induced obesity mouse model.36 Interestingly, decreased levels of miR-122 in the liver of CHC patients (which would be expected to be associated with lower plasma cholesterol levels) have recently been reported to correlate with a poor response to treatment.38

The effect of higher apo-B-associated cholesterol on improved SVR may be exerted at the stage of HCV entry. Non-HDL-C is an indirect method of quantifying apoB-associated lipoprotein particles. The key constituents of VLDL are apoB-100 (one molecule per VLDL particle), apoE and apoC. These determine receptor-mediated uptake and fate of cholesterol and TG within lipoprotein particles. HCV RNA in plasma has been shown to be associated with both apoB9 and apoE at a buoyant density similar to that of VLDL.10 This complex of HCV and host apolipoproteins and lipids may be termed a ‘lipo-viral particle (LVP)’. Recently it has been demonstrated in the in vitro cell culture infection model (HCVcc) that apoC1 is also a component of HCV and that anti apoC1 neutralized over 75% of infectious HCVcc particles.39 We hypothesize that non-HDL-C in VLDL (containing apoB-100, apoE and apoC1) compete with HCV LVP for hepatocyte entry via shared receptors. It has long been proposed that the LDL receptor may be one of the receptors for HCV.40, 41 Recent in vitro studies show that co-culture of primary hepatocytes with liver sinusoidal endothelial cells not only induces the hepatic expression of the LDL receptor but also HCV-like particle uptake.42 Likewise, manipulation of hepatocytes to increase LDL receptor expression found that HCV viral RNA accumulation increased or decreased in parallel with LDLr mRNA expression and LDL entry, implying that LDLr is involved in HCV entry into hepatocytes.43 An in vivo study has found that LDLr expression on mononuclear cells is significantly associated with HCV viral load, whereas genotype, age, body mass index and fibrosis were not, again implicating LDLr as one of the receptors involved in HCV entry.44 It is known that there is competition between VLDL remnants and chylomicron remnants for LDL receptor-mediated clearance by the liver.45 Both apoB and apoE are ligands for the LDLr and it is thus possible that there is competition in vivo between VLDL remnants and HCV LVP for LDLr-mediated uptake, which has been demonstrated in vitro.9 ApoE also interacts with heparin sulphate-proteoglycans (HSPG) and can be transferred to LDL receptor-related protein (LRP) for internalization and apoC1 interacts with HSPG in vitro,39 so apoC1 on HCV–LVP may use the HSPG–LRP pathway for viral entry. Furthermore, a recent report suggests that SR-B1, which is implicated as another receptor for HCV,46 plays an important role in the metabolism of VLDL remnants,47 so VLDL remnants may also compete with HCV LVP for hepatocyte uptake via SR-B1. Hence, higher non-HDL-C may interfere with the HCV life cycle by impairing HCV entry into hepatocytes. This hypothesis is consistent with the reported effect of IFNα on serum lipid profiles; serum TG levels, largely derived from VLDL, significantly increase following IFNα treatment48 due to a decrease in lipoprotein lipase activity.49

A recent meta-analysis in non-HCV infected individuals confirmed that polymorphisms in the APOE gene have a major influence on serum cholesterol levels.50 Seven per cent of the variation in total cholesterol in healthy Caucasian individuals is related to three different isoforms of the apoE protein.51 The wild-type protein is ε3, and the two variants are ε2 and ε4.52 The ε2 isoform binds poorly to LDLr and is associated with lower cholesterol and apoB.53 Those with the E2 allele have lower cholesterol and those with E4 allele to have higher cholesterol levels than the majority with the wild-type E3 allele.50 This pattern was observed in those HCV patients infected by HCV genotype 1, but not in those infected by HCV genotype 3 where the observed cholesterol levels in those with APOE*E3/E3 genotype and APOE*E3/ E4 genotype were lower than expected. Our observations demonstrate that HCV genotype 3 conveys a greater influence on serum total and non-HDL-C than the host APOE genotype. The difference in total and non-HDL-C levels between HCV genotypes 1 and 3 was greater than that observed between APOE genotypes. This study was underpowered to detect significant differences in treatment response rates between APOE genotypes. However, we did observe a trend for APOE*E2/E3 to be associated with improved IFN response which is intriguing. We have previously shown that APOE*E2 allele was associated with spontaneous viral clearance and hypothesized that the E2 allele may protect against viral persistence via defective binding of HCV to the cellular receptors such as LDLr and SR-B1.12, 46, 47

In summary, our study has shown that higher apoB-associated cholesterol is a significant determinant of SVR in CHC patients receiving anti-viral therapy with pegylated IFNα and ribavirin. We have also analysed the effect of HCV infection and APOE genotypes on serum cholesterol levels and shown that HCV genotype 3 exerts a greater influence over baseline cholesterol level than host APOE genotype. We have confirmed that HCV genotype 3 lowers serum cholesterol more than HCV genotype 1. Our data are consistent with the hypothesis that the effect of higher apo-B-associated cholesterol on improved SVR may be exerted at the stage of HCV entry, possibly due to competition between infectious low density HCV ‘LVPs’ and VLDL remnants for hepatocyte entry via shared receptors including LDLr, HSPG–LRP and SR-B1. This observation may be clinically important and needs to be validated in well designed prospective studies.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank Sister Margaret Hewett and Sister Kerry Baxter for monitoring and supporting patients through treatment. Declaration of personal interests: Margaret Bassendine has received in the past or currently receives grant/research support from Roche, Schering-Plough, Bristol-Myers Squibb and has served as a speaker and an advisory board member for Gilead Sciences and Schering Plough UK. Declaration of funding interests: DA Sheridan was supported by the Medical Research Council, grant number G0502028. DA Price was supported by a Wellcome Trust entry level training fellowship and by Newcastle Healthcare Charity. This study was funded in part by an unrestricted educational grant from Schering Plough Limited (UK).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References