Gut 60:680-687 doi:10.1136/gut.2010.222133
  • Hepatology

Insulin resistance and low-density apolipoprotein B-associated lipoviral particles in hepatitis C virus genotype 1 infection

  1. Margaret F Bassendine1
  1. 1Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UK
  2. 2Institute for Cell and Molecular Biosciences, Newcastle University, Newcastle upon Tyne, UK
  3. 3Division of Diabetes, Endocrinology and Metabolism, Imperial College London, London, UK
  4. 4Department of Clinical Biochemistry, Newcastle upon Tyne Hospitals NHS Foundation Trust, Royal Victoria Infirmary, Newcastle upon Tyne, UK
  1. Correspondence to Professor Margaret F Bassendine, Institute of Cellular Medicine, William Leech Building, The Medical School, Newcastle University, Framlington Place, Newcastle upon Tyne NE2 4HH, UK; m.bassendine{at}
  • Revised 28 July 2010
  • Accepted 24 August 2010
  • Published Online First 12 October 2010


Background The density of hepatitis C virus (HCV) in plasma is heterogeneous but the factors which influence this are poorly understood. Evidence from animal models and cell culture suggest that low-density apolipoprotein B (apoB)-associated HCV lipoviral particles (LVP) are more infectious than high-density HCV.

Objective To measure LVP in patients with chronic hepatitis C genotype 1 (CHC-G1) and examine metabolic determinants of LVP load.

Patients 51 patients with CHC-G1 infection.

Methods Fasting lipid profiles and homeostasis model assessment of insulin resistance (HOMA-IR) were determined in 51 patients with CHC-G1. LVP and non-LVP viral load were measured by real-time PCR of plasma at density <1.07 g/ml and >1.07 g/ml, respectively, following iodixanol density gradient ultracentrifugation. The LVP ratio was calculated using the formula: LVP/(LVP + non-LVP).

Results The mean LVP ratio was 0.241 but varied 25-fold (from 0.029 to 0.74). Univariate analysis showed that the LVP ratio correlated with HOMA-IR (p=0.004) and the triglyceride/high-density lipoprotein cholesterol (TG/HDL-C) ratio (p=0.004), but not with apoB. In multivariate analysis, HOMA-IR was the main determinant of LVP load (log10IU/ml) (R2=16.6%; p=0.037) but the TG/HDL-C ratio was the strongest predictor of the LVP ratio (R2=24.4%; p=0.019). Higher LVP ratios were associated with non-response to antiviral therapy (p=0.037) and with greater liver stiffness (p=0.001).

Conclusion IR and associated dyslipidaemia are the major determinants of low-density apoB-associated LVP in fasting plasma. This provides a possible mechanism to explain why IR is associated with more rapidly progressive liver disease and poorer treatment outcomes.

Significance of this study

What is already known about the subject?

  • Hepatitis C virus (HCV) RNA-containing particles circulate in the serum of patients with chronic hepatitis C (CHC) at a range of densities between 1.03 g/ml and 1.25 g/ml.

  • Experimental data indicate that the low-density HCV particles are infectious in animal models.

  • In vitro silencing apolipoprotein B (apoB) mRNA leads to a reduction in the secretion of HCV.

  • Insulin resistance in CHC genotype 1 (CHC-G1) infection predicts faster progression to fibrosis and cirrhosis and a poorer response to antiviral therapy.

What are the new findings?

  • The proportion of apoB-associated lipoviral particles (LVP ratio) varies at least 25-fold between patients.

  • Metabolic determinants of LVP in fasting CHC-G1 patients include insulin resistance and a high triglyceride/high-density lipoprotein-cholesterol ratio.

  • In patients with CHC-G1 a higher LVP ratio is associated with more fibrosis and a poorer response to antiviral therapy.

How might it impact on clinical practice in the forseeable future?

  • Measurement of low-density apoB-associated LVP provides additional prognostic information and needs to be further evaluated in clinical trials, in particular those using lipid-modulating and insulin-sensitising agents as adjuvant therapies.

The characterisation of hepatitis C virus (HCV) particles from the plasma of infected patients shows a highly heterogeneous population of viruses (buoyant density <1.03–1.25 g/ml) due to physical association of virions with host lipoproteins.1–3 HCV particles found in the low-density fractions of plasma are bound to very-low-density lipoprotein (VLDL), intermediate-density lipoprotein (IDL) and low-density lipoprotein (LDL) particles, each of which contains one molecule of the non-exchangeable structural protein, apolipoprotein B (apoB). Low-density virus particles that contain at least apoB, HCV RNA and core proteins have been termed lipoviral particles (LVP).4 Most LVP from the plasma of an immunodeficient HCV-infected patient without anti-HCV antibodies can be immunoprecipitated with antibodies against apoB.5 LVP may be reliably isolated by density gradient ultracentrifugation in iodixanol which maintains the integrity of the particles.5 This method permits LVP quantitation independent of immune complex formation in immunocompetent patients.4

The development of an efficient HCV cell culture (HCVcc) system has highlighted a critical role of VLDL assembly, maturation and secretion in the HCV life cycle.6 7 In vitro, two types of HCVcc particles are detected, with a small proportion of low-density particles of high infectivity whereas the majority of released particles lack infectivity.8 Silencing apoB messenger RNA in infected cells causes a 70% reduction in the secretion of both apoB-100 and HCV.9 In addition, using HCVcc, Lindenbach et al showed that the virus could establish a long-term infection in chimpanzees and in uPA+/+-SCID mice containing human liver grafts. The virus isolated from infected animals was of lower buoyant density and higher specific infectivity than HCVcc passaged in cell culture alone, providing further evidence that interaction with host triglyceride-rich lipoproteins plays an important role in the life cycle of HCV.10 This observation was in agreement with other in vivo evidence suggesting that lower density HCV particles are more infectious. When HCV particles derived from human plasma were administered to chimpanzees, a low buoyant density fraction was higher in infectivity than a fraction in which particles were predominantly of high density.11 12 Furthermore, phylogenetic analysis of an acutely infected individual revealed that the infecting viral quasispecies most closely matched quasispecies of low-density virus of the previous host.13

In view of the evident importance of lipoprotein metabolism in the HCV life cycle and the apparent higher infectivity of low-density HCV, we used iodixanol density gradient ultracentrifugation and sensitive real-time PCR to quantitate LVP in the fasting plasma of a large cohort of well-characterised patients with chronic hepatitis C genotype 1 (CHC-G1) and examined the host factors and metabolic determinants influencing LVP load.


Study cohort

Consecutive patients with CHC-G1 infection attending the viral hepatitis clinic at the Freeman Hospital, Newcastle upon Tyne were invited to attend the Clinical Research Facility (CRF), Royal Victoria Infirmary for a fasting blood test and were given a patient information leaflet explaining the study. Both treatment-naïve and previous non-responders to combination pegylated interferon-α and ribavirin antiviral therapy were eligible for inclusion. Patients were excluded if they were alcohol dependent, being treated with concurrent lipid-lowering therapy, co-infected with hepatitis B virus, hepatitis delta virus or HIV or had poor venous access due to injecting drug use. Non-responders were also invited to participate in a separate ongoing lipid-modulating intervention study (REC number-07/S0501/21) which, in addition to the above, excluded participants with a body mass index (BMI) ≥30 kg/m2. Fifty-one patients with CHC-G1 aged >18 years who provided written informed consent attended the CRF after a 12 h fast; 18 were non-responders with a BMI <30 kg/m2. The HCV cohort attending our viral hepatitis clinic comprises approximately 45% patients with CHC-G1, and the study participants were representative of both the regional and national HCV-infected population.14 15

Clinical and laboratory assessment

Each subject attending the CRF were reassessed for past alcohol intake and medication history and the following data were collected: sex, age, weight, height, waist and hip circumference. A serum sample was collected for lipid analysis. Total cholesterol, triglyceride (TG), HDL-cholesterol (HDL-C), non-esterified fatty acids and fasting glucose were measured by standard automated enzymatic methods using an Olympus AU 640 analyser (Olympus, Watford, UK). LDL cholesterol (LDL-C) was calculated using the Friedewald equation and non-HDL cholesterol (non-HDL-C) was obtained by subtraction of HDL-C from total cholesterol. Fasting insulin was measured by immunoassay (ELISA; Dako UK Ltd, Ely, UK) and assessment of insulin resistance (IR) was performed by calculation of Homeostasis Model Assessments of Insulin Resistance (HOMA-IR) and Quantitative Insulin Sensitivity Check Index (QUICKI) from the serum sample using the formulae: HOMA-IR=(fasting glucose (mmol/l)×insulin (mU/l)/22.5; QUICKI=1/(fasting glucose (log10 mmol/l)+fasting insulin (log10 mmol/l)). Apolipoprotein A-I and apoB were measured on each sample by automated rate nephelometric methods (BNII, Dade Behring Marburg GmbH, Marburg, Germany).

Liver stiffness was measured within 3 months of the patient's visit by the non-invasive method of ultrasound transient elastography using a Fibroscan with a standard probe (Echosens, Paris, France).16 HCV genotype was previously determined by the Health Protection Agency North East Laboratories, Newcastle upon Tyne using the line-probe assay (INNOLiPA HCV; Innogenetics, Ghent, Belgium).17

Venous blood and plasma preparation

In patients and healthy volunteers, 12 h fasting blood samples were collected in EDTA vacutainer tubes (BD Biosciences Ltd, Oxford, UK) and incubated at 37°C. The plasma was separated by centrifugation at 3000 rpm for 10 min at 37°C in a Rotanta 460R Hettich centrifuge (DJB Labcare, Buckinghamshire, UK). The blood was handled as close to body temperature as possible to avoid the formation of a cryoglobulin precipitate.18 Complete protease inhibitor cocktail (Roche Products Ltd, Welwyn Garden City, UK) was added to the plasma to prevent protein degradation.

Iodixanol density gradient ultracentrifugation

Plasma (0.5 ml) was added to 9.5 ml 12.5% iodixanol solution (2.1 ml iodixanol (60%) (Optiprep, Axis-Shield, Kimbolton, UK), 200 μl 100 mM EDTA, pH 8.0, 200 μl 0.5 M Tris-HCl pH 8.0, 7.15 ml 0.25 M sucrose, 25 μl 2 M MgSO4, 25 μl 2 M MgCl2) in polycarbonate centrifuge tubes (Beckman Coulter, High Wycombe, UK) and inverted several times to mix thoroughly before centrifuging at 50 000 rpm for 24 h at 4°C in a type Ti50 rotor and a L8-70M ultracentrifuge (Beckman Coulter). Gradients were harvested according to whether they were from the LVP assay development cohort (seven healthy volunteers and nine patients with CHC) or the validation cohort (51 patients with CHC-G1 infection). For the development cohort the gradients were harvested from the top by collecting 20×500 μl fractions and the density of each fraction was measured using a digital refractometer (Atago, Washington, USA). On the basis of the results obtained from the development cohort, the validation cohort gradients were harvested into two fractions: a top 3.5 ml low-density (<1.07 g/ml) fraction (LVP) and a bottom 6.5 ml high-density (>1.07 g/ml) fraction (non-LVP). The densities between these two fractions were measured. All fractions were pulse-vortexed to ensure thorough mixing.

SDS-PAGE and apoB western blotting

Iodixanol fraction samples were prepared by boiling in Laemmli buffer (4% sodium dodecyl sulphate (SDS), 20% glycerol, 10% 2-mercaptoethanol, 0.004% bromophenol blue, 0.125 M Tris HCl) for 10 min. Proteins were separated in 3–18% gradient gels by SDS-polyacrylamide gel electrophoresis (PAGE) on a Bio-Rad Protean II system. ApoB was detected by western blotting after transfer of the resolved proteins to a Hybond polyvinylidene difluoride membrane (GE Healthcare Life Sciences, Little Chalfont, UK) using an enhanced chemiluminescence (ECL) detection kit (GE Healthcare Life Sciences) and polyclonal anti-apoB antibody (Dako).19

HCV RNA extraction and quantitation by real-time PCR

A plasma sample was collected for viral analysis at the same time as the serum samples for lipid analysis. HCV RNA was extracted from whole and fractionated plasma by QIAamp MinElute Virus Spin Kit according to the manufacturer's protocol (Qiagen, Crawley, UK). HCV RNA was extracted from 200 μl of EDTA plasma or the iodixanol LVP/non-LVP fraction and eluted into 100 μl of buffer AVE (Qiagen). Extracted HCV RNA was quantitated by two-step real-time PCR for HCV as described previously.19 Reverse transcription was performed using the NCR-3 primer and AMV reverse transcriptase (Promega, Southampton, UK). The HCV positive-strand assay was calibrated against the WHO 3rd international standard for HCV RNA (NIBSC, Potters Bar, UK). Real-time PCR was conducted using an ABI Prism 7000 with Taqman Universal PCR Master Mix (Applied Biosystems, Warrington, UK). Determinations of duplicate tests were averaged. LVP is defined as the HCV RNA measured in the gradient with a density of <1.07 g/ml and non-LVP is the HCV RNA measured in the gradient with a density of >1.07 g/ml. The LVP ratio was calculated using the formula: LVP/(LVP + non-LVP).

Statistical analyses

All statistical analyses were performed using Minitab 15 software (Minitab Inc, State College, USA). The distribution of continuous data was assessed by normality tests. Age, waist circumference, BMI, total cholesterol, LDL-C and non-HDL-C conformed to a normal distribution. The remaining metabolic and viral variables showed positively skewed non-parametric distributions and were therefore either log10-transformed to parametric distributions before hypothesis tests were applied or non-parametric tests were used. Associations between metabolic parameters and viral load, LVP (log10 IU/ml) and LVP ratio were assessed by univariate linear regression analysis. Pearson correlation was used to test for the association between parametric viral and metabolic variables. These analyses yielded regression coefficients (r) and a probability value (p) from testing against the null hypothesis of no association. Spearman rank correlation analysis was used to test for the association between non-parametric viral and metabolic variables. p Values <0.05 were considered to be significant. From the univariate analysis, factors that correlated with p<0.05 were used in a stepwise multivariate linear regression analysis to determine which metabolic factors independently predict LVP (log10 IU/ml) and LVP ratio. Those factors that had p<0.05 after multivariate analysis were considered significant independent predictors of the variability of LVP (log10 IU/ml) and LVP ratio, with the degree of association indicated by the R2 value. LVP data from patients with CHC-G1 were divided into those with a low LVP ratio (<median) and those with a high LVP ratio (≥median). For parametric data, the F test was applied to test the assumption of equal variances and two-sample t tests were used to compare clinical and metabolic parameters between those with a high/low LVP ratio. For non-parametric data, Kruskal–Wallis tests were applied to compare groups.The χ2 test was used to test the association between high/low LVP ratio with early antiviral treatment responses, metabolic syndrome and liver stiffness.


Density distribution of plasma apoB on iodixanol density gradients

Plasma from a healthy volunteer was added to a range of isotonic iodixanol gradients with increasing concentrations of iodixanol: 6.25%, 12.5%, 18.75%, 25%, 31.25%, 37.5%, 43.75% and 50% and apoB was detected in the fractionated gradients by SDS-PAGE and western blotting. The optimum concentration of iodixanol was 12.5% as this formed the most linear density gradient. Increasing the concentration of iodixanol reduced the linearity of the density gradient and concentrated the apoB-containing fractions to the top of the gradient (data not shown).

Development cohort

Plasma from seven healthy volunteers and nine patients with CHC were fractionated by iodixanol density ultracentrifugation: 20×0.5 ml fractions were collected. ApoB was not detected in the healthy volunteers in fractions with a density >1.062 g/ml and was not detected in patients with CHC with a density >1.07 g/ml (see table 1 in online supplement). This density cut-off value in a 12.5% iodixanol gradient was subsequently used to define LVP in our assay.

Validation cohort

Iodixanol gradients were fractionated into the LVP fraction at a density of <1.07 g/ml (top 3.5 ml of the gradient) and the non-LVP fraction at a density of >1.07 g/ml (remaining 6.5 ml of the gradient).

Clinical and metabolic characteristics

The physical and metabolic characteristics of the 51 patients with CHC-G1 infection are shown in table 1. Forty-eight patients were Caucasian and the remaining three patients were Chinese (n=1), sub-Saharan African (n=1) and Eastern Mediterranean/Middle Eastern (n=1). BMI was normal (<25 kg/m2) in 24 patients (47%), overweight (25–30 kg/m2) in 20 (39%) and obese (>30 kg/m2) in 7 (14%) patients. According to the criteria of the International Diabetes Federation, 12 patients had metabolic syndrome.21 Of these, two patients were receiving treatment for type 2 diabetes mellitus.

Table 1

Clinical and metabolic characteristics of patients with hepatitis C virus genotype 1 (HCV-G1) infection

Viral parameters

Viral load, LVP and non-LVP load were measured in all 51 patients with CHC-G1 infection. The results of viral load and LVP load are summarised in figure 1A,B. Viral load and LVP load produced similar distributions. The majority (65%) of viral loads were >5.90 log10 IU/ml. The LVP ratio was calculated and the results are shown in figure 1C. The LVP ratio varied widely between patients ranging from 0.029 to 0.74 (mean 0.241; median 0.177).

Figure 1

Population distribution of viral parameters. (A) Total plasma hepatitis C virus (HCV) RNA viral load (log10 IU/ml). (B) Lipoviral particle (LVP) load (log10 IU/ml) found at a density of <1.07 g/ml. (C) LVP ratio (fraction of LVP load to LVP + non-LVP load (log10 IU/ml)).

Insulin resistance

As expected,22 HOMA-IR was strongly correlated with waist circumference (r=0.641, p<0.0001), BMI (r=0.526, p<0.0001) and positively correlated with age (r=0.324, p=0.020). There was no correlation between HOMA-IR and HCV viral load (figure 2A, r=0.078, p=0.585). Liver stiffness correlated with HOMA-IR (r=0.381, p=0.007).

Figure 2

Relationship between total viral load, lipoviral particle (LVP) load and LVP ratio with HOMA-IR. Open circles indicate non-responders, closed circles indicate those with an early virological response (EVR), open triangles indicate those not treated. Correlation between HOMA-IR and (A) viral load (log10 IU/ml) (r=0.078; p=0.585), (B) LVP load (log10IU/ml) (r=0.383; p=0.005) and (C) LVP ratio (r=0.397; p=0.004).

Metabolic determinants of LVP

Univariate analysis showed that LVP load (log10 IU/ml) in fasted patients correlated positively with glucose (table 2; r=0.307, p=0.028), insulin (table 2; r=0.393, p=0.005) and HOMA-IR (table 2; r=0.383, p=0.005, figure 2B). Univariate analysis showed that the LVP ratio correlated positively with glucose (table 2; r=0.311, p=0.026), insulin (r=0.393, p=0.005) and HOMA-IR (figure 2C; r=0.397, p=0.004). The LVP ratio also correlated positively with TG levels (figure 3A; r=0.320, p=0.022), the TG/HDL-C ratio (figure 3B; r=0.392, p=0.004) and with liver stiffness (table 2; r=0.419, p=0.003). Multivariate regression analysis showed that HOMA-IR was the only significant independent determinant of LVP load (log10 IU/ml) (R2=16.6%; p=0.037), whereas TG/HDL-C ratio was the only independent predictor of LVP ratio (R2=24.4%; p=0.019).

Table 2

Univariate analysis of viral load and LVP in CHC-G1 with host and metabolic parameters in patients with CHC-G1

Figure 3

Relationship between lipoviral particle (LVP) ratio with triglyceride levels and the triglyceride/HDL-C ratio. Open circles indicate non-responders, closed circles indicate those with an early virological response (EVR), open triangles indicate those not treated. Correlations between LVP ratio with (A) triglyceride (r=0.320; p=0.022) and (B) triglyceride/HDL-C (r=0.392; p=0.004).

The comparison of clinical and metabolic variables in patients with CHC-G1 with a low LVP ratio (defined as below the median value of 0.177; n=25) and those with a high LVP ratio (defined as above the median value of 0.177; n=26) is shown in table 3. Patients with a low LVP ratio had higher LDL-C (p=0.037) while those with a high LVP ratio had a significantly higher fasting plasma glucose (p=0.044), insulin (p=0.005) and HOMA-IR (p=0.008). The latter patients also had a higher TG level (p=0.015), lower HDL-C (p=0.015) and hence higher TG/HDL-C ratio (p=0.003). In addition, patients with CHC-G1 with a high LVP ratio had a larger waist circumference (p=0.037) and greater liver stiffness (p=0.001). Further analysis of the LVP ratio shows that patients with a high LVP ratio were more likely to have a liver stiffness measurement of >13 kPa (χ2=5.95; p=0.015), consistent with advanced fibrosis.

Table 3

Relationship between LVP ratio with metabolic variables, liver stiffness and EVR to antiviral therapy

LVP and treatment response

Data on the early virological response (EVR) to antiviral treatment with pegylated interferon-α and ribavirin was available for 42 patients with CHC-G1. Of these, 18 (43%) were non-responders with the remaining 24 (57%) having achieved a complete or partial EVR (HCV RNA negative or >2 log reduction in HCV RNA at 12 weeks). Eleven patients on treatment had metabolic syndrome and 10 of these patients were non-responders. The median LVP ratio in the non-responder group was significantly higher than in the EVR group (0.338 vs 0.201, p=0.031, figure 4). Conversely, patients with CHC-G1 with a low LVP ratio were more likely to have an EVR (p=0.037, table 3).

Figure 4

Relationship between lipoviral particle (LVP) ratio with early antiviral treatment outcome. LVP ratio and early antiviral treatment outcomes (EVR to non-responders, p=0.031). The box height represents the interquartile range (Q1–Q3), the line within the box is the median value, the lower whisker represents Q1–1.5 (Q3–Q1) and the upper whisker represents Q3+1.5 (Q3–Q1).


This study shows that IR is an independent predictor of LVP load and that the TG/HDL-C ratio is an independent predictor of the LVP ratio in patients with CHC-G1 infection. The LVP ratio is a relative measure of the density distribution of HCV in the plasma. This is the first study to measure LVP in a well-characterised cohort of patients with CHC using a method based on fractionation of iodixanol density gradients which maintains the integrity of these particles. This allows the use of sensitive real-time PCR methods for accurate determination of HCV RNA associated with low-density (<1.07 g/ml) apoB-containing lipoproteins (LVP load) and the proportion of total HCV RNA found in plasma as LVP (LVP ratio). Our study shows that the LVP ratio varied widely in fasting patients, ranging from 0.029 to 0.740 and that, on average, LVP contribute approximately a quarter (∼24%) of total viral load. Using univariate analysis, we found a positive correlation of LVP load with HOMA-IR. The LVP ratio correlated significantly with HOMA-IR, TG/HDL-C ratio and TG levels, but there was no association between total viral loads and host clinical or metabolic parameters. Using multivariate analysis, we showed that HOMA-IR accounts for approximately 16% of the variability of the LVP load. Interestingly, in the subgroup of 42 patients with CHC-G1 in the present study for whom data on response to antiviral treatment was available, the LVP ratio was significantly higher in non-responders than in those with a complete or partial EVR, consistent with the LVP fraction contributing to persistent infection. Furthermore, patients with CHC-G1 with a higher LVP ratio had more advanced fibrosis (liver stiffness >13 kPa).

IR in CHC is clinically relevant because it not only predicts faster progression to cirrhosis but also a poorer response to antiviral therapy. In subjects with CHC-G1, IR and overt diabetes mellitus are major determinants of advanced fibrosis.23–27 IR is also a risk factor for progression of fibrosis in HCV-infected patients following liver transplantation.28 IR is an independent predictor of a sustained virological response (SVR) in patients with CHC receiving pegylated interferon-α/ribavirin.22 29–32 Our results show that IR is the most important determinant of absolute LVP load in CHC-G1 infection, while the strongest predictor of LVP ratio was found to be the TG/HDL-C ratio which is characteristically increased in IR.

IR and TG levels have previously been associated with high viral load in some large cohort studies. In a study of 500 patients with CHC, those with IR had a high serum viral load (defined as a serum HCV RNA level >600 000 IU/ml) more frequently than those without IR (55.3% vs 42.3%, p<0.009).33 In another study of >500 patients with CHC, both HOMA-IR and TG levels were associated with higher HCV viral loads (p<0.05).34 In a recent study of >200 patients with CHC in a single US centre where patients with co-existing metabolic syndrome are more common, hypertriglyceridaemia correlated with higher viral loads (p=0.042).35

Epidemiological data indicate a strong risk for the development of IR and, ultimately, type 2 diabetes mellitus in patients with CHC infection.36–38 In a recent 5-year prospective follow-up study, HCV infection per se and genotype 1 were independent risk factors for the development of IR.39 IR in CHC is most strongly associated with genotypes 1 and 4 infection.33 HCV may promote IR through genotype-specific molecular mechanisms,40 41 although recent evidence suggests that IR in CHC is mainly in muscle and not liver.42 Furthermore, recent studies demonstrate that effective clearance of HCV with antiviral therapy correlates with an improvement in IR and reduced incidence of onset of type 2 diabetes mellitus.43 44

In normal individuals without CHC the liver secretes 1018 VLDL particles daily, with smaller VLDL2 being constitutively secreted and production of larger TG-rich VLDL1 being regulated by insulin in normoglycaemic subjects.45 46 IR results in overproduction of large low-density TG-rich VLDL1 particles.47 48 Kinetic studies have shown that <20% of large TG-rich VLDL1 are converted to LDL; most are cleared directly from the plasma.49 50 The positive correlation of the LVP ratio with the TG/HDL-C ratio, even in the fasting state, implies that HCV preferentially associates with triglyceride-rich lipoproteins which, in the fasting state, will predominantly be large VLDL—that is, VLDL1, not chylomicrons. Fatty acids for the biosynthesis of VLDL1 are derived from TG stored in lipid droplets,51 and HCV uses lipid droplets for the production of infectious virions.52 We have also recently shown in vivo that very-low-density HCV (<1.025 g/ml) surges 26-fold after a high fat meal.53 Collectively, our in vivo studies are consistent with LVP in the plasma preferentially associating with large TG-rich VLDL. Association between HCV and VLDL1 as LVP may enable these putative infectious particles to escape antibody neutralisation54 and be rapidly cleared by the liver via the apoE remnant pathway.55

Our data suggest that the measurement of LVP and calculation of the LVP ratio is clinically important and provides additional information when measured in combination with conventional HCV viral load. For example, incorporation of LVP measurement in therapeutic trials using insulin-sensitising or lipid-modulating drugs as adjuvant therapies31 56–58 may help identify patients who may benefit from this approach. The development of a simple and inexpensive method to measure LVP would facilitate the analysis of LVP in appropriately-powered larger prospective studies and would provide confirmation as to whether LVP is clinically useful in predicting patient responsiveness to antiviral regimens.

In summary, we report the first accurate measurements of LVP fraction of total hepatitis C viral load in a well-characterised cohort of patients with CHC-G1. Our results show that the LVP ratio varies at least 25-fold between patients and is significantly associated with metabolic parameters and important clinical outcomes. In particular, a higher LVP ratio is strongly associated with IR and higher TG levels, both components of the metabolic syndrome. Higher LVP ratios were also found in non-responders to antiviral treatment and those with more advanced fibrosis. Our in vivo study offers further insight into the life cycle of HCV and suggests that HCV preferentially associates with TG-rich VLDL1 which is modulated by IR. Our data support the hypothesis that promotion of IR by HCV infection may drive the production of more LVP, providing a possible explanation of why patients with CHC-G1 with IR have a worse prognosis.


The authors thank Fiona I Fenwick for her smooth running of the laboratories and are grateful to the hepatitis C patients for their generous donation of time and inconvenience.


  • Funding Supported by the Medical Research Council (G0502028) and the Newcastle upon Tyne Healthcare Charity. The funders had no role in the study design, data collection and analysis. HCT and SDT-R were supported by the NIHR Biomedical Facility at Imperial College London.

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the Northumberland Research ethics committee (REC number-07/H0902/45).

  • Provenance and peer review Not commissioned; externally peer reviewed.


Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of Gut.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.