Evidence from case control studies

Table 1 summarises the principal findings of cross sectional studies that assessed the associations between cardiovascular alterations and HCV infection. On the basis of the design and findings of the studies, we also categorised the evidence as ‘association’ or ‘strongly suggesting a causal effect.’ Specifically, causality is suggested when: (1) there is an improvement in morbidity/mortality after viral clearance and/or (2) there is a dose effect between HCV and clinical outcome.

Evidence from cohort and prospective studies

The strongest evidence of a potential link between HCV infection and both cardiovascular events and mortality emerged from studies evaluating the natural history of patients with CHC.

The first study showing that HCV infection may increase cardiovascular mortality was conducted by Guiltinan et al in 2008. The authors performed a retrospective cohort study on 10 259 anti-HCV positive and 10 259 age and gender matched anti-HCV negative US blood donors, followed from 1991 to 2002, and with a mean follow-up of 7.7 years.28 Interestingly, they observed that the HCV positive group had excessive mortality due not only to higher liver, drug, alcohol and trauma/suicide related events, as expected, but also to an increase in cardiovascular mortality (HR=2.21).28 Similar data were recently found in a community based long term prospective study in a cohort of 23 820 Taiwanese adults, enrolled from 1991 to 1992, and with follow-up data up to 2008.29 As in the US study, the authors observed that anti-HCV positives had a higher mortality from both hepatic and extrahepatic diseases compared with anti-HCV negatives, showing an adjusted HR of 1.50 for circulatory diseases.29 Interestingly, they also highlighted that this increase in mortality from circulatory diseases was maintained in anti-HCV patients with detectable HCV RNA, but not in those with undetectable HCV RNA.29

Similar data were reported when considering incidental stroke or stroke related mortality. Specifically, two recent population based cohort studies in Taiwan not only showed that HCV infection was an independent predictor of stroke occurrence after adjustment for cardiometabolic risk factors,30 but also that the risk of stroke was significantly reduced in HCV patients previously treated with an interferon based regimen.31 However, it should be emphasised that the authors did not report data on the outcome of antiviral therapy.31 Consonant with these findings, Lee et al confirmed anti-HCV positivity as an independent risk factor for lethal cerebrovascular events, and found that adjusted HRs progressively increased from anti-HCV positive/HCV RNA undetectable patients to patients with a low viral load and further to those who were highly viraemic, compared with anti-HCV negative patients.32 A recently published study from Taiwan33 investigated the occurrence of end stage renal disease, ischaemic stroke and acute coronary syndrome in 1411 HCV infected diabetic patients who received pegylated interferon plus ribavirin (treated cohort), 1411 untreated controls matched by propensity scores (untreated cohort) and 5644 diabetic patients without HCV infection (uninfected cohort). Interestingly, the authors concluded that antiviral treatment for HCV infection was associated with improved renal and cardiovascular outcomes in diabetic patients. However, lack of data on the outcome of the antiviral therapy could affect the interpretation of these results.

Finally, in a US observational cohort of more than 150 000 subjects, Butt et al showed a significantly higher risk of incidental CAD in HCV infected subjects compared with those who were uninfected, despite the better metabolic profile reported in the first group.34 However, these data were not confirmed by Forde et al in a large cohort of adults followed in general practices in the UK35 The authors reported that there was no difference in the incidence rates of myocardial infarction between HCV infected and uninfected patients during a median follow-up of 3.2 years.35 Similar to the case control studies, differences among studied populations and the methodologies used to assess viral and clinical parameters may explain the observed discrepancies.

Table 2 summarises the principal findings of the cohort studies that assessed the associations between cardiovascular outcomes and HCV infection. By using the same criteria quoted above, we discriminated the available evidence as simple ‘association’ or ‘strongly suggesting a causal effect’.

Potential mechanisms linking HCV infection to cardiovascular alterations

Obesity, IR, type 2 diabetes, hepatic steatosis, arterial hypertension, smoking and dyslipidaemia are frequent features in the general population, representing well known and clinically relevant risk factors for cardiovascular morbidity and mortality. Interestingly, some of these factors are reported most frequently in HCV infected compared with uninfected patients, due to the ability of HCV to directly and indirectly interfere with glucose and lipid metabolism, resulting in a high prevalence of IR, steatosis and type 2 diabetes.1–5 ,8 These features are of particular interest considering that insulin is able to regulate glucose, lipid and insulin metabolism, to modulate basic cellular function (cell proliferation and apoptosis), and to control nitric oxide production.36 Accordingly, IR is associated with impaired insulin signalling in metabolic tissues (pancreas, skeletal muscle, liver, adipose tissue), leading to hyperglycaemia and dyslipidaemia; in the vasculature, leading to vessel compliance impairment and hypertension; and in immune cells, leading to inflammation, all of which are capable of prompting the development of atherosclerosis.36 The development of type 2 diabetes and, therefore, hyperglycaemia, further increases the cardiovascular risk via different mechanisms, such as macrophage lipid uptake, leading to: foam cell formation; endothelial dysfunction; increased platelet activity; increased proteolytic activity; glycation of the extracellular matrix; stimulation and proliferation of smooth muscle cells; and increased inflammatory activity, all of which can increase the prevalence of the risk for atherosclerosis (ie, unstable inflammatory and lipid rich plaques).37

In addition, a recent study of chronically infected, cleared and never infected patients was the first to report a significant increase in mesenteric fat in HCV patients compared with controls.6 These data, together with recent evidence3 of an association between HCV viral load and visceral adiposity index, an index expression of both fat distribution and function, suggest a potential role of HCV in prompting the accumulation of visceral fat, a key cardiovascular risk factor that acts indirectly via induction of IR and directly via release of toxic free fatty acids and proinflammatory cytokines.38

Despite the reported high prevalence of these metabolic alterations in CHC, different studies have shown that HCV infected patients are characterised by a favourable lipoprotein profile: low levels of low density lipoprotein and very low density lipoprotein cholesterol. This is due to the interference of HCV with the lipid machinery, and may reverse after viral eradication.6 This good lipid profile, together with the higher prevalence of IR and diabetes, could therefore explain the similar prevalence of metabolic syndrome in CHC patients compared with non-infected controls,8 and could also suggest adopting a modified definition of metabolic syndrome for CHC patients.39

In addition to the above mentioned conventional cardiovascular risk factors, other metabolic alterations recently reported in HCV infected patients may contribute to the high cardiovascular morbidity and mortality in this setting. Some studies have found that the biochemical profile of patients with CHC of different HCV genotypes is characterised by lower than normal serum 25(OH)D levels and a high prevalence of vitamin D deficiency,40 ,41 and that G1 CHC patients self-reported a high industrial fructose intake.42 Interestingly, both of these features (ie, vitamin D deficiency and high industrial fructose consumption) are reported to be new and growing cardiovascular risk factors.43 ,44 In addition, the link between HCV infection and cardiovascular risk could be partially explained by genetic background. In fact, some gene variants associated with metabolic disturbances and/or with a higher spectrum of liver damage in CHC have also been associated with a higher cardiovascular risk, while others may be worth testing, considering the high rate of missing heritability reported in genome wide studies investigating the genetic basis of cardiovascular diseases.45 A few clinical studies have found that higher homocysteine levels and/or methylenetetrahydrofolate reductase (MTHFR) status are linked to HCV infection and its histological severity in terms of both steatosis and fibrosis,46–48 with hyperhomocysteinaemia and MTHFR status also being well known cardiovascular risk factors.49 Similarly, tumour necrosis factor α (TNFα) polymorphisms associated with higher TNFα activity and severity of liver disease in CHC,50 as well as peroxisome proliferator activated receptor gamma polymorphisms associated with steatosis in CHC,51 were also linked to an increased cardiovascular risk.52–55 Recent studies have identified variants in IL28B and PNPLA3 genes as associated with metabolic alterations and steatosis, and fibrosis severity in both CHC and NAFLD patients56–58 and, for PNPLA3 only, with atherosclerosis in NAFLD patients.59 Further studies are needed to assess the direct, or indirect (via promotion of metabolic disturbances), role of these gene variants on cardiovascular risk in CHC patients.

Data in the literature also suggest that HCV could prompt cardiovascular alterations via direct and indirect mechanism(s) outside the above mentioned metabolic alterations. The presence of indirect mechanisms is based on the assumption that infectious agents are recognised as being involved, via inflammation, in the pathogenesis of atherosclerosis. In HCV chronically infected patients, the pathogenesis of liver inflammation is principally immune mediated because the immune system is inadequate to eradicate the virus, leading to chronic inflammation via both innate natural killer mediated and adaptive T helper 1 mediated responses.60 This chronic inflammation is also enhanced by liver iron deposition, a feature frequently reported and associated with severity of liver histology in CHC,61 but also linked to a higher cardiovascular risk.62 There is some evidence that CHC patients have lower adiponectin and higher TNFα and interleukin 6 serum levels,63 ,64 with adiponectin inversely, and TNFα and interleukin 6 directly, related to cardiovascular alterations, compared with controls. We recently found that the risk of atherosclerosis is greater in CHC patients compared with those without severe liver fibrosis.13 Similarly, another study showed that the severity of myocardial injury was directly associated with liver necroinflammatory activity.27 These data suggest that the proinflammatory and profibrogenic environment prompting fibrogenesis in the liver of HCV infected patients may also be systemically activated, enhancing the development of atherosclerotic lesions. Furthermore, hepatic steatosis and visceral obesity may actively promote the development of atherosclerosis, independent of their well known association with classic cardiovascular risk factors, through a general enhancement of a cluster of pro-atherogenic factors, including reactive species of oxygen, inflammatory cytokines, homocysteinaemia, hypoadiponectinaemia, IR and features of the metabolic syndrome.12 However, when considering steatosis, it is worth discriminating between the viral steatosis of HCV genotype 3 and the metabolic steatosis of non-3 HCV genotypes. In fact, one could hypothesise that viral steatosis is not associated with IR, does not increase IR5 and does not affect the severity of liver disease,4 and that it does not participate in the increased cardiovascular risk observed in HCV infected patients. In addition, the presence of viral direct mechanisms leading to atherosclerosis is also suggested by some studies, which have reported a direct link between HCV viral load and vascular alterations. Adinolfi and colleagues12 found that CHC patients with carotid atherosclerosis had higher serum HCV RNA levels than those without. Similar results were obtained when the analysis was done for high or low viral load. These findings were further strengthened by a Japanese study that found a direct link between HCV RNA viral load and myocardial injury, assessed by myocardial scintigraphy.27 Boddi et al detected positive strand HCV RNA in carotid plaques from anti-HCV positive patients but not in plaques of anti-HCV negative patients.65 These interesting data suggest an active local HCV infection in vascular tissue65 although caution in interpreting these results is warranted given the lack of corroborative studies. Consequently, it is possible to hypothesise that a higher prevalence of IR, hyperglycaemia, liver steatosis, visceral obesity and perhaps vitamin D deficiency, together with industrial fructose intake and virus/inflammatory induced alterations, could result in a higher cardiovascular risk compared with uninfected populations, despite the virus induced good lipid profile (figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Potential mechanisms linking HCV infection to cardiovascular alterations. HCV is able to directly and indirectly interfere with glucose and lipid metabolism, resulting in a high prevalence of insulin resistance (IR), hyperglycaemia, liver steatosis, visceral obesity and vitamin D deficiency. These metabolic factors, together with genetic background (including IL28B and PNPLA3 polymorphisms) and lifestyle, act synergistically in increasing the cardiovascular risk of HCV infected patients. In addition, HCV itself could induce cardiovascular alterations via direct and indirect mechanism(s). On the one hand, proinflammatory and profibrogenic environments prompting fibrogenesis in the liver could also be systemically activated, thus enhancing the development of atherosclerotic lesions. On the other hand, the presence of viral direct mechanisms leading to atherosclerosis—that is, an active local infection in the vascular tissue—could be hypothesised. Access the article online to view this figure in colour.