| |
Fatty Liver (Steatosis) May Cause Fibrosis & Liver Cancer
|
| |
| |
Hepatitis C and steatosis: a reappraisal
Journal of Viral Hepatitis
Volume 13 Page 73 - February 2006
Lonardo1,2, P. Loria2, L. E. Adinolfi3, N. Carulli2 and G. Ruggiero3
1Unita Operativa di Medicina Interna e Gastroenterologia, Nuovo Ospedale Civile-Estense di Baggiovara, Modena; 2Dipartimento di Medicina Interna, Universita degli Studi di Modena e Reggio Emilia, Modena and 3Cattedra di Medicina Interna, Universita degli Studi di Napoli, Napoli, Italy
Summary.
The overall prevalence of steatosis in patients with Hepatitis C virus (HCV) chronic infection is 55.5% (range 34.8-81.2%). This is a two to threefold increase compared with the prevalence of steatosis in chronic hepatitides because of other aetiologies and of the figures expected on the grounds of a steatosis-HCV chance association. HCV genotype 3 (HCV-3) has specific epidemiological features; furthermore, as compared with HCV-non-3 genotypes, it is associated with a higher prevalence (74.1%vs 47.9%, P < 0.01) and with more severe grades of steatosis (prevalence of grade 3 steatosis 29.6 vs 5.5 P < 0.01). Host and viral factors play a role, although to a variable extent, in the pathogenesis of HCV-3 and non-3 steatosis. HCV load and body mass index are associated with steatosis in HCV-3 and in HCV-non-3 patients respectively. Serum cholesterol levels and liver steatosis at baseline follow an inverse relationship in HCV infection. As hypocholesterolaemia corrects only in those sustained responders to antiviral treatment both in genotype 3 and in non-3 genotypes, the occurrence of a virally induced, acquired and reversible hypobetalipoproteinaemia seems plausible. Steatosis affects the natural course of HCV infection: it is associated with fibrosis, a possible mediator of increased risk to develop type 2 diabetes, it impairs the response to antiviral treatment in HCV-3 patients and might constitute a risk factor for the development of hepatocellular carcinoma. These observations indicate the need to evaluate the efficacy of combined antiviral and 'metabolic' approaches vs standard antiviral regimes in patients with steatosis and HCV chronic infection.
Which are the risk factors for steatosis in HCV infection?
Various papers have addressed the issue of which risk factors are associated with steatosis in HCV infection. The results allow us to classify such factors to 'host-related'- including BMI/central adiposity - [2,5,15-17,22,23,25,27,30,31,34,36,40,42] and 'viral'- including genotype 3 [2,5,23,25,29-31,34-36,39,42]. It is of interest that a number of authors report exactly the same two risk factors [2,5,23,25,27,30,31,34,36,42] while few report only one of the two [15-17,22,29,35,39,40]. This impressive body of literature strongly suggests that steatosis should be studied separately in HCV genotype 3 ('viral steatosis') from non-3 infections ('metabolic steatosis'). By following such an approach, it has been convincingly shown that hepatic/blood viral load is the factor associated with steatosis in genotype 3-infected patients and BMI is associated to steatosis in 'non-3' patients (Table 3). Clinical experience suggests that serum cholesterol levels are low in some HCV-infected patients and that such an abnormality is corrected by effective antiviral treatment. Therefore, the following question should be addressed.
What are the consequences of steatosis in patients with HCV?
Steatosis observed in HCV-positive chronic hepatitis has the potential to affect the natural course of the infection through different routes (Fig. 1].
Fibrosis
Fibrosis - Several Authors have found that steatosis is associated with fibrosis in cross-sectional studies [2,5,22,23,25,34,36,40,45-47]. Two studies have reported that alcohol interacts with steatosis in the development of fibrosis [23,47]. Additional risk factors for fibrosis have been reported to be: age of the patients/disease duration [5,40], necroinflammation [25,45,47] and noninsulin dependent diabetes mellitus (NIDDM) [40]. Perhaps the best evidence for a link between steatosis and fibrosis comes from paired biopsies and interventional studies. Westin et al. [42] retrospectively analysed 98 patients who underwent dual liver biopsies before antiviral treatment. The median follow-up time was 5.8 years. The authors found that progressive fibrosis was more prevalent in patients whose initial biopsy showed steatosis, an effect seen mainly in genotype 3 infection. Castera et al. [48] studied paired biopsies from 96 patients with chronic hepatitis C and found worsening of steatosis to be the only factor independently associated with the progression of fibrosis in a multivariate analysis (OR 4.7, 95% CI 1.3-10.8; P = 0.0001). Hickman et al. [49] studied the effect of an effective 3-month weight reduction programme in 10 subjects with steatosis and chronic hepatitis C examining paired liver biopsies prior to and 3-6 months following intervention. Nine of 10 patients had a reduction in steatosis irrespective of viral genotype. In these subjects, the median modified Knodell fibrosis score decreased from 3 to 1 (P = 0.04) and activated stellate cells significantly decreased (P < 0.004) despite the persistence of the virus. Although a single study failed to disclose steatosis in the initial biopsy as a factor associated to progression of fibrosis [50], the steatosis-fibrosis association appears to be biologically plausible as fat is a recognized precursor of fibrosis in NAFLD (reviewed in Ref. 38). The mechanisms of fibrosis associated with steatosis in HCV have been reviewed elsewhere [4]. In short, with analogy to the pathogenesis of NAFLD, we speculate insulin resistance, oxidative stress leading to lipid peroxidation and cytokines to play a major role [4]. The fact that fibrosis appears to be more common and severe in HCV-related steatosis than in NAFLD suggests that HCV or viral products, the host's metabolic alterations and the antiviral inflammatory response represent an 'extra' source of biological mediators leading to increased-collagen deposition [4]. A HCV genotype-specific effect has been suggested, but which genotype is more fibrogenic is still controversial [8]. While three published studies suggest HCV-3 to be more fibrogenic [34,42,48] a single study blames genotype 1 [36]. The mechanism through which steatosis causes fibrosis might be via the induction of apoptosis [51] or, similar to what was observed in NASH [52], via increased oxidative stress and hepatic stellate cell activation [21]. While it remains to be ascertained if abstinence from alcohol is an effective means of preventing fibrosis in these patients, it appears wise to advise HCV-positive patients to quit drinking.
Response to antiviral treatment
An impaired therapeutic effect of antiviral treatment has been consistently observed in patients with HCV and steatosis [2,53,54]. It has been speculated that this phenomenon is linked to steatosis and might be differentiated from those which are linked to obesity [7]. However, there is such a close association between steatosis and obesity, even in genotype 3 infection [36], that studies addressing steatosis in HCV-positive subjects with a normal BMI are needed. Indeed, a logistic regression model provided an odds ratio (OR) of the weight effect of 0.90 per each 10 Kg weight increase (95% CI 0.79-1.04) [55]. In the largest multi-centre study published so far, the impact of steatosis on the response rate to antiviral treatment was substantial in that the OR for SVR was 0.48 (CI 0.35-0.66) [22]. These data have also been confirmed by other authors [53]. By studying 28 patients with genotype 1 and 34 with genotype 3 HCV, these authors found that in patients with HCV genotype 1, there was no change in hepatic steatosis after treatment, irrespective of the treatment response. Among those infected with genotype 3, SVR significantly reduced steatosis (P < 0.001), but there was no change in steatosis among those without a SVR. By logistic regression analysis, SVR was the only variable predictive of improvement in hepatic steatosis (OR = 36, 95% CI = 2.7-481, P = 0.007). Castera et al. [56] found that the percentage of SVR in those patients with steatosis is 16.5%. This figure is 30% less than the expected response rate following the current gold standard combination of peginterferon (PEG-IFN) and Ribavirin this being 54-61% (ranges 42-51% for patients with genotype 1 infection through 78-82% of cases with genotypes 2 or 3) (reviewed in Ref. 57). In Poynard's experience [2], steatosis is more prevalent in nonresponders (65%) than among sustained responders (47%, P < 0.001). However, steatosis-associated resistance to antiviral treatment is specific for metabolic steatosis because the viral steatosis observed in genotype 3 is not associated with lower sustained response [2].
Noninsulin dependent diabetes mellitus
Several lines of evidence support a close relationship between HCV chronic infection and NIDDM [3,58]. Up to one-third of patients with HCV chronic infection develops NIDDM. This prevalence is much higher than that observed in the general population and in patients with other chronic liver diseases such as hepatitis B virus, alcoholic liver disease and primary biliary cirrhosis. Further, HCV seropositivity in patients with NIDDM appears to be higher than in the general population. Postliver transplantation NIDDM also appears to be higher among patients with HCV. Finally, HCV infection substantially increases the risk of developing NIDDM. As both NIDDM and HCV chronic infection are closely linked to steatosis, steatosis has been postulated to be a mediator of insulin resistance in HCV-positive subjects [4]. Although some studies failed to demonstrate that this is the case [59,60], our data indicate that homeostasis model assessment-insulin resistance (HOMA-IR) in HCV-positive subjects is higher than that observed in healthy and FHBL control subjects and is in the same order of magnitude to that observed in NAFLD (A. Lonardo, unpublished observation), a condition widely accepted to be a prototype insulin-resistant state [61]. Clearly, additional studies are needed in this area before any firm conclusions can be drawn.
Hepatocellular carcinoma
A single study has reported steatosis to be a risk factor for the development of HCC in HCV-positive patients. Ohata et al. [32] followed-up for up to 15 years 161 patients with chronic HCV infection and found that the presence of steatosis was significantly associated with the incidence of HCC (risk ratio 2.81, 95% CI 1.24-6.37 P = 0.0135) in a multivariate analysis. The mechanisms through which hepatic steatosis might contribute to hepatocarcinogenesis remain unknown. HCV might exacerbate oxidative stress produced by steatosis: e.g. HCV core protein may alter the oxidant/antioxidant state in the liver in the absence of inflammation and may interact with retinoid X receptor alpha, a transcriptional regulator that controls many aspects of cell proliferation, differentiation and lipid metabolism [62]. Although Ohata's data are clinically relevant and biologically plausible, they need to be confirmed also in other geographic areas where HCC has a lesser incidence.
In conclusion, steatosis has been convincingly demonstrated to significantly impact on the natural history of HCV infection. Therefore, it is of great clinical relevance to determine whether life-style alterations might affect the natural history of steatosis.
Do life-style changes impact on steatosis in HCV-positive patients?
If we assume that genotype 3 directly causes steatosis and that, in contrast, steatosis associated with 'non-3' genotypes (particularly genotype 1) occurs mostly as a consequence of the host's risk factors for steatosis [2,8], we would expect no benefit from body weight reduction in patients infected with HCV genotype 3. Contrary to this prediction, Hickman et al. [63] found a significant reduction in the grade of steatosis in their overweight or obese patients who were ineligible for, or nonresponders to, antiviral treatment. Interestingly, this was found not only in four HCV genotype 1 but also in those seven genotype 3 patients with steatosis after a weight reduction programme based on a 15-month diet and exercise intervention. These findings need to be confirmed by larger studies. However, it is of interest that in patients with genotype 3 infection on antiviral treatment, after disappearance of HCV RNA from serum, the same associations were observed with metabolic parameters than were seen in patients infected with non-3 genotypes [2]. Taken collectively these data probably indicate that a clear-cut distinction between 'viral' and 'metabolic' steatosis should be viewed with caution. It appears likely that both host and viral factors are present, though to a variable extent, both in HCV genotype 3 and in 'non-3' steatosis. Before drawing any conclusions, an effort should be made to understand the reasons why discrepancies exist among the various studies about HCV-related steatosis.
Can discrepancies be reconciled and any conclusions be drawn?
The discrepancies in the results reported by various authors who have studied steatosis in HCV-patients can be traced back to differences in some of the features of the case series, including BMI, alcohol consumption, prevalence and extent of steatosis in relation to genotypes, treatment schedules, and criteria of exclusion (i.e. patients with NIDDM). Some of these items, such as prevalence, body distribution and severity of adiposity - probably the most powerful determinant of hepatic steatosis - reflect the varying prevalence rates of obesity in the general population, with studies from USA and Australia reporting higher rates than European studies [8]. In addition, the role of alcohol is also inherently difficult to be ascertained as patients might alter their drinking habits when notified that they have chronic HCV hepatitis [8]. Furthermore, different drinking patterns i.e. in the fasting as opposed to the postprandial phase might differently affect the risk of hepatotoxicity [64]. Failure to study epidemiological, pathogenic and clinical patterns separately in 3 vs'non-3' genotypes might be another reason why earlier studies appear to have provided less dependable results than modern studies.
If we are to make an effort to reconcile these discrepancies, the following conclusions can be drawn: on the grounds of its high prevalence rate and of its heavy impact on the history of treated and untreated HCV chronic infection, steatosis represents a distinct disease entity. Its pathogenesis is diverse and involves both host (prevalently observed in those patients infected with non-3 genotypes) and viral factors (more important in untreated HCV genotype 3 infection) though to a variable extent as a function of the genotype involved. Alcohol consumption should prudently be discouraged and specific follow-up schedules need to be instigated for patients displaying steatosis in the context of HCV infection, because of their inherent risk for fibrosis progression and perhaps of their increased oncogenic risk. Finally, numerous patients do not benefit from current antiviral combination regimes. They should be offered the chance of a 'metabolic' treatment aimed at steatosis, similar to those are currently being tested in nonalcoholic fatty liver disease. The impact of combined antiviral and 'metabolic' approaches should also be evaluated vs current standard antiviral treatment.
What is the background of the epidemiological burden imposed by hepatitis C?
Hepatitis C virus is a small enveloped virus belonging to the Flaviviridae. Based on molecular biology techniques six different (1-6) genotypes are recognized. Over the past decades, there have been two major HCV epidemics. The first one took place in the 1960s; it was mainly because of transmission of HCV genotype 1 through medical procedures. The second occurred in the 1980s, was transmitted by needle-sharing among drug addicts and was because of HCV genotype 3 [9]. The natural course of HCV infection is variable and modulated by the interaction of host [age, gender, HLA, genetic polymorphisms in TGF-B1 and angiotensinogen, TNF-a, Body mass index (BMI)] and viral factors (viral load) [10-12]. However, following acute infection that occurs through the parenteral route, chronicity ensues in approximately 85% of those who are exposed. Given that it affects approximately 3% of the world's population, HCV infection represents a leading cause of chronic hepatitis, cirrhosis, end-stage liver failure necessitating transplantation and hepatocellular carcinoma (HCC). Our weapons against HCV infection have a limited efficacy, are costly and are not devoid of side-effects. A recent study performed in 327 patients referred to a liver clinic after a positive result for antibody against HCV found that 72% were not treated on account of failure to adhere to evaluation procedures, patient preference or normal liver enzymes [13]. Only 23% were treated and 13% had a sustained viral response.
The varying patterns of transmission of 3 vs non-3 HCV genotypes justifies a study of the extent and prevalence of steatosis as a function of the genotype.
Are the prevalence and the extent of steatosis related to HCV genotype?
The reported prevalence of steatosis in patients with HCV infection [2,5,14-36] ranges from 34.8 [32] to 81.2% [28] dependent on the characteristics of the patients studied (i.e. alcohol consumption, percentage with obesity, diabetic, hyperlipidemic cases, etc.) that probably mirror local differences in lifestyles and prevalence of disease. By summing up 25 studies, collectively including 6400 patients, the overall prevalence of steatosis in patients with HCV infection is 55.54% (Table 1). This figure represents a two to threefold increase compared with the prevalence of steatosis in autoimmune (17%), hepatitis B and cryptogenic (27%) chronic hepatitides and to the prevalence expected by chance in of nonalcoholic fatty liver disease and HCV infection [1,4]. Table 2 highlights the different histological features of steatosis in various steatogenic liver diseases [37,38].
Adinolfi et al. [18,23] were first in reporting that the genotype was a major factor associated with steatosis in HCV infection. This finding has been confirmed by several authors and by pooling published data [2,5,27,34-36,39] it can be calculated that steatosis has a prevalence of 415/560 (74.10%) in genotype 3 vs 1179/2460 (47.92%) in non-3 genotype (P < 0.01). Recognition that genotype 3 might be specifically associated with steatosis has prompted further correlative studies of the extent and prevalence of steatosis according to the genotype. Combining all the studies that specify the extent of steatosis divided into three grades, one mild to three severe [34-36], an interesting observation can be made. While the prevalence of grade 2 steatosis is in the same order of magnitude in non-3 genotypes and genotype 3 (20%vs 23%P = ns), there is a significant prevalence of non-3 genotypes in steatosis grade 1 (73.88%vs 46.92%P < 0.01) and of genotype 3 in steatosis grade 3 (29.6 vs 5.52 P < 0.01). Therefore, it appears to have been convincingly shown that in the setting of genotype, steatosis is both more prevalent and more severe than in infection with non-3 genotypes. This conclusion reasonably leads us to next address the risk factors for steatosis as a function of HCV genotype.
Is hypolipidaemia relevant to the pathogenesis of steatosis observed in the setting of HCV chronic infection?
Data correlating serum cholesterol levels and liver steatosis in HCV infection follow an inverse relationship at baseline [2,27,29]. hypocholesterolaemia is corrected only in those patients with sustained virological response (SVR) following IFN-Ribarin treatment both in genotype 3 and in non-3 genotypes [2]. Such a phenomenon is strongly suggestive of an acquired (i.e. virus-induced) hypobetalipoproteinaemia (HBL). Familial (genetic) HBL (FHBL) is an autosomal co-dominant disease whose prevalence in the general population is unknown. FHBL is often because of apo B gene mutations that result in the synthesis of truncated apo Bs. Given that truncated apo Bs are dysfunctional, FHBL is characterized phenotypically by serum cholesterol and serum triglyceride levels <95th percentile. Steatosis in this disease results from impaired export of triglycerides from hepatocytes into the bloodstream because of truncated apo Bs. Not surprisingly, in FHBL, subjects with steatosis have significantly lower cholesterol serum levels than subjects without ultrasonographic evidence of steatosis (A. Lonardo, P. Loria, L.E. Adinolfi, N. Carulli, G. Ruggiero, unpublished data). FHBL represents an extraordinary model of human disease to address the relationship between serum cholesterol levels and intrahepatic fat accumulation. An Austrian study performed in 137 IFN-naive patients with chronic hepatitis C suggests that in addition to inducing steatosis HCV-3a lowers serum cholesterol, an effect which is fully reversible only in those with a SVR [41]. These findings are not accounted for by the effect of antiviral drugs because interferon induces - via inhibition of LDL - an increase mostly in those HCV genotypes 1 patients with lower triglyceride pretreatment levels [43]. The exact mechanism(s) through which HCV induces acquired HBL are unknown. Interactions with apo B and microsomal triglyceride transfer protein have been proposed (reviewed in Ref. 3). One study reported that higher serum LDL cholesterol at baseline predicts a response to IFN therapy [44]. These data have not been confirmed by a second study [24] perhaps because low LDL cholesterol levels are associated with other predictors for poor response to IFN therapy such as advanced fibrosis and high HCV load. In any case, a critical issue to be evaluated is whether steatosis represents an innocent bystander in the course of HCV infection.
|
|
| |
| |
|
|
|
