| |
67% Steatosis in HCV/HIV Coinfected: : Liver inflammation, HCV genotype 3, and BMI are associated with steatosis, a common finding in HCV-HIV-coinfected patients.
|
| |
| |
"Prevalence of and Factors Associated With Hepatic Steatosis in Patients Coinfected With Hepatitis C Virus and HIV: Agence Nationale pour la Recherche contre le SIDA et les hepatites virales CO3 Aquitaine Cohort"
JAIDS Journal of Acquired Immune Deficiency Syndromes:Volume 45(2)1 June 2007pp 168-173
Neau, Didier MD, PhD*; Winnock, Maria PhD; Castera, Laurent MD; Bail, Brigitte Le MD, PhD; Loko, Marc-Arthur MD; Geraut, Laurent MD*; Dupon, Michel MD*; Ragnaud, Jean-Marie MD*; Lacoste, Denis MD; Lafon, Marie-Edith MD, PhD#; Bioulac-Sage, Paulette MD, PhD; Dabis, Francois MD, PhD; the Groupe d'Epidemiologie Clinique du SIDA en Aquitaine
From the *Federation des Maladies Infectieuses et Tropicales, Centre Hospitalo-Universitaire (CHU), Hopital Pellegrin, Bordeaux, France; INSERM U593, Institut de Sante Publique, d'Epidemiologie et de Developpement (ISPED), Universite Victor Segalen, Bordeaux, France; Services d'Hepato-Gastroenterologie, CHU, Hopital Saint-Andre et Hopital Haut-Leveque, Bordeaux, France; Laboratoire d'Anatomie Pathologique, CHU/Universite Victor Segalen, Hopital Pellegrin, Bordeaux, France; Departement de Medecine Interne et de Maladies Infectieuses, Hopital Saint-Andre, Bordeaux, France; Centre d'Information et de Soins de l'Immunodeficience Humaine (CISIH), Centre Hospitalo-Universitaire (CHU), Hopital Pellegrin, Bordeaux, France; and #Laboratoire de Virologie EA 2968, CHU/Universite Victor Segalen, Hopital Pellegrin, Bordeaux, France.
Abstract
Background: Hepatic steatosis is a common feature in liver biopsies from patients with chronic hepatitis C and is associated with fibrosis progression. Patients with HIV infection and hepatitis C virus (HCV) coinfection have more rapid progression of liver fibrosis than patients with HCV infection alone. The prevalence and factors associated with hepatic steatosis are not well defined in HCV-HIV-coinfected patients.
Methods: Steatosis was assessed among 148 HCV-HIV-coinfected patients of the Aquitaine Cohort. Steatosis was graded as follows: none, mild (1%-10% of hepatocytes), moderate (11%-30%), severe (31%-60%), and massive (more than 60%). Epidemiologic, clinical, biologic, and therapeutic data were retrieved from the cohort database to investigate the risk factors.
Results: Steatosis was present in 67% of patients (95% confidence interval [CI]: 59% to 74%) and was at least moderate in 30% (95% CI: 23% to 38%). Steatosis was macrovesicular or mixed (macro- and microvesicular) in 40.5% and 52.8% of patients, respectively. Necroinflammatory activity was the only factor independent of steatosis (adjusted odds ratio = 5.3, 95% CI: 1.6 to 17.9). When necroinflammatory activity was removed from the model, HCV genotype 3 and body mass index (BMI) were significantly associated with steatosis.
Conclusions: Liver inflammation, HCV genotype 3, and BMI are associated with steatosis, a common finding in HCV-HIV-coinfected patients.
Hepatitis C virus (HCV)-induced liver disease is now an important cause of mortality in HIV-infected patients.1 Indeed, HIV infection has been associated with more rapid progression of fibrosis in patients with chronic hepatitis C.2,3 Consequently, cirrhosis and hepatocellular carcinoma occur after a shorter duration of HCV infection in HCV-HIV-coinfected patients.4-6 The mechanisms involved in the rapid progression of liver fibrosis in HIV-infected patients remain unclear, however. The course of HCV-related hepatic disease varies markedly from one patient to another. In patients who are infected only with HCV, several factors, including age at exposure, duration of infection, alcohol intake, and male gender, have been shown to be independently associated with fibrosis progression.7-9 Even in the absence of these factors, however, disease progression may occur in some patients, suggesting a role of other factors that remain to be identified.
Hepatocellular steatosis is defined by the accumulation of lipid droplets (mainly triglycerides) in hepatocytes. The size and distribution of these droplets may be an indication of pathogenesis.10 Indeed, microvesicular steatosis is related to defective β-oxidation of free fatty acids and may be associated with liver failure in conditions such as acute fatty liver of pregnancy and Reye syndrome, whereas macrovesicular steatosis results from a complex combination of pathogenic alterations that include increased delivery, inadequate oxidation, and reduced secretion of various forms of lipids in the liver.11 Macrovesicular steatosis is found in a wide range of clinical settings, including excessive alcoholic intake, obesity, hyperlipidemia, type 2 (noninsulin-dependent) diabetes, and chronic hepatitis C, where it has been reported with a prevalence ranging from 30% to 70%.12 This latter observation is of clinical relevance from diagnostic and prognostic points of view. Indeed, there is increasing evidence that steatosis is associated with fibrosis progression in patients with chronic hepatitis C.13-17 Two distinct forms of hepatocellular steatosis can be observed in patients with chronic HCV infection. Classic metabolic risk factors for steatosis such as obesity, diabetes type 2, hyperlipidemia, and excessive alcohol intake account for most cases of steatosis in patients infected with HCV non-3 genotypes. In contrast, in patients infected with HCV genotype 3, steatosis is thought to be induced by the virus itself through a direct cytopathic effect, as suggested by several lines of evidence, including the following: (1) the development of hepatic steatosis in transgenic mouse lines expressing HCV core protein;18 (2) the significant relation between the presence and severity of steatosis and HCV genotype 3 RNA plasma viral load, although no such relation has been found with other HCV genotypes;13,19-21 and (3) the disappearance of steatosis in HCV genotype 3-infected patients who had a sustained viral clearance after antiviral therapy19,20,22,23 and recurrence at relapse.21,24
In HIV-infected patients, steatosis was a common finding in liver biopsies and autopsies before the era of highly active antiretroviral therapy (HAART), with prevalence rates of more than 30%.25 HAART is an exogenous factor of liver steatosis related to the use of nucleoside reverse transcriptase inhibitors (NRTIs), which are potentially responsible for mitochondrial injury.26 Patients receiving HAART also have potential risk factors of developing steatosis, such as hyperlipidemia and decreased insulin sensitivity, which are associated with the use of protease inhibitors (PIs).
The aim of our study was to evaluate the prevalence of steatosis in HCV-HIV-coinfected patients included in the Aquitaine Cohort. The relation between steatosis and viral or host factors, including the use of HAART, was also further investigated.
DISCUSSION
Our study suggests that steatosis is a common histologic feature in HCV-HIV-coinfected patients, with a prevalence (67%) higher than reported in previous studies.29-31 Indeed, in a study of 112 patients in the United States, Sulkowski et al29 reported hepatic steatosis (involving at least 5% of hepatocytes) in 18% of patients. In 2 US-based studies on 92 and 106 patients, steatosis (involving at least 1% of hepatocytes) was found in 47% and 56% of patients, respectively.29,30 In 3 studies (2 from United States and 1 from France), however, steatosis (involving at least 1% of hepatocytes) was present in 61% to 72% of patients; these latter results are close to those observed in the Aquitaine Cohort.32-34 Differences in steatosis grading make it difficult to compare these findings. For example, in some studies, patients had mild steatosis involving <33% of hepatocytes, but no further details on the distribution of steatosis within this grade were provided, thus limiting the interpretation of these findings from a clinical standpoint.30,32,33 In the present study, steatosis was graded according to the previously validated METAVIR scoring system, and 30% of patients had steatosis that could be considered significant (involving at least 11% of hepatocytes).14,28 The high proportion of black patients included in 3 studies from the United States (93%, 31%, and 47%, respectively) may be another explanation for the lower reported prevalence of steatosis. Indeed, it has been recently suggested that subjects of black race have a lower incidence of liver steatosis than whites.35 Most of our patients in southwestern France are white; thus, this factor could not be investigated in this study. Finally, the higher proportion of patients infected with HCV genotype 3 in our study than in the studies published in the United States may also account for the higher prevalence of steatosis, as observed in the other French study.32
Several factors have been reported to be independently associated with steatosis in HCV-HIV-coinfected patients, although not systematically investigated.29-34 They include white race, age, BMI, hyperglycemia, lower levels of high-density lipoprotein cholesterol, presence of lipodystrophy, HCV genotype 3, HCV plasma viral load, stavudine or didanosine use, and ferritin increase. In the present study, in which most of these parameters were studied in univariate and multivariate models, histologic activity was the main determinant of steatosis. In HCV-monoinfected patients, several studies reported an association between histologic activity and steatosis in univariate13,21 and multivariate36 analyses. In coinfected patients, steatosis has also been associated with histologic activity and fibrosis in a study by Sulkowski et al.29 Several mechanisms may account for the relation between steatosis and histologic activity. HCV infection is associated with increased cytokine production enhancing histologic inflammation and leading to increased lipid peroxydation.37 Also, HCV core protein has been shown to induce oxidative stress in vitro and in vivo.38 In HIV-HCV-coinfected patients, histologic inflammation may be related not only to HCV infection but to the use of antiretroviral drugs. Although no relation was found between steatosis and antiretroviral treatment, we cannot rule out the implication of the latter. Indeed, microvesicular steatosis may be induced by NRTIs, which inhibit mitochondrial DNA polymerase-γ.26 Macrovesicular steatosis could be linked to the use of PIs and associated with decreased insulin sensitivity, development of diabetes mellitus, and substantial increases in plasma lipids. It must be stressed, however, that in the present study, when histologic activity was removed from the model, HCV genotype 3 and BMI were significantly associated with steatosis, a finding consistent with those of Bani-Sadr et al,32 who reported similar risk factors for steatosis as in HCV-monoinfected patients.
Overall, hepatic steatosis was frequently observed in our cohort. Investigating a large set of determinants, steatosis was associated with liver inflammation. As identified in HCV-monoinfected patients, genotype 3 and BMI were also risk factors for steatosis. Further investigations are required to achieve better understanding of steatosis in HCV-HIV-coinfected patients, a complication that impairs clinical management.
RESULTS
Study Population
A total of 148 patients were available for the study (Table 1). Their median age was 38.8 years, and 64% were male. The main source of HIV infection was intravenous drug use (65.5%). BMI was normal (<25 kg/m2) in 84% of the patients. The median CD4 count was 500 cells/μL, ranging from 78 to 1644 cells/μL. The HIV plasma RNA level was ≦50 copies/mL in 44% of the patients. Forty-three percent of the patients were infected with HCV genotype 1, whereas 31% were infected with HCV genotype 3; 13% of the patients had received interferon monotherapy without virologic efficacy at least 2 years before the biopsy. All patients who underwent histologic evaluation were eligible for anti-HCV treatment and reported an alcohol intake <40 g/d during the 6 months before the biopsy.
Hyperglycemia (>5.8 mmol/L) and hypertriglyceridemia (>1.5 mmol/L) were observed in 20% and 40% of the patients, respectively. Ninety percent of the patients were receiving HAART at the time of the biopsy; 40% were treated with a combination that included a PI, whereas 36% were treated with a nonnucleoside reverse transcriptase inhibitor (NNRTI).
Prevalence, Type, and Severity of Hepatic Steatosis
The prevalence and severity of hepatic steatosis are reported in Figure 1. Steatosis was present in 99 patients (67%, 95% confidence interval [CI]: 59% to 74%). Forty-five patients (30%, 95% CI: 23% to 38%) had significant fat accumulation in more than 11% of hepatocytes, corresponding to moderate, severe, or massive steatosis. Steatosis was macrovesicular or mixed in 36 (40.5%) and 47 (52.8%) patients, respectively. Only 6 patients (6.7%) had isolated microvesicular steatosis.
Determinants of Hepatic Steatosis
In univariate analysis, significant hepatic steatosis (at least moderate) was associated with HCV genotype 3 (odds ratio [OR] = 2.3, 95% CI: 1.1 to 4.8), increased AST level >2.5 times the normal value (OR = 2.9, 95% CI: 1.1 to 7.4), BMI (OR = 2.6, 95% CI: 1.1 to 6.4), severe necroinflammatory activity (METAVIR score A3; OR = 5.9, 95% CI: 1.8 to 19.7), and NNRTI exposure (OR = 0.4, 95% CI: 0.2 to 0.9). No association was detected with exposure to stavudine, an NRTI used in 43% of our patients. Necroinflammatory activity was severe (METAVIR score A3) in 16 (35%) of 45 patients with significant hepatic steatosis compared with 14 (13%) of 103 patients without significant steatosis (P = 0.004). Severe fibrosis or cirrhosis (METAVIR score F3 or F4) was observed in 19 (42%) of 45 patients with significant steatosis compared with 28 (27%) of 103 patients without significant steatosis (P = 0.07).
In multivariate analysis (Table 2), severe necroinflammatory activity (METAVIR score) was the only factor independent of significant steatosis (adjusted odds ratio [AOR] = 5.3, 95% CI: 1.6 to 17.8). The association with HCV genotype 3 was close to statistical significance (P = 0.08). When necroinflammatory activity was removed from the model, however, genotype and BMI were significantly associated with steatosis.
METHODS
Study Population
The Aquitaine Cohort is a prospective hospital-based cohort of HIV-1-infected patients under routine clinical management. It was initiated in 1987 at the Bordeaux University Hospital and 4 other public hospitals in the Aquitaine region of southwestern France by the Groupe d'Epidemiologie Clinique du SIDA en Aquitaine (GECSA).27 All adults who are inpatients or outpatients of the participating hospital wards with HIV-1 infection confirmed by Western blot testing and who have given informed consent are enrolled in the cohort, whatever their clinical stage, gender, or HIV transmission category. Also, information from at least 1 follow-up visit after the baseline assessment or a known date of death must be available. At each hospital contact, a standardized questionnaire, including epidemiologic, clinical, biologic, and therapeutic data, is filled in by clinicians for each patient included in the cohort and entered into the database. The schedule of follow-up visits is based on clinical practice, and an active search of patients lost to follow-up is performed annually.
Between January 1999 and December 2003, HIV-infected patients had a liver biopsy performed at the Bordeaux University Hospital in case of chronic HCV infection, defined as positive serum antibodies to HCV by an enzyme-linked immunosorbent assay (ELISA) third-generation test (Ortho HCV 3.0 ELISA, Monolisa antiHCV; Sanofi Diagnostics Pasteur, Paris, France) and detectable plasma HCV RNA by a sensitive polymerase chain reaction (PCR) technique (Cobas Amplicor HCV 2.0; Roche Diagnostics, Branchburg, NJ). Patients were eligible for inclusion in the study if they had a liver biopsy available and if they had never been previously treated with interferon or peginterferon-α and ribavirin. They could have received interferon monotherapy at least 2 years before the liver biopsy but without documented virologic efficacy. Serum hepatitis B surface antigen was negative in all the patients. Epidemiologic, clinical, biologic, and therapeutic data within 3 months of the biopsy were retrieved from the cohort database.
Steatosis Grading and Histologic Evaluation
A fragment of at least 10 mm was fixed in 10% formalin, paraffin embedded, and serially sectioned. Sections were stained with hematoxylin-eosin-saffron, Masson trichrome, picrosirius red, and Gordon and Sweet stains. Biopsy samples were read by the same 2 experienced pathologists (PBS, BLB), using the METAVIR scoring system.28 Stage of fibrosis was defined as F0 (none), F1 (portal fibrosis without septa), F2 (few septa), F3 (numerous septa without cirrhosis), and F4 (cirrhosis). Grade of activity was defined as A0 (none), A1 (mild), A2 (moderate), or A3 (severe). Steatosis was graded as follows: none, mild (involving 1%-10% of hepatocytes), moderate (involving 11%-30% of hepatocytes), severe (involving 31%-60% of hepatocytes), or massive (involving more than 60% of hepatocytes). Steatosis was considered significant when it was at least moderate according to this classification. In addition, the type of steatosis (macrovesicular, microvesicular, or mixed) was characterized.
Statistical Analysis
For statistical analysis, it was considered that steatosis was significant if it was at least moderate. Patients without or with mild steatosis were compared with patients with moderate, severe, or massive steatosis. Variables considered for analyses of steatosis-determining factors were age at liver biopsy, gender, body mass index (BMI), source of HIV infection, Centers for Disease Control and Prevention (CDC) HIV classification, plasma HIV RNA level, CD4 cell count, HCV genotype, previous anti-HCV treatment, aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma-glutamyl transpeptidase, glycemia, total cholesterol, triglycerides, use of antiretroviral drugs before liver biopsy, histologic activity, and fibrosis. Continuous variables were expressed by median and range, and categoric variables were expressed by absolute figures and percentages. Univariate comparisons between patients with (at least moderate) and without (none or mild) significant steatosis used χ2 and Mann-Whitney tests for qualitative and quantitative variables, respectively.
Logistic regression models were fitted to identify determinants of hepatic steatosis. Variables included in multivariate models were those with P values ≦0.25 in univariate analyses. Statistical analyses were performed using SAS software, version 8.1 (SAS Institute, Cary, NC).
|
|
| |
| |
|
|
|
