The purpose of this study is to compare the liver toxicity in HIV-infected patients with chronic hepatitis B and/or hepatitis C, who start a new antiretroviral drug regimen, as well as the influence of the degree of pre-existing liver fibrosis on the incidence of liver toxicity.

In the last years, various clinical trials and studies have evaluated the incidence of hepatic toxicity (HT) associated with the commonly used antiretroviral drugs in the HIV/hepatitis C virus (HCV)-infected population. Unfortunately, clinical trials that compared hepatic safety of these antiretrovirals include a low number of coinfected patients (van Leth 2004, Molina 2010, Ortiz 2008, Rockstroh 2012). According to recent cohort studies, rates of grade 3 or 4 transaminase elevations (TE) observed in patients receiving ritonavir-boosted protease inhibitors PI/r as lopinavir (LPV/r), atazanavir (ATV/r), fosamprenavir (FPV/r) and darunavir (DRV/r) are similar to those observed in individuals that receive efavirenz (EFV) or raltegravir (RAL) (Pineda 2008, Palacios 2006, Macías 2011, Neukam 2011). However, this conclusion may not be exact due to methodological problems. In this context, currently available data on severe TE are derived from cohort studies that were conducted in different populations, with distinct methods and/or data that was collected retrospectively.

Little information is available on whether the stage of liver damage influences the incidence of HT caused by antiretroviral drugs. Although two studies found an association between advanced fibrosis and higher rates of TE (Aranzabal 2005, Mira 2006), data obtained by further studies are contradictory (Pineda 2008, Palacios 2006, Macías 2011, Neukam 2011) and there are still open questions. Thus, the number of cirrhotic patients is considerably small in these cohorts, however, cirrhosis represents an important comorbidity in HIV. In these individuals, plasmatic concentrations of antiretroviral drugs could reach toxic levels which would be reflected in TE (Barreiro 2007).

Therefore, studies conducted in the same population and with the same methodology are required in order to obtain precise and reliable information on hepatic safety of all commonly used antiretroviral drugs in the clinical practice. These findings could contribute to a better understanding of differences between antiretrovirals and could therefore improve the individualization of antiretroviral therapy in individuals with chronic hepatitis B and/or C. Therefore, the HEPAVIR group of the Andalusian Society of Infectious Diseases (SAEI) has established a prospective cohort: The HEPAVIR Cohort.

Hypothesis: The incidence of severe TE associated with antiretroviral therapy in the clinical practice in hepatitis B virus and/or hepatitis C virus-coinfected patients could be different according to the administered drug.

Primary objective: To determine the incidence of grade 3 or 4 TE associated with antiretroviral therapy in the clinical practice in patients with viral hepatitis.

Secondary objectives:

• Identification of factors associated with grade 3 or 4 TE
• Analysis of the association between TE and pre-existing hepatic damage
• Evaluate the incidence of serious symptoms, including hepatic decompensations, acute fulminant hepatitis and death due to liver failure, associated with the antiretroviral drug
• Determination of the incidence of grade 3 or 4 bilirubin elevations associated with the antiretroviral drugs
• Evaluation of efficacy of the antiretroviral drugs used in the study

Scheduled visits: 0, 4, 12, 24, 36 and 48 weeks.

Definition of grade 3 or 4 laboratory abnormalities/data dictionary: Grade 3 transaminase elevations (TE) are considered when elevations between 5 and 10 times above the upper level of normality (ULN) are presented in patients with normal baseline levels of alanine-aminotransferase (ALT) and aspartate-aminotransferase (AST). Grade 4 TE are defined as ALT or AST values >10 times of the ULN. In patients with elevated baseline ALT or AST levels, 3.5- to 5-fold increase from baseline levels are considered grade 3 TE and >5-fold elevations grade 4 TE, respectively. Grade 4 total bilirubin elevations were defined as increases of total bilirubin ≥5 mg/dl.

Definition of hepatic fibrosis:

• advanced fibrosis: F3 as determined by liver biopsy or 11 kilopascals as determined by transient elastometry
• cirrhosis: F4 as determined by liver biopsy or 14.6 kilopascals as determined by transient elastometry

Variables collected within in the cohort:

• primary outcome variable: AST, ALT
• epidemiological variable: age, sex
• variables related to hepatitis C virus-infection: infection route, genotype, grade of hepatic fibrosis and method used for its determination, baseline Child-Pough-Index, previous hepatic decompensations
• variables related to HIV-infection: Centers for Disease Control and Prevention (CDC) category, HIV viral load, cluster of differentiation 4 (CD4) cell count, previous and new antiretroviral drugs
• analytical variables: platelets, cholesterol, bilirubin, gamma-glutamyltransferase (GGT), alkaline phosphatase
• clinical variables: alcohol intake
• time to TE
• percentage of patients who discontinued antiviral therapy due to TE

Quality assurance and data checks: Data will be obtained from controlled databases at the participating centers. Databases will be monitored and controlled by queries every six months. Descriptive statistics will be be applied in order to detect transcription errors.

Source data verification: not planned.

All grade 3 or 4 adverse events, as well as all unexpected events, will be reported to the Andalusian Center for Pharmacovigilance (Centro Andaluz de Farmacovigilancia, www.cafv.es).

A sample size of 500 is planned for this study.

Statistical analysis:

The outcome variables of this study will be the development of grade 3 or 4 transaminase elevations (TE) and grade 4 total bilirubin elevations (TBE) during follow-up. Comparative analyses of TE and TBE will be carried out between the different drug groups. Additionally, the relationship between the outcome variable and the following potential predictors will be assessed: age, sex, previous intravenous drug use, alcohol consumption, baseline ALT levels, baseline CD4 cell count, CDC category C, undetectable plasma HIV-RNA, HCV genotype, previous ART, type of drug, as well as significant fibrosis and cirrhosis at initiation of the new ART regimen. Continuous variables will be expressed as median [interquartile range (IQR)] and categorical variables as number [percentage; 95% confidence interval (CI)]. The density of incidence of grade 3-4 TE will be calculated as number of cases per 100 person-years of the follow-up. Continuous variables will be compared using the Student's t-test for normal distribution and the Mann-Whitney U-test otherwise, whereas the categorical variables will be analyzed applying the χ2-test or the Fisher's test, when applicable. The Wilcoxon Signed Rank test and the McNemar test will be applied to compare repeated measurements in continuous and categorical variables, respectively. Those factors that show an association in the univariate analysis with a p<0.2, as well as those with a biologically possible influence, will be entered into a logistic regression model in order to identify independent risk factors for grade 3-4 TE and grade 4 TBE. The adjusted odds ratio (AOR) and the respective 95% CI will be calculated. All p values <0.05 will be considered statistically significant. Data will be analysed using the SPSS statistical software package release 19.0 (SPSS Inc., Chicago, IL, USA) and STATA 9.0 (StataCorp LP, College Station, TX, USA).

Inclusion Criteria:

• Older than 18 years
• HIV-1 infection as confirmed by ELISA and western blot
• Chronic HCV infection as confirmed by HCV antibodies in plasma, as well as a positive HCV viral load determined by polymerase chain reaction OR chronic hepatitis B infection as confirmed by HBsAg
• Treatment-naive or pretreated patients who start a new antiretroviral regimen that includes at least one drug that has not been received by the patient before
• At least one week of exposure to new regimen
• Liver biopsy or transient elastometry determination within 12 months prior to treatment initiation

Exclusion Criteria:

• Pregnancy
• Treatment against hepatitis C virus infection
• Presence of opportunistic infections, including tuberculosis, neoplasia, autoimmune diseases. Patients receiving primary or secondary chemotherapy against an opportunistic process are not included.
• Any liver disease of vascular, metabolic, biliary, autoimmune or tumoral origin
• Patients who are not able to provide written informed consent to participate in the study
• Lack of scheduled clinical visits including blood analysis throughout study period

Patients seen at the infectious disease unit of a terciary care center