Generalized Linear Model (GLM) framework for the association of host variables and viral strains with liver fibrosis in HCV/HIV coinfected patients.

Chronic hepatitis C virus (HCV) infection is the main cause of advanced and end-stage liver disease world-wide, and an important factor of morbidity and mortality in Human Immunodeficiency virus-1 (HIV-1) co-infected individuals. Whereas the genetic variability of HCV has been studied extensively in monoinfected patients, comprehensive analyses of both patient and virus characteristics are still scarce in HCV/HIV co-infection. In order to find correlates for liver damage, we sought to analyze demographic, epidemiological and clinical features of HCV/HIV co-infected patients along with the genetic makeup of HCV (viral subtypes and lineage studied by nucleotide sequencing and phylogenetic analysis of the NS5B region). We used the Generalized Linear Model (GLM) methodology in order to integrate data from the virus and the infected host to find predictors for liver damage. The degree of liver disease was evaluated indirectly by means of two indexes (APRI and FIB-4) and accounting for the time since infection, to estimate fibrosis progression rates. Our analyses identified a reduced number of variables (both from the virus and the host) implicated in liver damage, which included the stage of HIV infection, levels of gamma-glutamil transferase and cholesterol, and some distinct HCV phylogenetic clades.

Relating the liver damage with hepatitis C virus polymorphism in core region and human variables in HIV-1-coinfected patients.

Hepatitis C virus (HCV) infection is the most important cause of chronic hepatitis, cirrhosis and end-stage liver disease leading to liver transplantation worldwide. Chronic infection by HCV causes liver fibrosis, which is accelerated by unknown mechanisms in patients with human immunodeficiency virus-1 (HIV-1) coinfection. Although the genetic variability of both HCV and HIV has been extensively studied in the context of monoinfections, more limited data is available regarding HCV-HIV coinfection. HCV disease progression among HIV coinfected patients may be influenced not only by demographic, epidemiological and clinical background variables, but also by genetic differences in infecting viruses. To explore this issue, we carried out a study in coinfected patients trying to associate the degree of liver damage to several demographic, clinical, and epidemiological characteristics of the patients, and also to the genetic variability of HCV between patients. For this purpose, we have applied different statistical techniques including the statistical generalized linear model (GLM) framework. The stage of fibrosis was indirectly measured by noninvasive means using the indexes Forns, APRI and FIB-4. HCV genetic variability between patients was estimated by sequencing the core region and by reconstructions of consensus maximum parsimony phylogenetic trees with 50% and 75% bootstrap majority rules. The results showed a direct correlation of the fibrosis biomarkers with the AST/ALT ratio, MoftIDU and with 3a HCV genotype clades, among others.