Skip to content


  • Poster presentation
  • Open Access

Risk factors for end-stage liver disease among HIV and hepatitis C virus co-infected patients in the Spanish VACH Cohort

  • 1,
  • 2,
  • 3,
  • 4,
  • 5,
  • 6,
  • 7,
  • 8,
  • 9,
  • 10,
  • 11,
  • 12,
  • 13,
  • 14,
  • 15,
  • 16,
  • 17,
  • 18,
  • 19 and
  • 20
Journal of the International AIDS Society200811 (Suppl 1) :P134

  • Published:


  • Portal Hypertension
  • Chronic Liver Disease
  • Hepatic Encephalopathy
  • Spontaneous Bacterial Peritonitis
  • Hepatorenal Syndrome


There is increasing evidence supporting the hypothesis of a beneficial effect of highly active antiretroviral therapy (HAART) on the evolution and outcome of chronic liver disease (CLD) caused by hepatitis C virus (HCV) in HIV co-infected patients. The relative merit of different drugs or drug classes is, however, less well studied.


We performed a cross-sectional study on the VACH Cohort, a multicenter cohort of HIV-infected individuals in Spain, to ascertain the possible associations between exposure to protease inhibitors (PI) or to non-nucleoside analogues (NAN), and the outcome of HCV CLD. We selected HCV co-infected patients who had ever initiated HAART and who had at least one follow-up visit [treatment evaluable (TE)]. We defined our main "exposure" variable as the total time of treatment with any of, either, PI's or NAN's. We evaluated "outcome" as the occurrence of end stage CLD (ESLD), defined as any of: ascites, oesophageal varices, hepatic encephalopathy, hepatorenal syndrome, portal hypertension, hepatocellular carcinoma, or related diagnoses (i.e. upper gastrointestinal bleeding, spontaneous bacterial peritonitis).

Summary of results

Out of 15,183 patients in the VACH database, 6,004 were TE, of whom 2,669 (44,4%) were HCV co-infected; 128 patients (4.4%) developed an ESLD. Table 1 shows some important features of this sub-cohort, according to study group. The following factors (marked with * in Table 1) were associated with ESLD in the univariate analyses: being HBsAg-positive, nadir of CD4+ cell count, age and prior diagnosis of AIDS. Exposure to either PI's or NAN's was not. In a regression model, only HBsAg, age and nadir CD4+ cell count remained associated with ESLD. In a sensitivity analysis, Kaplan-Meier curves of time from initiation of HAART to diagnosis of ESLD, restricted to patients only exposed to one class of drugs, showed no differences between PI's and NAN's.

Table 1



prior AIDS*

ever IVDU


mean age*

1st CD4 cell count

log10 last viral load

nadir CD4 count*




















In conclusion, we found no evidence to support the hypothesis of a different effect of PI's and NAN's on the occurrence of ESLD among HIV and HCV co-infected individuals. Hepatitis B co-infection, more profound immunosupression and older age were associated with this outcome.

Authors’ Affiliations

Hospital de Sierrallana, Torrelavega, Spain
Hospital Virgen de la Luz, Cuenca, Spain
Hospital Gregorio Marañon, Madrid, Spain
Hospital Infanta Cristina, Badajoz, Spain
Hospital Virgen del Rocío, Sevilla, Spain
Hospital Infanta Elena, Huelva, Spain
Hospital La Fe, Valencia, Spain
Hospital General, Granollers, Spain
Hospital Joan XXIII, Tarragona, Spain
Hospital Virgen del Rosell, Cartagena, Spain
Hospital de Valme, Sevilla, Spain
Hospital SAS, Jerez, Spain
Hospital Clinico, Puerto Real, Spain
Hospital Clinico, Valencia, Spain
Hospital Santa Creu i Sant Pau, Barcelona, Spain
Hospital Vall d'Hebron, Barcelona, Spain
Hospital General, Castellon, Spain
Hospital de Cabueñes, Gijon, Spain
VACH Data Management, Cartaya, Spain
Hospital de Basurto, Bilbao, Spain


© Teira et al; licensee BioMed Central Ltd. 2008

This article is published under license to BioMed Central Ltd.