Risk of Severe Clinical Events After Sustained Virological Response Following Direct-Acting Antiviral Therapy in HIV and Hepatitis C Virus Coinfected Participants

Mathieu Chalouni; Linda Wittkop; Firouzé Bani-Sadr; Karine Lacombe; Laure Esterle; Camille Gilbert; Patrick Miailhes; David Zucman; Marc Antoine Valantin; Sylvie Brégigeon-Ronot; Philippe Morlat; Eric Billaud; Lionel Piroth; Alissa Naqvi; Philippe Sogni; Dominique Salmon


HIV Medicine. 2021;22(9):791-804. 

In This Article

Population and Methods


HIV/HCV-coinfected patients from the ANRS CO13 HEPAVIH cohort (ClinicalTrials.gov Identifier: NCT03324633), a multicentre, prospective, nationwide French cohort study, were included. Participants were followed semi-annually for cirrhotic participants and annually for non-cirrhotic participants, with supplementary visits scheduled during HCV therapies. Written informed consent was obtained from each participant included in the cohort. The study followed the ethical principles of the World Medical Association (Declaration of Helsinki) and was approved by an Institutional Review Board (Comité de protection des Personnes Ile de France III, Paris, France).

Participants were included if they reached SVR after a treatment by DAA. For each outcome, participants with a history for the investigated event were excluded. Participants were classified as having reached SVR if they had an undetectable HCV RNA at least 12 weeks after the end of the anti-HCV treatment. The SVR date was defined as 12 weeks after the end of the anti-HCV therapy.


The outcomes were time between SVR and the occurrence of a liver-related event, death from any cause, AIDS-defining event, NLNA cancer and ischaemic event. Causes of death were validated by an adjudication committee of the cohort. A liver-related event was defined according to International Classification of Diseases, Tenth Revision (ICD10) codes as the occurrence of a liver decompensation [liver encephalopathy (ICD10: 10019660), hepatorenal syndrome (ICD10: 10019846), oesophageal varices bleeding (ICD10: 10058122; 10056084; 10030210; 10055968), oedemato-ascitic decompensation (10003445; 10014210) and icterus (ICD10: 10021206; 10023126; 10023139; 10057488)], a hepatocellular carcinoma (HCC) (ICD10: 10008593; 10049010; 10024659; 10027761; 10051969), or a death due to one of these causes. For ascites, it was checked that no cancers or cardiovascular events occurred in the 6 months before the occurrence of the event, and for icterus that no atazanavir treatment was taken, which could have caused the event. The occurrence of an AIDS-defining event was defined using the Centers for Disease Control and Prevention (CDC) criteria. The NLNA cancers were all cancers excluding HCC, Kaposi sarcoma, cervical cancer and non-Hodgkin's lymphoma. Ischaemic events were defined as the occurrence of an ischaemic cardiac angina, transient ischaemic attacks, ischaemic heart diseases, strokes, myocardial infraction or death from a cardiovascular cause.

Statistical Analysis

Incidences rates of each event and their confidence intervals at 95% were estimated using Poisson models. Incidence rates were estimated overall and according to participants' characteristics selected a priori for each event. Age (< 50, 50–60, ≥ 60 years), sex, alcohol consumption (never, past or current consumption as declared by participants), CD4 count (< 200, 200–500, ≥ 500 cells/μL), undetectable HIV viral load and cirrhosis were selected for every event. Cirrhosis status was defined by a cohort-specific algorithm: an F4 classification on the Metavir Score by liver biopsies at any time before DAA treatment, liver stiffness ≥ 12.5 kPa (assessed by FibroScan) in a 3-month window before the initiation of DAA, or a fibrosis-4 (FIB-4) index value ≥ 3.25 in a 3-month window before DAA initiation for participants without an available liver stiffness value. In addition, HCV genotype (1, 3, others) liver stiffness (< 12.5, ≥ 12.5 kPa), FIB-4 (1.45, 1.45–3.25, ≥ 3.25), platelets (< 150, ≥ 150 cells/μL), albumin (< 40, ≥ 40 g/L) and gamma-glutamyl transferase (GGT) (< 50, ≥ 50 IU/L) were included for the analysis of liver-related events. For ischaemic events, body mass index (BMI; < 18.5, 18.5–25, ≥ 25 kg/m2), tobacco consumption (never, past, current), total cholesterol (< 5.0, ≥ 5.0 mmol/L), HDL cholesterol (< 1.0, ≥ 1.0 mmol/L), LDL cholesterol (1.4, ≥ 1.4 mmol/L) and arterial hypertension were selected. Finally, for NLNA cancers, BMI, tobacco consumption, metabolic syndrome (defined following the National Cholesterol Education Program recommendations[21]) and CD4 count nadir (< 150, 150–300, ≥ 300 cells/μL) were added. For overall mortality, every variable included for the analysis of event incidence rates was included. All variables were measured at time of SVR, except for variables reflecting liver damage (cirrhosis, liver stiffness, FIB-4, platelets, albumin and GGT), which were measured at time of DAA treatment initiation as their values could have evolved during and after anti-HCV treatment.

Cumulative probabilities of events over time were estimated according to treatment generation and cirrhosis status using Aalen–Johansen methodology.

To identify factors associated with the risk of each event, incidence rate ratios (IRRs) were estimated for every variable selected. The IRRs were estimated using Poisson models adjusted on age and sex.