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

Discussion

In a population of HIV/HCV-coinfected participants, the most frequent event after SVR was overall mortality, whereas liver-related events and even more AIDS-defining events were the least observed events. The incidence rates of every event were higher in cirrhotic than in non-cirrhotic participants. After SVR, markers of liver disease (cirrhosis and liver stiffness) and those related to HIV infection (CD4 count) were associated with the risk of liver-related events; age and markers of liver disease (cirrhosis, liver stiffness and GGT) were associated with increased risk of overall mortality; higher total cholesterol was associated with increased incidence of ischaemic events; and age and CD4 count were associated with the risk of NLNA cancers.

We estimated incidence rates of liver-related events of 5.9 (3.3–10.3) per 1000 PY after SVR induced by DAA treatment, with higher incidence in cirrhotic participants [vs. 16.0 (7.2–35.7)] than in non-cirrhotic participants [1.0 (0.1–6.9)]. These results are similar to those reported in previous studies. In participants reaching SVR after DAA treatment, incidence of liver decompensation was 5.5 per 1000 PY,^[11] and incidence of HCC was 19.0 per 1000 PY in participants with an advanced fibrosis.^[16] In HIV-coinfected participants treated with DAA, Cormà-Gomez et al.^[22] observed a higher incidence of liver-related complications, 12.0 per 1000 PY, which could be due to the inclusion solely of participants with an advanced fibrosis. In participants treated with DAA, the incidence rates of liver-related events were higher in cirrhotic than in non-cirrhotic participants as described previously.^[11,23] We identified cirrhosis, liver stiffness and lower CD4 count to be associated with an increased risk of liver-related events. Liver stiffness^[4,16,17,24] or factors reflecting the advancement of the liver damage^[14,23,25] were previously described as factors associated with the risk of liver events. In HIV-coinfected participants, the association between CD4 count and risk of liver-related event has already been shown. In the NA-ACCORD collaboration, CD4 count < 500 cells/μL was associated with higher risk of HCC.^[26] CD4 count < 200 cells/μL was associated with the risk of liver-related event in HIV/HCV-coinfected participants both with and without DAA treatment.^[27,28] The association between low CD4 count and liver-related events could be due to other mechanisms not taken into account here, as our models were adjusted for age and sex only. Indeed, in cirrhotic HIV-infected participants, low CD4 count may also reflect portal hypertension or splenic sequestration^[29] but also longer HIV duration, comorbidities induced by HIV, lower CD4/CD8 ratio or detectable HIV viral load. However, we did not find an association between detectable HIV viral load and the risk of liver-related events.

For overall death, we estimated a high incidence of overall mortality of 22.2 per 1000 PY, which was even higher in cirrhotic participants (44.7 per 1000 PY) than in non-cirrhotic participants (13.3 per 1000 PY). These incidence rates were close to that estimated by Backus et al.^[16] in HCV-monoinfected participants (26.0 per 1000 PY) and also in non-cirrhotic participants (11.8 per 1000 PY),^[30] but higher than that estimated by Carrat et al.^[11] (9.5 per 1000 PY) in HCV-monoinfected participants who also reported a higher mortality rate in cirrhotic (14.9 per 1000 PY) than in non-cirrhotic participants (4.8 per 1000 PY). As in previous studies, age at time of SVR, cirrhosis, liver stiffness and GGT increased were associated with an increased risk of overall mortality.^[16,31,32]

For ischaemic events and NLNA cancers, to our knowledge, no other studies estimated their incidence rates or identified factors associated with their occurrence after SVR induced by DAA treatment. In HCV-monoinfected people who did not reach SVR, Chew et al.^[33] estimated a higher incidence of atherosclerotic cardiovascular disease events (13.2 per 1000 PY), whereas in HIV/HCV-coinfected participants who reached SVR after an interferon-based treatment, the incidence was slightly lower (5.2 per 1000 PY), potentially due to younger age.^[34] Finally in HIV-monoinfected participants, the incidence of ischaemic events was lower than that estimated by Ryom et al.^[35] (5.3 per 1000 PY). The incidence of NLNA cancers was close to that estimated by Berenguer et al.^[36] in HIV/HCV-coinfected people who reached SVR after an interferon-based treatment (10.4 per 1000 PY). In this study we identified total cholesterol > 5.0 mmol/L as the only factor associated with an increased risk of ischaemic events. High total cholesterol has been identified as a risk factor for ischaemic events in HCV-infected participants.^[37,38] Previous studies also identified LDL cholesterol as a factor associated with the risk of ischaemic events.^[37,38] Participants with LDL cholesterol > 1.4 mmol/L had increased incidence rates [9.1 (4.9–16.9)] compared with participants with lower values [1.9 (0.3–13.2)]. However, the association was not statistically significant, potentially due to the low number of events observed and a lack of statistical power. The only factor identified as associated with an increased risk of NLNA cancers after SVR was CD4 count < 500 cells/μL. The increased risk of NLNA cancers due to low CD4 count had highly be demonstrated.^[39]

The main limitations of this study are, first, despite a relatively long follow-up of participants after SVR in both groups, we observed few events. This low number of observed events led to a reduced statistical power which could have resulted in the inability to identify other factors associated with risk of the events studied. Due to the low number of events, we chose to use a Poisson model to identify factors associated with the risk of the event, as this model is adapted for rare events. Second, the aim of this study was to identify factors associated with the risk of different events. The statistical methods used have therefore been adapted to identify these factors, but they did not take into account potential confounding factors and could have resulted in the identification of factors reflecting other mechanisms.

The main strength of this study is the size of the studied population and the relatively long follow-up of participants after they reached SVR. Indeed, participants were followed for 3.0 (1.7–3.8) years. This relatively long follow-up allowed us to observe more events. In addition, to our knowledge, no other studies have explored the factors associated with the risk of non-liver-related complications related to HCV after SVR, except for overall mortality, whereas, as observed in our study, these complications are the more frequent.

To conclude, in a population of HIV/HCV-coinfected participants who reached SVR after treatment with a DAA, we observed lower incidence of liver-related or AIDS-defining events than of NLNA events. Severity of liver damage was associated with the risk of liver-related events and overall mortality but not with the risk of non-liver-related complications, whereas CD4 count, reflecting HIV infection control, was associated with risk of liver-related events and NLNA cancers. For ischaemic events, total cholesterol was the only factor associated with an increased risk. The results of this analysis are exploratory and have to be confirmed.

HIV Medicine. 2021;22(9):791-804. © 2021 Blackwell Publishing