| |
Atherosclerotic Cardiovascular Events in Patients Infected With Human Immunodeficiency Virus and Hepatitis C Virus - undetectable HIV viral load reduced risk
|
| |
| |
Download the PDF here
"Undetectable baseline HIV RNA was predictive of a lower risk of coronary and/or cerebral ASCVD events: whereas undetectable baseline viral load (HR 0.41, 95% CI, 0.18-0.96) was associated with coronary and/or cerebral events."
Predictors of ASCVD in our study were similar to those in previous studies [4, 14, 15, 21], underlining the role of traditional CVD risk factors (aging, total cholesterol, and prior CVD events) in HIV and HIV-HCV coinfected populations. As in previous studies [34-36], we found that undetectable baseline HIV RNA was associated with a reduced risk of coronary and/or cerebral ASCVD events. We did not find any association between HCV-related factors and risk of ASCVD events, particularly between DAA treatment and ASCVD events. However, HCV SVR has been associated with a reduction in CVD events in HCV-monoinfected patients in previous studies that also included nonatherosclerotic CVD events [9, 18, 25, 27]. Of note, the definition of CVD events in these studies was broader, including heart failure [9, 26, 27], arrhythmia [9], and cardiac valvulopathy [9], only in cirrhotic patients [9, 26]. In addition, the proportions of patients reaching SVR and treated by DAA were low in our study, which could limit statistical power to identify an association between SVR and ASCVD.
CONCLUSIONS: This large, nationwide, prospective multicenter cohort including HCV-HIV coinfected patients with a long follow-up, showed a high incidence of ASCVD events, especially acute coronary syndrome and PAD. Traditional CVD risk factors were predictive of coronary and/or cerebral ASCVD events, whereas undetectable baseline HIV RNA was protective. PAD ASCVD events, in particular lower limb artery disease, was predominant and requires active diagnosis and intensive management, while controlling modifiable traditional cardiovascular risk factors. HCV-related factors did not appear to be associated with the risk of cardiovascular events. The effect of HCV and DAA on ASCVD events remains to be elucidated
----------------------------
Clinical Infectious Diseases, Volume 72, Issue 9, 1 May 2021 - Boun Kim Tan,1,2,3,a, Mathieu Chalouni,4,a Dominique Salmon Ceron,2,3 Alexandre Cinaud,3,5 Laure Esterle,4 Marc Arthur Loko,4 Christine Katlama,6,7 Isabelle Poizot-Martin,8,9 Didier Neau,10,11 Julie Chas,12 Philippe Morlat,4,11,13 Eric Rosenthal,14,15 Karine Lacombe,7,16 Alissa Naqvi,17 Karl Barange,18 Olivier Bouchaud,19,20 Anne Gervais,21 Caroline Lascoux-Combe,22 Daniel Garipuy,23 Laurent Alric,24,25 Cécile Goujard,26,27 Patrick Miailhes,28 Hugues Aumaitre,29 Claudine Duvivier,30 Anne Simon,31 Jose-Luis Lopez-Zaragoza,32 David Zucman,33 François Raffi,34,35 Estibaliz Lazaro,11,36 David Rey,37 Lionel Piroth,38,39 François Boué,27,40 Camille Gilbert,4 Firouzé Bani-Sadr,41,42 François Dabis,4 Philippe Sogni',3,43,44 Linda Wittkop,4,45,a and Franck Boccara46,47,a; for the ANRS CO13 HEPAVIH study group
Abstract
Background
An increased risk of cardiovascular disease (CVD) was reported in patients coinfected with human immunodeficiency virus (HIV) and hepatitis C virus (HCV), without identifying factors associated with atherosclerotic CVD (ASCVD) events.
Methods
HIV-HCV coinfected patients were enrolled in the Agence Nationale de Recherches sur le Sida et les hépatites virales (ANRS) CO13 HEPAVIH nationwide cohort. Primary outcome was total ASCVD events. Secondary outcomes were coronary and/or cerebral ASCVD events, and peripheral artery disease (PAD) ASCVD events. Incidences were estimated using the Aalen-Johansen method. Factors associated with ASCVD were identified using cause-specific Cox proportional hazards models.
Study Design and Participants
Agence Nationale de Recherches sur le Sida et les hépatites virales (ANRS) CO13 HEPAVIH is a prospective, multicenter, nationwide, observational cohort study in HIV-HCV coinfected subjects that was initiated in France in December 2005. In this analysis, we used the database from this study, censored in December 2016, to identify patients who were chronically infected by HCV at the time of inclusion and had undergone ≥1 year of follow-up. Participants were included from December 2005 and followed through November 2016.
The primary outcome was total ASCVD events, defined as the composite of:
1. Coronary and/or cerebral ASCVD events: any death from CVD related to myocardial infarction or sudden cardiac death with proven coronary artery disease; any acute coronary syndrome (with or without ST-segment elevation myocardial infarction or unstable angina pectoris requiring hospitalization); any coronary revascularization (percutaneous coronary intervention or coronary artery bypass graft); or any ischemic stroke (transient or complete); and/or
2. PAD ASCVD events presumed to be of atherosclerotic origin: symptomatic PAD with or without transluminal arterial angioplasty, arterial bypass graft, or thrombectomy; and subclinical carotid artery stenosis documented by ultrasound (defined as ≥50% lumen stenosis) and aortic atherosclerotic disease (defined as an abdominal aortic aneurysm). Descending thoracic aneurysm and renal or mesenteric artery disease were excluded.
Secondary outcomes were the components of the primary outcome, taken separately. Other adjudicated CVD events were episodes of heart failure requiring hospitalization and venous thromboembolic disease requiring hospitalization.
Results
At baseline, median age of the study population (N = 1213) was 45.4 (interquartile range [IQR] 42.1-49.0) years and 70.3% were men. After a median follow-up of 5.1 (IQR 3.9-7.0) years, the incidence was 6.98 (95% confidence interval [CI], 5.19-9.38) per 1000 person-years for total ASCVD events, 4.01 (2.78-6.00) for coronary and/or cerebral events, and 3.17 (2.05-4.92) for PAD ASCVD events.
Aging (hazard ratio [HR] 1.06; 95% CI, 1.01-1.12), prior CVD (HR 8.48; 95% CI, 3.14-22.91), high total cholesterol (HR 1.43; 95% CI, 1.11-1.83), high-density lipoprotein cholesterol (HR 0.22; 95% CI, 0.08-0.63), statin use (HR 3.31; 95% CI, 1.31-8.38), and high alcohol intake (HR 3.18; 95% CI, 1.35-7.52) were independently associated with total ASCVD events, whereas undetectable baseline viral load (HR 0.41, 95% CI, 0.18-0.96) was associated with coronary and/or cerebral events.
Conclusions
HIV-HCV coinfected patients experienced a high incidence of ASCVD events. Some traditional cardiovascular risk factors were the main determinants of ASCVD. Controlling cholesterol abnormalities and maintaining undetectable HIV RNA are essential to control cardiovascular risk.
Early and widespread access to antiretroviral therapy has contributed to a decrease in fatal events related to AIDS, improving the lifespan of people living with human immunodeficiency virus (HIV) (PLWH) [1, 2]. This increase in longevity has contributed to the transformation of HIV infection into a chronic disease, at a cost of an excess of noncommunicable diseases [1] such as cardiovascular disease (CVD) [1-4], and leading to increased morbimortality in PLWH [2, 5]. The pathophysiology of HIV-related atherosclerosis in PLWH on antiretroviral therapy is multifactorial [2]. Many traditional CVD risk factors affect PLWH [1-4], including arterial proinflammatory retrovirus activity [6], immune activation [1], drug toxicity [7], and disturbed lipid metabolism [7]. Hepatitis C virus (HCV) coinfection affects 6.2% of PLWH [8], and increases the risk of liver-related events and death [9]. Chronic HCV infection also contributes to atherosclerosis [10-12], myocardial infarction [13], and other CVD events [9]. Several studies have thus highlighted an increased risk of CVD events with HIV-HCV coinfection [14-19], especially compared with HIV alone [3, 4, 17, 20]. This coinfection may act synergistically to induce endothelial dysfunction [21], vascular inflammation [22], and insulin resistance [9]. Meanwhile, the advent of direct-acting antivirals (DAA) in 2014 has revolutionized the care of patients with HCV [23], allowing the achievement of a sustained virologic response (SVR) for almost all treated patients [24]. Hence, HCV SVR could decrease CVD events in HCV monoinfected patients [25-27] and in HIV-HCV coinfected patients [18]. However, in HIV-HCV coinfected patients, factors related to the occurrence of atherosclerotic CVD (ASCVD) have not yet been fully identified. Consequently, we sought to describe the incidence of total ASCVD events (including coronary, cerebrovascular, and peripheral artery disease [PAD]) and identify predictors of such events in this population.
RESULTS
Baseline Characteristics
The study population comprised 1213 patients (Figure 1). Baseline characteristics are reported in Table 1. Median age was 45.4 (IQR 42.1-49.0) years and 70.3% were men. The prevalence of current smoking was 70.2%, cirrhosis 18.9%, diabetes 5.9%, and prior CVD 2.7%; 4.1% were on statins. Most of the patients (68.8%) were on HIV antiretroviral therapy.
DISCUSSION
In this large, nationwide, prospective cohort of HIV-HCV coinfected patients in the ANRS CO13 HEPAVIH cohort, we observed a high incidence of ASCVD events, with similar incidences of coronary artery disease and PAD ASCVD events. Multivariable analyses demonstrated that some traditional risk factors (lipids and prior CVD) were predictive of total ASCVD events. Undetectable baseline HIV RNA was predictive of a lower risk of coronary and/or cerebral ASCVD events.
The high incidence of total ASCVD events observed in our cohort (6.98 per 1000 person-years) is in line with a previous study [7]. Other studies focused on either all or specific CVD events in PLWH [3, 4, 14-18]. Some of these studies showed that chronic HCV infection could promote CVD in PLWH without focusing on atherosclerosis [4, 14, 17, 18]. In the international Data Collection on Adverse Events of Anti-HIV drugs (D:A:D) cohort of PLWH, in which 19% were coinfected by HCV, the incidence of ASCVD was 5.34 per 1000 person-years (from January 2009 to February 2016), based on a composite outcome that included myocardial infarction, sudden cardiac death, stroke, and invasive cardiovascular procedures (coronary bypass, coronary angioplasty, and carotid endarterectomy) [7]. Compared with the D:A:D composite outcome, our definition of total ASCVD events also included PAD ASCVD events. Our PAD events predominantly included symptomatic lower limb artery disease (17 of 20, 85%).
How HCV could confer an increased risk of lower limb artery disease is unclear [10]. Epidemiologically, HCV appeared to be a specific risk factor associated with PAD in a recent study in PLWH [20], but the pathophysiology of HCV-related vasculopathy remains an issue. Few data are available regarding the descriptions of PAD in PLWH [20, 31] or with HIV-HCV coinfection [18]. Our incidence of PAD (3.17 per 1000 person-years) is lower than in the North American Veterans Aging Cohort Study (VACS) (11.9 per 1000 person-years) [20], but higher than in the Spanish nationwide HIV-HCV cohort Grupo de Estudio de la Sociedad Española de Enfermedades Infecciosas y Microbiología Clínica (GESIDA; 1.1 per 1000 person-years) [18]. This may be because only administrative data were used to identify PAD in the VACS cohort [20], whereas inpatient and outpatient medical files were used to confirm PAD in our study and in the GESIDA cohort [18]. In addition, there were different proportions of smokers, patients with diabetes, and illicit drug users (70.2%, 5.9%, and 5.5%, respectively, in our study vs 70.3%, 7.5%, and 18% in the VACS cohort [4, 14, 20], and 2.3% for diabetes and 80.4% for illicit drug use in GESIDA [18]). Again, nadir CD4 cell count could have also influenced the incidence of PAD, which was 119 cells/µL in VACS [20], 150 cells/µL in ANRS CO13 HEPAVIH [24], and 203 cells/µL in GESIDA [18]. Intravenous drug or cocaine use and cryoglobulinemia did not appear to be associated with PAD, even though they have been previously described [32, 33].
Surprisingly, our incidence of stroke (0.79 per 1000 person-years) was lower than in D:A:D [7] or VACS [4] (1.8 and 2.2 per 1000 person-years, respectively). This could be explained by a worse cardiovascular risk profile in VACS [4] and GESIDA [18] (with double the prevalence of hypertension and diabetes) compared with patients in the present study. Moreover, we defined ischemic stroke-as part of our composite outcome of total and coronary and/or cerebral ASCVD events-whereas other cohorts included ischemic and nonischemic strokes and some liver-related encephalopathies [15, 17, 20].
Predictors of ASCVD in our study were similar to those in previous studies [4, 14, 15, 21], underlining the role of traditional CVD risk factors (aging, total cholesterol, and prior CVD events) in HIV and HIV-HCV coinfected populations. As in previous studies [34-36], we found that undetectable baseline HIV RNA was associated with a reduced risk of coronary and/or cerebral ASCVD events. We did not find any association between HCV-related factors and risk of ASCVD events, particularly between DAA treatment and ASCVD events. However, HCV SVR has been associated with a reduction in CVD events in HCV-monoinfected patients in previous studies that also included nonatherosclerotic CVD events [9, 18, 25, 27]. Of note, the definition of CVD events in these studies was broader, including heart failure [9, 26, 27], arrhythmia [9], and cardiac valvulopathy [9], only in cirrhotic patients [9, 26]. In addition, the proportions of patients reaching SVR and treated by DAA were low in our study, which could limit statistical power to identify an association between SVR and ASCVD.
This study was not designed to identify causal associations, rather it aimed to identify independent predictors of ASCVD events. Statin use was identified as a predictor of ASCVD events in multivariable analysis; this may reflect indication bias, as statins might have been prescribed because of the presence of traditional cardiovascular risk factors. Our relatively low rate of statin use (4.1%) is in line with other international, multicenter cohorts [37, 38], indicating underuse of statins in primary and secondary prevention. However, statin use has been associated with improved virologic response rates to HCV treatment [39]. No data on antiplatelet drugs were available, which may account for the high incidence of PAD ASCVD events, including lower limb artery disease. Moreover, given the absence of a control group, the implication of HCV infection in the development of PAD ASCVD events is purely speculative. Current smoking was not associated with ASCVD events in our multivariable analyses, whereas high alcohol intake was, especially for PAD ASCVD events. A confusion bias may explain this result by simultaneous addiction and the high proportion of current smokers (70.2%) compared with those who have never smoked (20%) [9, 25]. The number of patients on DAA medications may not have been sufficient to determine an effect of HCV SVR in reducing the incidence of ASCVD at the end of the inclusion period in 2016. Consequently, follow-up after SVR may have been too short to observe a reduction in the incidence of ASCVD, as previously [40]. The present cohort is young (median age 45.4 years), which may explain the over representation of STEMI versus NSTEMI, which is usually predominant in older subjects. However, we could not exclude the possibility that NSTEMI related to type II myocardial infarction (nonobstructive coronary artery disease with troponin elevation was not included in the present study) or patients who were not hospitalized could have been missed. Finally, the relatively small number of ASCVD events could have resulted in unidentified predictive factors because of low statistical power.
CONCLUSIONS
This large, nationwide, prospective multicenter cohort including HCV-HIV coinfected patients with a long follow-up, showed a high incidence of ASCVD events, especially acute coronary syndrome and PAD. Traditional CVD risk factors were predictive of coronary and/or cerebral ASCVD events, whereas undetectable baseline HIV RNA was protective. PAD ASCVD events, in particular lower limb artery disease, was predominant and requires active diagnosis and intensive management, while controlling modifiable traditional cardiovascular risk factors. HCV-related factors did not appear to be associated with the risk of cardiovascular events. The effect of HCV and DAA on ASCVD events remains to be elucidated.
|
|
| |
| |
|
|
|
