| |
Is the Risk of Myocardial Infarction in People With Human Immunodeficiency Virus (HIV) Associated With Atazanavir or Darunavir? A Nested Case-Control Study Within the French Hospital Database on HIV
|
| |
| |
Download the PDF here
Download the PDF here
We found no significant association between the risk of MI in HIV-infected individuals and exposure to atazanavir or darunavir. We used several approaches to minimize biases including the choice of a case-control design nested within the FHDH cohort [14], use of controls matched for age and sex, which are 2 important risk factors for MI [18], as well as adjustment for both HIV-related parameters and classic MI risk factors. These classic risk factors were associated with the risk of MI in our study, supporting the reliability of our data. We did not include recurrent MI to avoid a possible selection bias, because the treatment choice could be influenced by a previous MI. Of note, however, durations of atazanavir and darunavir exposure in those exposed were relatively short with mean duration of 2.1 years and 1.2 years, respectively, with somewhat larger CIs for darunavir. Therefore, we cannot exclude that longer duration of exposure would lead to different results.
---------------------------
Editorial - Darunavir and Cardiovascular Risk: Evaluating the Data to Inform Clinical Care
JID 15 February 2020 - Virginia A. Triant1,2,3,4, and Mark J. Siedner1,3,4,5
1Division of Infectious Diseases, Massachusetts General Hospital, Boston, Massachusetts, USA, 2Division of General Internal Medicine, Massachusetts General Hospital, Boston, Massachusetts, USA, 3Mongan Institute, Massachusetts General Hospital, Boston, Massachusetts, USA, 4Harvard Medical School, Boston, Massachusetts, USA, 5Africa Health Research Institute, KwaZulu-Nata, South Africa
see main article below
People with human immunodeficiency virus (PWH) confront an increased risk of aging-associated chronic diseases including cardiovascular disease (CVD) [1]. Studies have delineated a complex, multifactorial mechanism contributing to increased CVD risk in PWH [1]. Understanding the factors that contribute to this risk-particularly those unique to human immunodeficiency virus (HIV)-is important to enable HIV caregivers to better estimate risk, prevent and manage disease, and improve health for PWH.
Antiretroviral therapy (ART) has been implicated as contributing to CVD risk in PWH. Historically, protease inhibitors (PIs) used in earlier treatment eras have been associated with an increased risk of myocardial infarction (MI) that increases with cumulative use of the drugs and is explained only in part by dyslipidemia [2]. The nucleoside reverse-transcriptase inhibitor abacavir has also been associated with MI risk in some studies; risk is increased with recent use, and it appears to abate after discontinuation, in contrast to cumulative use with the PI class [3]. In addition, contemporary PIs have now been assessed, with a Data Collection on Adverse events of Anti-HIV Drugs (D:A:D) study demonstrating increased MI risk with exposure to darunavir but not to atazanavir [4].
Cardiovascular risk associated with PIs remains highly relevant in the contemporary treatment era. Although European and US HIV treatment guidelines favor integrase-strand transfer inhibitor (INSTI)-based regimens for initial ART in ART-naive PWH, the European AIDS Clinical Society guidelines include darunavir in recommended regimens; older PIs, including atazanavir, are in wide use in lower-resource settings as standard, second-line therapy as per World Health Organization guidelines. Moreover, recent data showing weight gain after INSTI initiation suggest that some medications in this increasingly used class may have adverse metabolic effects that will need to be considered when selecting ART regimens [5].
In this issue of the Journal of Infectious Diseases, Costagliola et al investigated the risk of MI associated with darunavir and atazanavir in the large, well established, multicenter French Hospital Database on HIV (FHDH-ANRS CO4). The study was a nested case-control study of PWH who had an MI event between 2006 and 2012, with the study period selected to reflect PI availability in France. Of 81 294 eligible patients, 408 had an MI; 1250 controls without MI and followed contemporaneously were matched to the cases by age and sex. Myocardial infarction was identified initially by International Classification of Diseases (ICD)-10 code I21, with each event adjudicated by a cardiologist. Antiretroviral (ARV) drug exposure was defined as cumulative exposure. Models were built adjusting for ARVs with or without confounding covariates and replicated limiting each model to include only the ARVs used in the D:A:D models (abacavir, lopinavir-ritonavir, and indinavir). They found no significant association between either darunavir or atazanavir and odds of MI in any of the models (darunavir adjusted odds ratio [aOR] = 0.51, 95% confidence interval [CI] = .11-2.32, atazanavir aOR = 1.54, 95% CI = .87-2.73 in fully adjusted models accounting for all other ARVs and confounders not on the causal pathway).
The absence of a significant association of darunavir with MI in the FHDH study represents an apparent discordance with the results of the recent, large D:A:D study, which demonstrated a significant, approximately 50% increased risk of MI per 5 years of darunavir exposure. Although the results of the present study do not support the findings of the D:A:D study, they also do not contradict the results. As noted in the discussion of the FHDH study, the CI was wide and, notably, encompassed the entirety of the D:A:D study CI (FHDH CI = .11-2.32, D:A:D CI = 1.13-2.02), suggesting that even though the estimates were qualitatively different, the margin of error encompassed both negative and positive associations with the outcome. Despite a large cohort with ample follow-up time, the median duration of darunavir exposure in the FHDH study was only 1 year for MI cases and 1.2 years for controls without MI, and the number of patients exposed to darunavir was relatively low (10% of cases and 9% of controls). Studies examining the association of PIs on MI risk have consistently shown increased risk to be associated with increased duration of exposure, including in prior FHDH studies [6], prior D:A:D studies [2], and the recent D:A:D study [4], which compared duration of less than 1 year to more than 6 years. It is possible that with longer duration of darunavir exposure in the current study, and a larger sample size of individuals exposed to it, a qualitatively different or statistically significant effect would be observed.
The DAD and FHDH studies were similar in many aspects, including baseline characteristics, HIV-related parameters, and CVD risk profiles of the cohorts. Events in both studies were centrally validated or adjudicated. The D:A:D study used a broader outcome definition (MI, stroke, sudden cardiac death, or invasive cardiovascular procedures), but in sensitivity analyses with the definition limited to MI, results were similar to the original analyses, with a statistically significant increased MI risk with darunavir. Although modeling strategies differed, including with regards to how to handle covariates on the causal pathway from ARV exposure to MI (included at baseline in D:A:D but only in sensitivity analyses in FHDH), the FHDH study included models that replicated the ARVs included in the D:A:D study. The FHDH study did not report CVD risk profiles by regimen, but confounding by indication could have influenced results in either cohort. Nonetheless, the results of the D:A:D study were published after the observation period of the FHDH study, and the D:A:D study tested several interactions that did not suggest confounding by indication.
There were some important differences that could further explain the contrasting results. Factors that differed between the studies included geographic distribution, study period (2006-2012 for FHDH vs 2009-2016 for D:A:D), method of MI outcome ascertainment (ICD code for FHDH vs clinical variables from standardized data collection forms for D:A:D), median duration of darunavir exposure (1 year for FHDH vs 2.56 years for D:A:D), and inclusion of bilirubin as a covariate (included in the D:A:D study). Moreover, the FHDH study imputed values for parameters missing for fewer than 50% of individuals, which has the potential to misestimate confounding factors. Importantly, given the matching approach of incidence density sampling used, the odds ratio from the FHDH study can be compared with the incidence rate ratio in the D:A:D study.
Additional reports are conflicting with regard to the potential cardiovascular effects of darunavir. Darunavir has been associated with worsening of CVD risk factors and surrogate markers of atherosclerosis in some other investigations [7-9]. Lipid alterations are known to occur in the setting of darunavir use [10]; yet, in the D:A:D study, the effect of darunavir on MI was not modified by dyslipidemia. Several published studies have failed to show a significant association of darunavir with MI [11, 12]. A study of Janssen-sponsored clinical trials, postmarketing pharmacovigilance data, and administrative claims data showed that rates of a composite CVD outcome did not increase with increasing yearly intervals of exposure to darunavir, but the study was limited by median trial duration of 2 years and missing data in some trials [11]. In contrast to darunavir, prior studies have consistently failed to show increased CVD risk with atazanavir use or have shown decreased risk compared with other regimens [13, 14]. Atazanavir appears to be the outlier with regards to CVD risk in the PI class, and its apparent protective association has been hypothesized to result from antioxidant and anti-inflammatory effects of unconjugated bilirubin [15-18].
Reviewing the findings of the FHDH and D:A:D studies underscores several factors important to consider in interpretation of longitudinal observational cohorts. In contrast to randomized controlled trails, which lend strong support to causal relationships, questions addressed by cohort studies lend less support for a causal association due both unmeasured and residual confounding. For these study designs, findings are bolstered through reproducibility, when multiple studies address the same question. In cases such as the D:A:D study and FHDH, when results are discordant-or at least not concordant-and a clear methodological explanation does not explain the differing results, it is reasonable to maintain a conservative approach while awaiting further data.
In summary, this important study by Costalgiola et al underscores the complexity of mitigating CVD risk in HIV and of interpreting observational cohort studies. Given the wide CIs reported on the relationship between darunavir use and MI, and contrasting results with other similar studies, additional studies with longer follow-up time and more events in darunavir-exposed individuals may offer further clarification. Until that time, it is plausible that CVD risk increases with darunavir exposure, and this should be considered as a factor during selection of ARVs. Specifically, prolonged darunavir use could be considered as an HIV-Related CVD Risk-Enhancing factor per the American Heart Association statement on HIV and CVD [1], and clinicians might consider selecting an alternate ARV in PWH at high underlying CVD risk. While we await further data to help clarify this issue, supported practices to reduce CVD risk, such as achieving and maintaining virologic suppression, careful attention to risk factor management, and high suspicion for CVD risk among PWH, should continue to drive treatment decisions for this high-risk population.
--------------------------------
Is the Risk of Myocardial Infarction in People With Human Immunodeficiency Virus (HIV) Associated With Atazanavir or Darunavir? A Nested Case-Control Study Within the French Hospital Database on HIV
Dominique Costagliola,1, Valérie Potard,1,15 Sylvie Lang,14 Nathalie de Castro,2 Laurent Cotte,3 Xavier Duval,4 Claudine Duvivier,5 Sophie Grabar,1,6
Murielle Mary-Krause,1 Marialuisa Partisani,7 Sylvie Ronot-Bregigeon,8 Anne Simon,9 Pierre Tattevin,10, Laurence Weiss,11 David Zucman,12
Christine Katlama,1,9 François Raffi,13 and Franck Boccara14; on behalf of ANRS CO4 FHDH
Abstract
Background
The Data Collection on Adverse Events of Anti-HIV Drugs (DAD) study has reported an increased risk of cardiovascular diseases in people with human immunodeficiency virus who were exposed to darunavir (DRV) but not to atazanavir (ATV). Our objective was to evaluate associations between ATV or DRV exposures and the risk of myocardial infarction (MI) in a nested case-control study within ANRS-CO4 French Hospital Database on HIV (FHDH).
Methods
Cases were individuals who had a first validated MI between 2006 and 2012. Up to 5 controls were selected at random with replacement among individuals with no history of MI, followed at the time of MI diagnosis, and matched for age and sex. Conditional logistic regression models were used to adjust for potential confounders (MI risk factors and HIV-related parameters) and for cumulative exposure to each antiretroviral drug (ARV).
Results
Overall, 408 MI cases and 1250 controls were included: 109 (27%) cases and 288 (23%) controls had been exposed to ATV, and 41 (10%) cases and 107 (9%) controls had been exposed to DRV. There was no significant association between exposure to ATV (adjusted odds ratio [OR] = 1.54; 95% confidence interval [CI], .87-2.73) or DRV (adjusted OR = 0.51; 95% CI, .11-2.32) and the risk of MI.
Conclusions
In FHDH, exposures to ATV or to DRV were not significantly associated with the risk of MI, adjusting for complete ARV history, contrary to the analysis in DAD.
People with human immunodeficiency virus (PWH) have a higher incidence of myocardial infarction (MI) compared with the general population [1-4]. The use of first-generation protease inhibitors (PIs)-indinavir, lopinavir, and amprenavir/fosamprenavir-has been associated with an increased risk of MI [5, 6], but today these PIs are no longer recommended, at least in high-income countries [7]. More recent PIs, atazanavir and darunavir, are potent antiretroviral drugs (ARVs) with more favorable lipid profiles than lopinavir [8, 9]. The Data Collection on Adverse Events of Anti-HIV Drugs (DAD) study, an international PWH multicohort study, has reported an association between exposure to darunavir and the risk of cardiovascular disease (CVD) in PWH, whereas no association was evident with atazanavir [10]. This analysis accounted for exposure to only 3 drugs: lopinavir, indinavir, and abacavir. Earlier analyses in DAD have found these drugs to be associated with an increased risk of CVD [5]. In contrast, a study based on pooled data from 19 Janssen-sponsored clinical trials, postmarketing, and epidemiological data did not suggest an important risk of CVD for users of darunavir [11]. Finally, a cohort study found a reduced risk of CVD with atazanavir versus other PIs including darunavir among treatment-naive PWH [12]. In this context of discordant results, and thanks to the large French Hospital Database on HIV (FHDH ANRS CO4), we aimed to evaluate associations between exposure to atazanavir or darunavir and the risk of MI in a nested case-control study within FHDH.
METHODS
The French Hospital Database on HIV
The FHDH is a hospital-based, open multicenter cohort, and inclusion have been ongoing since 1989 [13]. Individuals are eligible if they have documented human immunodeficiency virus (HIV)-1 or HIV-2 infection and give written informed consent to participate. Data are collected prospectively on standardized forms, which include demographic characteristics, the date and type of clinical events recorded according to the International Disease Classification (ICD), antiretroviral therapy (ART), and biological markers. The FHDH project has been approved by the French data protection authority (Commission National de l’Informatique et des Libertés on November 27, 1991, Journal Officiel, January 17, 1992).
Study Design
We conducted a case-control study nested in the FHDH and restricted to HIV-1 infected individuals. We chose this approach because of the time-varying nature of ARV use, the large size of the cohort, and the long duration of follow-up. Moreover, compared with a full cohort approach using a survival analysis with time-dependent variables, a nested case-control analysis provides estimates of odds ratios (ORs) from conditional logistic regression models that are unbiased estimates of relative risk [14]. We focused on the 2006-2012 period because atazanavir was available in France beginning in 2002 and darunavir was available beginning in 2006, initially through expanded access programs. During this period, there were 81 294 PWH with at least 1 follow-up visit in the FHDH.
Case Definition
Cases comprised individuals enrolled in the FHDH who had a first prospectively reported MI between January 2006 and December 2012. The diagnosis of MI (ICD I21) was confirmed by a cardiologist (F.B.) who was provided with cardiac signs and symptoms, troponin and/or creatinine kinase levels, and electrocardiographic findings, as recorded in the medical records. We used the American College of Cardiology/European Society of Cardiology definition [15]. Only definite and probable cases of MI and possible death from MI were included. The index date was the date of MI diagnosis.
Selection of Controls
Up to 5 controls enrolled in the FHDH were randomly selected using the incidence density sampling method [16], among individuals with no history of MI, who were under follow-up at the time of MI diagnosis in the corresponding case (±6 months) in the same clinical center, and matched for sex and age (±3 years).
Potential Confounders
It was important to consider pre-existing risk factors for MI that might have influenced the choice of ARV during the study period. Therefore, we explored traditional risk factors unlikely to be on the causal pathway between PI exposure and the risk of MI: body mass index (BMI) (<21/21 ≤ BMI < 24/24 ≤ BMI < 27/≥27), smoker status (no/yes), family history of premature coronary artery disease (CAD), hypertension or antihypertensive treatment (no/yes), current cocaine and/or intravenous drug use (no/yes). We also studied the potential effect of (1) geographic origin (sub-Saharan Africa/other) and (2) the following HIV-related variables on the risk of MI: prior acquired immune deficiency syndrome (AIDS) status, CD4 T-cell nadir, CD4 T-cell count, CD8 T-cell count, CD4 to CD8 T-cell ratio, and viral load (VL). Hypercholesterolemia or hypertriglyceridemia or lipid-lowering treatment and diabetes or antidiabetic treatment, which could be on the causal pathway between ARV exposure and the risk of MI, were collected to be used in sensitivity analyses if a significant association was found between exposure to any ARV and the risk of MI.
Data Collection
The date of MI diagnosis (for cases), sex, age, geographic origin, HIV parameters, and a detailed history of prescribed ARV up to the index date were extracted from the FHDH and validated in the medical records by trained medical assistants using a predefined case report form for both cases and controls. We also collected smoking status, family history of premature CAD, hypertension, antihypertensive medication, diabetes, antidiabetic treatment, hypercholesterolemia (including lipid-lowering treatment), and hypertriglyceridemia from the medical records. All biological parameters were measured within 3 months before the date index.
Statistical Analysis
Characteristics of cases and controls were compared by using univariable conditional logistic regression. Exposure to each ARV was considered as the cumulative duration of exposure per 5 years up to the index date. Conditional logistic regression models were used to quantify the relation between exposure to atazanavir or darunavir and the risk of MI. In a first multivariable model, the exposures of all ARVs were included. In a second multivariable model, we included exposure to all drugs plus the potential confounders. These confounders were classic MI risk factors, geographic origin, and HIV-related variables that differentiated cases and controls. In a sensitivity analysis, diabetes or antidiabetic treatment, hypercholesterolemia (including use of lipid-lowering drugs), and hypertriglyceridemia were added in the multivariable model. To replicate an analysis that is close to that conducted by the DAD study, we performed a sensitivity analysis accounting only for exposures to lopinavir, indinavir, and abacavir. Because parameters with missing data can influence the results, all values missing for fewer than 50% of individuals were replaced by using a multiple imputation method, and missing values were randomly sampled from their predicted distributions [17]. Ten sets of imputations were used to create 10 complete datasets. All 10 datasets were analyzed and combined with Rubin’s rules. SAS software (version 9.4; SAS Institute Inc., Cary, NC) was used for all statistical analyses.
RESULTS
Baseline Characteristics of Participants
Overall, 408 MI cases were validated and matched to at least 1 control (1250 controls, mean of 3 controls per case). Characteristics of cases and controls are shown in Table 1. The median year of MI diagnosis was 2008 (interquartile range [IQR], 2007-2009). At enrollment in FHDH, 81% of participants were ARV naive. In the case population, median age was 49 years and 88% of individuals were men. At the index date, the cases were different from the controls on a majority of HIV-related factors including a lower CD4 nadir, a higher CD8 level, a higher proportion of cases with VL >50 copies/mL, and with a history of AIDS-defining event, and those less likely to be from sub-Saharan origin. The classic MI risk factors were more frequent among the cases, except for obesity. The individuals with BMI <21 kg/m2 were more frequent among the cases.
People with human immunodeficiency virus (PWH) have a higher incidence of myocardial infarction (MI) compared with the general population [1-4]. The use of first-generation protease inhibitors (PIs)-indinavir, lopinavir, and amprenavir/fosamprenavir-has been associated with an increased risk of MI [5, 6], but today these PIs are no longer recommended, at least in high-income countries [7]. More recent PIs, atazanavir and darunavir, are potent antiretroviral drugs (ARVs) with more favorable lipid profiles than lopinavir [8, 9]. The Data Collection on Adverse Events of Anti-HIV Drugs (DAD) study, an international PWH multicohort study, has reported an association between exposure to darunavir and the risk of cardiovascular disease (CVD) in PWH, whereas no association was evident with atazanavir [10]. This analysis accounted for exposure to only 3 drugs: lopinavir, indinavir, and abacavir. Earlier analyses in DAD have found these drugs to be associated with an increased risk of CVD [5]. In contrast, a study based on pooled data from 19 Janssen-sponsored clinical trials, postmarketing, and epidemiological data did not suggest an important risk of CVD for users of darunavir [11]. Finally, a cohort study found a reduced risk of CVD with atazanavir versus other PIs including darunavir among treatment-naive PWH [12]. In this context of discordant results, and thanks to the large French Hospital Database on HIV (FHDH ANRS CO4), we aimed to evaluate associations between exposure to atazanavir or darunavir and the risk of MI in a nested case-control study within FHDH.
Impact of Atazanavir or Darunavir Exposure on the Risk of Myocardial Infarction
As shown in figure 1, no association was found between atazanavir exposure or darunavir exposure and the risk of MI in all the models including the sensitivity analyses. For atazanavir, the univariable OR was 1.32, the OR adjusted for exposure to other ARVs was 1.39, and the OR in the fully adjusted model was 1.54. The corresponding figures in analyses accounting only for exposure to abacavir, indinavir, and lopinavir were 1.27 and 1.29. For darunavir, the univariable OR was 1.14, whereas the OR adjusted for exposure to other ARVs was 0.61, and the OR in the fully adjusted model was 0.51. The corresponding figures in analyses accounting only for exposure to abacavir, indinavir, and lopinavir were 0.86 and 0.79. It is interesting to note that the OR for darunavir was smaller in the analysis accounting for complete treatment history than in the analysis similar to DAD, ie, accounting only for exposure to lopinavir, indinavir, and abacavir, whereas the reverse was observed for atazanavir. In the models including diabetes, hypercholesterolemia, and hypertriglyceridemia, the ORs were 1.22 (95% confidence interval [CI], .68-1.82) for atazanavir exposure and 0.44 (95% CI, .09-2.05) for darunavir exposure.
Discussion
We found no significant association between the risk of MI in HIV-infected individuals and exposure to atazanavir or darunavir. We used several approaches to minimize biases including the choice of a case-control design nested within the FHDH cohort [14], use of controls matched for age and sex, which are 2 important risk factors for MI [18], as well as adjustment for both HIV-related parameters and classic MI risk factors. These classic risk factors were associated with the risk of MI in our study, supporting the reliability of our data. We did not include recurrent MI to avoid a possible selection bias, because the treatment choice could be influenced by a previous MI. Of note, however, durations of atazanavir and darunavir exposure in those exposed were relatively short with mean duration of 2.1 years and 1.2 years, respectively, with somewhat larger CIs for darunavir. Therefore, we cannot exclude that longer duration of exposure would lead to different results.
In the univariable analysis of the DAD study over the period 2009 to 2016, the univariable incidence rate ratio (IRR) of CVD event per 5 years of darunavir exposure was 1.93 (95% CI, 1.63-2.28) [10], whereas in our study over 2006 to 2011, the univariable OR of MI was 1.14 (95% CI, 0.36-3.59) per 5 years of darunavir exposure. Given our study design (a case-control study nested in a cohort study with incidence density sampling for the selection of controls), the OR provides an estimate of the IRR that is estimated in the DAD study, and one can directly compare the 2 parameters [19]. In both studies, the multivariable analyses accounting for potential confounders allowed us to reduce the IRR of CVD or MI associated with darunavir exposure; however, in the DAD study, cumulative use of darunavir remained associated with an increased risk of CVD. The discrepancy between the 2 studies could be explained by differences in the definitions of the event and/or in the way of accounting for potential confounders and/or by a problem of under notification of CVD diagnosis in DAD. The DAD study assessed the risk of all cardiovascular events including MI, strokes, sudden cardiac deaths, and invasive cardiovascular procedures. However, in DAD, when the analysis was restricted to MI only, the association with darunavir exposure remained significant (multivariable IRR = 1.51; 95% CI, 1.13-2.02). In terms of confounders, the DAD investigators chose to account for exposure to only 3 ARVs (abacavir, lopinavir, and indinavir). This is a potential important bias, because exposure to any antiretroviral may influence more recent exposure and therefore confound the association. It is a strength of our study to have accounted for the complete antiretroviral history. It would be interesting to see whether, similar to our analysis, accounting for complete treatment history in DAD would reduce the risk for darunavir. In the DAD study, the events are reported electronically on specific forms and are centrally validated. However, because centers reporting MI are those caring for HIV infection, some MI case may not have been reported. There is potential for selection bias if physicians tend to more systematically report MI cases exposed to darunavir because of its known impact on lipid profile. In our study, a similar issue may arise, but we checked that control are truly control with no MI prior to the index date and found that 4 of those selected to be control had had an MI and were therefore not elligible to be a control (no MI reported in the database) and eventually found to have had MI, and these were considered as cases in the analysis. The relatively small number of MI in DAD over 2009-2016 (n = 432 with no prior CVD over a median follow-up of 7.0 years in 35 711 participants) compared with our study over 2006-2011-whereas DAD is conducted in countries that tend to have a higher incidence of MI compared with France-is an indication that this phenomenon of underreporting might be present. Our study design has the advantage to permit the prospective collection of MI cases from all FHDH participating centers, as well as to validate the cases as cases, but also to validate that the controls did not had an MI, which would be very costly in a cohort approach. Finally, one should note that although the direction of the association is different between the 2 studies, the CI for darunavir exposure in the DAD study is completely enclosed in the CI of the present study, and it is possible that the 2 studies are not discrepant.
The study conducted by Janssen, which contains analyses of 19 Janssen-sponsored clinical trials, analyses of postmarketing pharmacovigilance database, and analyses of administrative claim database, does not suggest that a CVD should be considered an important risk for users of darunavir [11]. Regarding the analysis in the study of the US Veterans [12], the authors compared the risk of cardiovascular outcomes in naive PWH initiating ART with a regimen including either atazanavir (n = 1529) or other PIs (n = 2053), in the majority PIs known to be associated with an increased risk of CVD (lopinavir [n = 1087], indinavir [n = 98], and fosamprenavir [n = 169] or amprenavir [n = 11]; for a total of 1365 [66.5%]), and only a limited number of darunavir users (n = 424, 20.7%). Given that, it is difficult to conclude on the risk associated with darunavir exposure from this study.
Conclusions
In conclusion, in this large, case-control study nested within FHDH, we found no evidence of excess risk of MI after exposure to darunavir or atazanavir, which is reassuring because PIs remain widely used in real-life management of PWH.
|
|
| |
| |
|
|
|
