| |
Impact of atazanavir-based HAART regimen on the carotid intimaĞmedia thickness of HIV-infected persons: a comparative prospective cohort - pdf attached
|
| |
| |
Download the PDF here
AIDS: 27 November 2010. Concise Communication
de Saint-Martin, Luca,b,c; Bressollette, Luca,c,d; Perfezou, Pascalec,e; Bellein, Veroniquea,b,c; Ansart, Severinea,c; Vallet, Sophiea,c,f; Pasquier, Elisabetha,b,c
aCHRU Brest, France
bUniversite de Bretagne Occidentale, Faculte de Medecine et des Sciences de la Sante, EA3878, IFR148, Brest, France
cUniversite Europeenne de Bretagne, France
dLaTIM Ğ INSERM U650, Brest, France
eCHIC, Quimper, France
fUniversite de Brest, Faculte de Medecine et des Sciences de la Sante, EA3882, IFR148, Brest, France.
"The cIMT, a well known and validated surrogate marker of atherosclerosis, decreased significantly among patients on the HAART regimen including ATV/r, whereas a slight increase is observed in the control group. This difference in the cIMT course appeared after 6 months of follow-up, when the cases had been ATV/r exposed since 289 days.....However, this study has some limitations (see text of article).....the study was not randomized.....this modification in the HAART regimen could have been brought about by diet or physical activity recommendations, two relevant parameters, which were not collected in our study..... if an interventional correction of the cardiovascular risk is necessary, the results of the AVATAR study favor a switch to a low cardiovascular impact HAART regimen, like HAART including ATV/r, which amounts to the usual cardiovascular prevention. Of course, these data must be confirmed by further randomized studies."
Abstract
Objective: With the advent of highly active antiretroviral therapy regimens, it is crucial to consider their long-term benefits to risk ratios among HIV-infected persons. The impact of protease inhibitors on the cardiovascular risk is controversial.
Design: This observational cohort was designed to investigate the cardiovascular impact of boosted atazanavir (ATV/r), a protease inhibitor that does not provide major dyslipidemia or insulin resistance.
Setting: This study was carried out at the University Hospital of Brest (France).
Patients: Among the 229 HIV-infected persons of the cohort, 33 cases treated by ATV/r-containing regimen since less than 6 months were compared to 99 age-matched and sex-matched ATV/r naive controls.
Intervention: None.
Main outcome measure: The main outcome measure was carotid intimaĞmedia thickness (cIMT) at the baseline, 6, 12, and 18 months.
Results: Although the cIMT was not different at inclusion (0.633 ± 0.05 vs. 0.666 ± 0.09, P = 0.07), the cIMT course significantly decreased (P = 0.018) in cases at 18 months. The differences remained significant even after adjustment on the variables that differed between cases and controls (P < 0.1) at inclusion (high-density lipoprotein cholestrol, cardiovascular family history) and the cumulated and current exposure to the nucleosidic reverse transcriptase inhibitor, nonnucleosidic reverse transcriptase inhibitor, and protease inhibitor class.
Conclusion: Despite similar HIV and cardiovascular characteristics at baseline, cIMT decreased after 6 months of follow-up among the patients exposed to ATV/r, even after adjustment for the exposure to the three antiretroviral classes. Considering the shortcomings of this study, especially the absence of randomization and the heterogeneity of the control group, the benefit of ATV/r treatment in patients with high cardiovascular should be confirmed by randomized trials.
Introduction
With the advent of HAART regimens, it is crucial to consider their long-term benefits to risk ratios in HIV-infected patients. If the high cardiovascular risk among these patients is no longer discussed [1,2], the mechanism is controversial:
1. the consequence of an HIV-associated chronic inflammation [3Ğ7], of dysimmunity [8Ğ10], or of co-infection [11];
2. a high frequency of standard risk factors (sex, smoking, etc.) among HIV-infected population [8,12,13]. However, the Framingham risk score underestimates sub-clinical atherosclerosis observed in HIV-infected population [14];
3. HAART-induced metabolic abnormalities [7,15], especially with protease inhibitor [16], even a metabolic syndrome [17,18], lipodystrophy [19], endothelial dysfunction [20,21], or acquired thrombophilia [22], even though the impact of confounders is discussed [13,23,24]. In this case, apart from the usual healthy lifestyle recommendations [25] (giving up smoking, diet, physical exercise), the interventional strategy is not clear: switching from a HAART regimen generally active and well accepted to a new regimen [26], which induces less metabolic abnormalities but without assurance of a favorable clinical impact [27]; giving a specific treatment (statins [28], salicylic acid [22,27]), or only cardiovascular monitoring.
Since 2003, thanks to atazanavir (ATV), a new protease inhibitor, it is possible to assess the impact of a protease inhibitor-containing HAART regimen that does not provide major dyslipidemia [29] or insulin resistance [30] thanks to a validated surrogate investigator of atherosclerosis: the carotid intimaĞmedia thickness (cIMT).
Discussion
The cIMT, a well known and validated surrogate marker of atherosclerosis, decreased significantly among patients on the HAART regimen including ATV/r, whereas a slight increase is observed in the control group. This difference in the cIMT course appeared after 6 months of follow-up, when the cases had been ATV/r exposed since 289 days. The HIV and cardiovascular baseline characteristics were similar and these results did not change after the planned adjustments. The glycemia, the only parameter significantly different during the study (-0.44 mmol/l in case group), had no impact on the cIMT difference observed between cases and controls.
However, this study has some limitations. First, the study was not randomized. Thus, it is impossible to exclude the hypothesis that the ATV/r regimen may have been especially prescribed to patients with a high vascular risk, even though the cardiovascular baseline characteristics, including the Framingham score, the metabolic syndrome status, cardiovascular treatments, and cIMT were not significantly different. Incidentally, with larger samples, the baseline characteristics might have been different. Moreover, this modification in the HAART regimen could have been brought about by diet or physical activity recommendations, two relevant parameters, which were not collected in our study.
On the contrary, the heterogeneity of the control group should lead to sub-analyses according to the different HAART regimens. However, sufficient statistical power cannot be attained due to the small size of the samples. Moreover, the insufficient number of cases and controls, particularly at the end of the study, could explain why the mechanisms of the positive impact of ATV/r remain unclear. In fact, a major metabolic change is not obvious.
Thus, if an interventional correction of the cardiovascular risk is necessary, the results of the AVATAR study favor a switch to a low cardiovascular impact HAART regimen, like HAART including ATV/r, which amounts to the usual cardiovascular prevention. Of course, these data must be confirmed by further randomized studies.
Results
At least two cIMT measures were available for 97 controls and 32 cases. The cIMT at 18 months was available for only 12 cases and 52 controls. The main characteristics at inclusion and during the study are summarized in Table 1.
At the time of inclusion, among the cardiovascular risk factors, only the HDL-cholesterol level was significantly lower in the case group (1.23 vs. 1.62, P = 0.01). The percentage of the metabolic syndrome status (33.3 vs. 30.5) and the Framingham score (-3.32 vs. -3.36) were the same for cases and controls. No patient had previous or present CVD symptoms. As regards the current HAART regimen, two NRTI were associated with one protease inhibitor in 21 cases and in 25 controls (65.6 vs. 25.7%, P < 0.001), with one NNRTI in two cases and 42 controls (6.2 vs. 43.3, P < 0.001). Two cases (6.2%) and 15 (15.4%) controls (P = 0.008) received only NRTI, two controls and one case had another HAART regimen. Six (18.7%) cases and 13 (13.4%) controls were not treated (P = 0.3). The control group had a lower cumulated and current protease inhibitor exposure than the case group and an increased exposure to the NNRTI class. However the results were not statistically significant as regards the NRTI cumulated exposure (P = 0.35). The comparison between the 45 controls and the 16 cases, whose cIMT measure at 18 months was missing, gave the same results.
During the study, only the glucose level was significantly higher in the case group (5.95 vs. 5.51, P = 0.03). Of course, the protease inhibitor and NNRTI exposure were different between both groups, the cases being exclusively exposed to ATV/r, while the NRTI exposure was the same. Nevertheless, the use of a once-a-day drug (didanosine or tenofovir) was more frequent with ATV/r-based regimen than with zidovudine (Table 2, http://links.lww.com/QAD/A92). Ten controls (10.1%) did not receive HAART during the study; the three others were treated by boosted ATV/r and NRTI, 162, 186, and 504 days after inclusion, respectively. For two of them, two cIMT measures were missing.
Exposure to cardiovascular drugs was the same before and during the study (Table 2, http://links.lww.com/QAD/A92).
Although not different at inclusion (0.633 ± 0.05 in cases vs. 0.666 ± -0.09 in controls, P = 0.07), the cIMT significantly decreased at 12 (0.636 vs. 0.676; P = 0.05) and 18 months (0.611 vs. 0.675, P = 0.018), but not at 6 months (0.642 vs. 0.660) in cases when compared to controls (Fig. 1). After adjustment by the variables that differed between cases and controls (P < 0.1) at inclusion (HDL-c, cardiovascular family history) and the protease inhibitor, NRTI, and NNRTI cumulated and current exposures, the differences remain significant.
| |
| |
| |
|
|
|
