| |
Mild renal impairment is associated with calcified plaque parameters assessed by computed tomography angiography in people living with HIV
|
| |
| |
Download the PDF here
Our data highlight a robust relationship between subclinical renal impairment and overall coronary artery disease burden among PLHIV. This finding emphasizes the importance of identifying renal disease in the early stages and that potential intervention with medical therapy and referral to nephrologist may be needed in such patients. Further research is needed to understand the interplay between renal impairment and cardiovascular disease in HIV and clinical impact of this relationship in HIV......In studies performed in the general population, CKD is associated with higher rates of CVD events [6,10,11]. Moreover, the most common cause of mortality and morbidity in individuals with CKD is CVD rather than progression to ESRD [5]. Traditional risk factors such as hypertension, diabetes, and obesity play an intricate role in both CVD and CKD. In CKD, nontraditional risk factors such as albuminuria, disorders of mineral metabolism, oxidative stress, and inflammation have been shown to be associated with increased risk of CVD events [4,12,13]. Furthermore, carotid artery studies in the general population have shown that individuals with CKD have significantly increased calcification and alteration in plaque composition and vulnerability for plaque rupture [14,15]. In one study, carotid intima–media thickness was inversely correlated with eGFR [15]. Bundy et al. [4] showed that lower eGFR was significantly associated with coronary artery calcification in non-HIV participants in the Chronic Renal Insufficiency Cohort (CRIC) study.
'Multivariable regression modeling to evaluate the effect of estimated glomerular filtration rate on Agatston score and coronary plaque types..............In the model, TDF tended to be associated with total coronary segments with plaque, coronary segments with calcified plaque, and Agatston score. HIV viral load was also significantly and independently related to total coronary segments with plaque and Agatston score............'
Here for the first time, we show that mild subclinical renal impairment in PLHIV is independently correlated with coronary plaque assessed using CT angiography. Both creatinine and more strongly eGFR significantly correlated with critical indices of overall coronary plaque severity, including total severity of coronary plaque, total coronary segments with plaque, as well as calcium-specific indices including coronary segments with calcified plaque and Agatston score, but not with measures of noncalcified plaque, in PLHIV. The correlation between eGFR and key indices of calcified plaque, including the Agatston score, remained significant after multivariable regression analysis adjusting for total Framingham point score, BMI, CD4+ cell count, HIV viral load, duration of HIV diagnosis, and protease inhibitor, TDF, abacavir, and efavirenz use, suggesting a strong independent association, even among this group, chosen to exclude those with overt renal disease. These data provide further supporting evidence that mild subclinical renal disease, often common in HIV, may have important consequences for the development of coronary artery disease in this population.
In the current study, we relate eGFR (kidney function blood test) using the well accepted CKD-EPI formula, which accounts for race and sex, to measures of calcified and noncalcified coronary plaque in HIV-infected participants without a known history of CVD. Here, we demonstrate that mild renal impairment is significantly correlated with calcified plaque parameters, but not to noncalcified plaque parameters in PLHIV. This relationship to calcified but not to noncalcified plaque has not previously been compared in PLHIV. Calcified plaque is thought to relate to traditional CVD risk factors, whereas noncalcified plaque to other factors including immune activation in HIV[9,20,21]. Calcified plaque, more representative of traditional risk factors, is an important predictor of CVD events among non-HIV participants [22,23]. Such data is not yet available in PLHIV, but the ongoing REPRIEVE study will assess differential relationships of calcified and noncalcified plaque as well as eGFR to CVD events in HIV (NCT#02344290). Our finding suggests that early onset of subtle renal impairment can contribute to the development of calcified plaque in PLHIV. Importantly, we control for total Framingham point score, a global index of traditional CVD risk, and demonstrate persistent, independent associations of eGFR to coronary plaque and overall plaque burden, highlighting the independent effects of eGFR apart from other risk factors. Potential mechanisms by which subclinical renal disease and atherosclerosis are associated are likely multifactorial. Traditional risk factors such as age, sex, race, diabetes, hypertension, tobacco use and increased chronic inflammation from HIV alone can contribute to CKD and atherosclerotic disease [7,20,24–26]. Furthermore, HIV-related complications such as lipodystrophy and metabolic syndrome are also known to contribute to CVD [27–29]. Importantly, there may be a heightened interaction between mild renal impairment, immune and inflammatory factors, which drive increased coronary plaque in HIV. Indeed, in the current study, we show that HIV viral load is highly related to calcified plaque parameters consistent with prior data [18], suggesting that both traditional risk factors, including eGFR and nontraditional immune factors contribute to calcified plaque in this population.
HIV negative controls were included in the current analysis, and demonstrate generally similar, but less robust associations between coronary plaque and eGFR. It is unknown why subtle renal impairment relates more strongly to plaque in HIV patients. HIV and non-HIV groups may differ in the biological mechanisms of plaque formation. Alterations in kidney function resulting from virological factors may contribute to immune activation and other mechanisms of advanced plaque formation [38,39]. Future research is needed to further elaborate on this issue.
Relationship of estimated glomerular filtration rate to antiretroviral therapy use among HIV-infected participants
eGFR was lower among participants receiving protease inhibitors and TDF treatments. PLHIV on efavirenz had significantly higher eGFR (Supplemental Table 1, http://links.lww.com/QAD/B376). There was no significant correlation between eGFR and nucleoside reverse transcriptase inhibitors (NRTI), non-nucleoside reverse transcriptase inhibitors (NNRTI), abacavir, or raltegravir.
Correlation between creatinine and estimated glomerular filtration rate with coronary plaque type in people living with HIV and HIV negative controls
In PLHIV, eGFR was inversely and significantly related to total severity of coronary plaque score, total coronary segments with plaque, number of coronary segments with calcified plaque, and Agatston score (Table 3). In contrast, eGFR did not correlate with coronary segments with noncalcified plaque in PLHIV. In the HIV negative group, directionally similar but less robust relationships were seen relating eGFR to plaque parameters. The relationship was the strongest between eGFR and number of coronary segments with calcified plaque, which trended toward significance. Creatinine related significantly to multiple plaque parameters in PLHIV, but these relationships were less robust than with eGFR (Table 3). The same was seen among the non-HIV group.
We stratified eGFR to indicate normal and impaired renal function, defined by eGFR less than 90 ml/min per 1.73 m2. PLHIV with eGFR less than 90 ml/min per 1.73 m2, had significantly higher total severity of coronary plaque score compared with PLHIV with eGFR greater than or equal to 90 ml/min per 1.73 m2, 3.46 ± 4.98 and 1.73 ± 3.07, P = 0.02, respectively (Fig. 1). Among HIV-negative controls, the difference in total severity of plaque score was less comparing participants with eGFR less than 90 ml/min per 1.73 m2 and participants with eGFR greater or equal to 90 min per 1.73 m2, and did not reach statistical significance (Fig. 1).
|
|
| |
| |
|
|
|
