Antiretroviral Therapy: From Heart Risk Factor to Heart Protector?

Table of Contents



Multiple studies indicate that certain antiretrovirals raise chances of cardiovascular morbidity and mortality. At the same time, diverse research indicates that combination antiretroviral therapy lowers cardiovascular risk. Studies addressing this question have been pooled in several systematic reviews and meta-analyses. One meta-analysis of recent observational studies determined that recent abacavir or protease inhibitor use approximately doubled chances of myocardial infarction. Every additional year of lopinavir or indinavir therapy also independently raised MI risk. A second meta-analysis determined that HIV-positive antiretroviral-naive people had a 60% higher risk of cardiovascular disease than did HIV-negative people, while antiretroviral-treated people had a doubled risk compared with the HIV-negative group. Antiretroviral-treated people had about a 50% higher cardiovascular disease risk than did treatment-naive HIV-positive people. This analysis could not factor in other cardiovascular risks, such as smoking, which may be more prevalent in people with HIV and have nothing to do with cART. In this second meta-analysis, each year of treatment with protease inhibitors, nucleoside reverse transcriptase inhibitors, or nonnucleoside reverse transcriptase inhibitors added to cardiovascular disease risk. Studies are divided on whether a lower viral load or higher CD4 count cuts cardiovascular risk.

In 2010 EuroSIDA investigators found that a lower CD4 count inflated chances of every non-AIDS event analyzed, except one -- cardiovascular disease.1 Bolstering their finding with results of two other studies,2,3 the EuroSIDA team noted "there is, to date, no strong evidence linking cardiovascular disease with immunodeficiency." But even as the EuroSIDA team steered their paper into print, a small army of other researchers was amassing data pointing in the opposite direction.

If a low CD4 count tips the scales toward cardiovascular disease, one would expect combination antiretroviral therapy (cART) to ease cardiovascular risk by boosting CD4 tallies -- and maybe via other mechanisms. Yet studies from the United States,4 France,5 and the international Data Collection on Adverse Events of Anti-HIV Drugs (DAD) Study Group2 from the early 2000s all implicated cART -- and specifically protease inhibitor (PI)-based cART -- in surging cardiovascular rates seen in people with HIV. Other research tied certain nucleoside reverse transcriptase inhibitors (NRTIs) to heart disease. Ten years later, however, the French team found evidence that a viral load above 50 copies/mL hiked the risk of myocardial infarction 50%.6

A stockpot of other data simmered throughout these years, as researchers refined their multivariate recipes. Sometimes data implicating cART in heart matters bubbled to the top; sometimes cART emerged as an essential ingredient of a heart-healthy recipe. So where are we today, in 2013? Does antiretroviral therapy inflate chances of cardiovascular morbidity and mortality? Or does cART help HIV-positive people trim their coronary risk? The answers to those two questions would be yes and yes.

The Department of Health and Human Services (DHHS) Panel on Antiretroviral Guidelines for Adults and Adolescents believes stavudine, zidovudine, abacavir, efavirenz, and all ritonavir-boosted PIs can send lipids awry and so pose a cardiovascular threat to people taking those drugs (see Table 2 in the first review article in this issue).7 Franck Boccara, an HIV cardiology expert at Saint Antoine University Hospital in Paris, and colleagues suggest "the 2 most important and recent observational cohorts8,9 with a sufficient duration of exposure to PIs showed that the duration of exposure was associated with an increased risk for MI."10

Yet suspending cART in the SMART treatment interruption trial hoisted hazards of cardiovascular events more than 50% compared with taking steady cART.3 And when AIDS Clinical Trials Group (ACTG) investigators randomized antiretroviral-naive people to NRTIs plus efavirenz, NRTIs plus lopinavir/ritonavir, or efavirenz plus lopinavir/ritonavir, they found that all three regimens rapidly improved endothelial function (measured as brachial artery flow-mediated dilation), and that improvement persisted through 24 weeks of follow-up.11 Only one factor appraised predicted improved arterial function -- viral suppression. (For details of this study, see below under the subhead "Low viral load: low cardiovascular risk (usually).")

Cardio Meta-Analyses With (Slightly) Different Outcomes

When matters get this messy, who can resist a meta-analysis? But meta-analyses can get murky, too: Because they ask different questions, use different methods, and examine different studies, they can reach different conclusions. The earliest such effort -- already a decade old -- focused on 30 randomized double-blind trials involving the first four PIs: indinavir, ritonavir, saquinavir, and nelfinavir.12 Comparing trial participants who took a PI with those who took only NRTIs, these researchers found no higher MI risk with PIs (relative risk [RR] versus NRTIs 1.69, 95% confidence interval [CI] 0.54 to 7.48). The absolute difference in MI risk in PI takers was +0.77 per 1000 person-years, meaning an excess MI rate below 1 MI per 1000 people each year.

Three meta-analyses focused solely on abacavir,13-15 the NRTI famously yoked to higher MI risk in a DAD study.9,16 These three studies came from the FDA,13 the ACTG,14 and abacavir's maker, GlaxoSmith-Kline.15 None of them turned up any evidence that abacavir predisposes people to heart attacks.

A team from Stanford University offered the latest meta-analysis of cardiovascular risk with cART17 and compared their findings with those of the most comprehensive abacavir analysis13 and the early PI analysis.12 The Stanford group criticized both of these meta-analyses, noting they did not assess study quality or the likelihood of publication bias. These investigators winnowed a field of 1458 articles to 27 studies published through June 2011, only one of them a randomized controlled trial.

The Stanford researchers could combine data from only a handful of these studies for each of the risk profiles they explored.17 Two studies of cumulative exposure to NRTIs reached opposite conclusions on whether abacavir or didanosine magnifies MI risk, DAD saying those NRTIs did,9 the French national team saying they did not.8 Pooled analysis of two studies16,18 determined that abacavir use within the last 6 months almost doubled MI risk (RR 1.91, 95% CI 1.50 to 2.42) (Figure 1). Three studies of recent didanosine use8,16,19 could not be combined by meta-analysis, but together they indicated a "harmful association" between didanosine and MI risk. No studies yielded evidence that other NRTIs imperil heart health.

Meta-Analysis of PI and Abacavir Impact on MI Risk
Figure 1. Meta-Analysis of PI and Abacavir Impact on MI Risk Meta-analysis of recent observational studies determined that recent abacavir or protease inhibitor (PI) use approximately doubled chances of myocardial infarction (MI).17 Every additional year of lopinavir or indinavir therapy also independently raised MI risk. (Risks for abacavir, lopinavir, and indinavir calculated as relative risk (RR); risk for recent PI use calculated as odds ratio (OR). See text for 95% confidence intervals.)

The Stanford team melded data from a DAD study9 and a French study8 to determine that every additional year of lopinavir use boosted MI chances more than 20% (RR 1.22, 95% CI 1.01 to 1.47) (Figure 1). Every additional year of indinavir use inflated MI chances a little more than 10% (RR 1.11, 95% 1.05 to 1.17). Another DAD analysis figured that every additional year of exposure to PIs as a class significantly raised MI risk.20 Combining three studies21-23 that calculated odds ratios for recent PI use, the Stanford statisticians reckoned a doubled MI risk with recent PI use (OR 2.13, 95% CI 1.06 to 4.28). Combining 6 studies21-26 by a different method, they confirmed a significantly higher MI risk with recent PI use (P = 0.003).

The Stanford group does a good job not only sifting through these hazards, odds, and oddities, but also explaining what they mean:17

  • Evidence from observational studies implicated both PIs and abacavir in myocardial infarction risk.
  • Evidence from randomized trials did not.
  • Randomized trials offer the least biased approach to reckoning cardiovascular risk.
  • But none of the clinical trials analyzed was designed for that purpose, and none lasted very long.
  • Observational studies include a much larger and more representative patient sample than clinical trials.
  • But observational studies are fraught with confounders that cannot be adjusted away by savvy statisticians.
  • Also, combining evidence from several studies is hard because the studies differ in design and analytical plan.

Keeping all those caveats in mind, the Stanford investigators "believe there is still uncertainty whether ART leads to increased cardiovascular risk, and if so, the magnitude of that risk."17 But the observational studies analyzed yield enough good data "to warrant further study in prospective studies designed to assess cardiovascular risk from ART."

People dissatisfied with "uncertainty" after all these numbers get crunched down to bite-sized portions can consult yet another meta-analysis of studies weighing cardiovascular risk with cART.27 Of course this second meta-analysis, by researchers at the University of New South Wales, catechizes mostly the same studies as the first meta-analysis,17 so the findings are largely concordant. But the Australian team asked some different questions and used somewhat different methods, so their findings do not perfectly mirror those of the Stanford team.

The Australian meta-analysis focused on 23 studies, including 2 randomized trials, published before August 2010.27 Unlike the Stanford group,17 the Australian investigators combined studies with different risk metrics (odds ratios, relative risks, or hazard ratios). While myocardial infarction was the outcome in the Stanford study, the main Australian outcome was "cardiovascular disease," meaning coronary artery atherosclerosis. Whereas both groups assessed the impact of individual antiretrovirals and antiretroviral classes, the Australians also compared outcomes in antiretroviral-naive people, antiretroviral-treated people, and HIV-negative people. In that analysis, the antiretroviral-naive group had about a 60% higher cardiovascular disease risk than people without HIV (RR 1.61, 95% CI 1.42 to 1.83) and antiretroviral-treated people had a doubled risk (RR 2.00, 95% CI 1.70 to 2.37). cART-treated people had about a 50% higher cardiovascular disease risk than treatment-naive HIV-positive people (RR 1.52, 95% CI 1.35 to 1.70). Notably, though, this analysis could not factor in other cardiovascular risks, such as smoking, which may be more prevalent in people with HIV and have nothing to do with cART.

In the Australian meta-analysis, every added year of PI therapy upped the cardiovascular disease risk 11% (RR 1.11, 95% CI 1.05 to 1.17)27 (Figure 2), about the same as each year of lopinavir therapy jacked MI risk in the Stanford inquest17 (Figure 1). Each year of NRTI therapy boosted cardiovascular disease risk 4%, a relative risk just beyond the confines of statistical significance (RR 1.04, 95% CI 0.99 to 1.09). Each year of nonnucleoside therapy budged cardiovascular risk 5%, a statistically significant impact (RR 1.05, 95% CI 1.01 to 1.10). DAD saw the same per-year MI risk with NNRTIs, but the association stopped short of statistical significance: relative rate 1.05, 95% CI 0.98 to 1.33.)20

Meta-Analysis of CVD Risk With Each Year of Treatment
Figure 2. Meta-Analysis of CVD Risk With Each Year of Treatment Meta-analysis of studies assessing cardiovascular disease (CVD) risk in antiretroviral-treated people found that each year of treatment with a protease inhibitor (PI), lopinavir (LPV), a nucleoside reverse transcriptase inhibitor (NRTI), abacavir (ABC), or a nonnucleoside reverse transcriptase inhibitor (NNRTI) raised that risk.27 The association with NRTIs fell just short of statistical significance. See text for 95% confidence intervals.

Every year taking lopinavir inflated chances of cardiovascular disease 19% (RR 1.19, 95% CI 1.03 to 1.39), and each year of abacavir boosted chances 5% (RR 1.05, 95% CI 1.02 to 1.16).27 The lopinavir and abacavir risks of cardiovascular disease diverge from the lopinavir- and abacavir-linked MI risks in the Stanford study17 (Figure 1). In the Australian analysis, people who took a PI regimen ran about a 40% higher risk of cardiovascular disease than people taking a non-PI combo (RR 1.41, 95% CI 1.21 to 1.65).

The Australian team stresses that cART has improved the "quality and length of life" in people with HIV.27 And "it is possible," they caution, "that the use of ART increases life expectancy and hence increases the average age of those taking ART in comparison to the reference group, which may lead to confounding of results."

cART and Cardiology: A Balancing Act

So what's the bottom line? No one suggests stopping or delaying or interrupting cART to trim the risk of heart disease, even in people with an otherwise foreboding risk profile. When US antiretroviral guideline writers began recommending cART for everyone with HIV, regardless of CD4 count, the first reason they listed is the higher risk of AIDS, cardiovascular disease, and other non-AIDS diseases with untreated infection or uncontrolled viremia.7

But what about avoiding or switching from PIs or abacavir in people with an otherwise high risk? Certainly one would not want to avoid all PIs when considering cardiovascular risk. A big DAD analysis sniffed out not a whiff of evidence that cumulative atazanavir treatment boosts MI or stroke risk.28

A comparable analysis has not weighed the impact of darunavir/ritonavir on cardiovascular risk. But a review of trials involving two NRTIs plus a ritonavir/boosted PI as first-line therapy found darunavir/ritonavir comparable to atazanavir/ritonavir in 48-week lipid readings, and superior to lopinavir/ritonavir or amprenavir/ritonavir in triglyceride or total cholesterol results.29 A 13-person study tracked lipids and cystatin C in people who switched from lopinavir/ritonavir or amprenavir/ritonavir to darunavir/ritonavir.30 Total cholesterol, low-density lipoprotein (LDL) cholesterol, and triglycerides all improved through 12 months of follow-up, as did cystatin C. (High levels of cystatin C have been linked to cardiovascular disease, kidney disease, and death.)

What about abacavir in people with a high background cardiovascular risk? US antiretroviral advisors demoted abacavir/lamivudine from a "preferred" to an "alternative" NRTI backbone7 because of some evidence indicating worse virologic outcomes with abacavir/lamivudine than with tenofovir/emtricitabine.31,32 But after reviewing studies of cardiovascular risk with abacavir, these experts concluded that "to date [in February 2013], no consensus on the association between abacavir use and MI risk or the mechanism for such an association has been reached."7

The DAD Study Group, whose two big analyses first turned the spotlight on abacavir as a possible MI risk factor,9,16 stressed in their later report that the overall MI rate in this population was low -- 3.2 events per 1000 person-years.9 In other words, 3 of 1000 people in the DAD cohort (0.3%) died of an MI every year. And "any toxicities of antiretroviral drugs must always be interpreted in the context of the benefits that these drugs provide," the DADmasters added.9

A team of top-drawer HIV researchers distilled HIV-related cardiovascular risk variables into a list of seven -- three that raise risk and four that lower risk (Figure 3).33 Antiretroviral therapy figures in most of these risk factors in one way or another. On the increased-risk side of the equation, cART can contribute to dyslipidemia, insulin resistance, and body shape changes. And because cART prolongs survival with HIV, it paradoxically favors the higher risk of cardiovascular death that comes with older age. On the decreased-risk side of the equation, all four factors involve cART.

Positive and Negative CVD Risk Factors Negative CVD Risk Factors in People Taking cART
Figure 3. Positive and Negative CVD Risk Factors Negative CVD Risk Factors in People Taking cART Top HIV clinicians, cardiologists, and epidemiologists proposed this scheme summarizing the balance between decreased and increased cardiovascular risk in people with HIV.33 Most factors on both sides of the equation involve cART.

In an interview in this issue, the University of Wisconsin's James Stein stresses that viral suppression remains the overriding principle of antiretroviral. "As a cardiologist," he says, "I would never tell an HIV treater or a patient with HIV infection that they can't start an antiretroviral because it raises their heart disease risk so much that it will overshadow the risk of uncontrolled HIV infection." But if two drugs are good candidates for viral control and one carries some cardiovascular disease risk, he would opt for the antiretroviral with a cleaner cardiovascular risk profile in someone with an already high risk of cardiovascular disease.

Current US antiretroviral guidelines lean toward favoring cART-induced viral control as one way to cut cardiovascular risk, citing multiple lines of evidence suggesting "that early control of HIV replication with ART can be used as a strategy to reduce risk of CVD, particularly if drugs with potential cardiovascular toxicity are avoided."7 But no study demonstrates that cART prevents heart disease, these experts caution. And "for HIV-infected individuals with a significant risk of CVD, as assessed by medical history and established estimated risk calculations, risk of CVD should be taken into consideration when selecting a specific ART regimen."7

Low Viral Load: Low Cardiovascular Risk (Usually)

A well-planned cART regimen usually boosts CD4 tallies and curbs HIV replication. Abundant research addresses whether those responses directly affect cardiovascular risk. By and large the answer seems to be yes, though several key studies say no. Sorting out the reasons for these divergent results is tough, but close analysis offers some clues.

Recent French and US studies appraised the impact of viral replication on a clinical endpoint -- myocardial infarction,6,34 and a European-Canadian-Australian cohort study gauged the impact of viral load on cardiovascular death.35 A case-control study within the French Hospital Database on HIV determined that a viral load above 50 copies/mL upped chances of incident myocardial infarction 50%.6 This study involved 289 people with a new MI between January 2000 and December 2006, matched by age, sex, and clinical center to 884 HIV-positive people with no MI history.

Median age was 47 in people with an MI and 46 in those without an MI. A higher proportion of control patients had a body mass index in the overweight range, but otherwise classic cardiovascular risk factors were more prevalent in the case group, including current smoking (64% versus 40%, P = 0.028), family history of premature coronary artery disease (18% versus 7%, P < 0.001), hypertension (20% versus 12%, P = 0.001), current cocaine or injection drug use (13% versus 9%, P = 0.041), and diabetes (16% versus 10%, P = 0.036). Fasting glucose and lipid measurements were significantly worse in people who had an MI.

Statistical analysis that considered antiretroviral exposure, CD4 and CD8 counts, and a mélange of classic risk factors determined that a current viral load above 50 copies/mL (versus below) independently raised the odds of a new MI 51% (adjusted odds ratio 1.51, 95% CI 1.09 to 2.10). Cumulative exposure to PIs more than doubled MI odds in this analysis (adjusted odds ratio 2.23 per 10 years, 95% CI 1.17 to 4.24), but abacavir exposure did not.

A study of 6517 HIV-positive people in two tertiary-care Boston hospitals found links between higher viral load and incident MI in statistical models that did not include CD4 count, but not in the models that did factor in CD4 tallies.34 The Boston team checked records of HIV-positive people in care sometime between December 1998 and February 2008 to see how many suffered an acute MI. Age averaged 53.7 years in the 273 people who had an MI and 45.7 in the 6244 who did not. Women made up almost one third of the study group; 55% were white, 24% black, and 18% Hispanic. Half of these people smoked (55% with an MI and 50% without an MI). Classic cardiovascular risk factors were consistently more prevalent in the MI group.

Statistical analysis accounting for classic risk factors, antiretrovirals, CD4 count, and viral load determined that a load above 100,000 copies/mL predicted acute MI, but not significantly (adjusted odds ratio [aOR] 1.63, 95% CI 0.91 to 2.93, P = 0.10). In statistical models not including CD4 count, a higher viral load invariably altered odds of acute MI:

  • Above 100,000 copies: aOR 2.16, 95% CI 1.26 to 3.69, P = 0.01
  • Every 10-fold higher viral load: aOR 1.23, 95% CI 1.04 to 1.44, P = 0.01
  • Every 10-fold higher peak viral load: aOR 1.23, 95% CI 1.04 to 1.44, P = 0.02
  • Less than 400 copies: aOR 0.60, 95% CI 0.38 to 0.93, P = 0.02

More HIV RNA may also inflate prospects of cardiovascular death, according to a 23-cohort European-Canadian-Australian CASCADE collaboration analysis involving 9858 people with an estimated date of HIV seroconversion.35 Compared with people whose viral load lay below 100,000 copies/mL while not on cART, those with a load above that level while not on cART had almost a 6 times higher risk of death from cardiovascular disease (adjusted hazard ratio 5.81, 95% CI 1.59 to 21.24) and those with a viral load above 100,000 copies/mL while on cART had almost a 5 times higher risk (adjusted hazard ratio 4.70, 95% CI 1.25 to 17.73). However, the study group included only 36 people who died from heart disease. And this analysis uncovered no link between CD4 count and death from cardiovascular disease.

Not all big studies tie viral load to cardiovascular disease or death. Two notable exceptions are the SMART study3 and a big DAD analysis.20 SMART randomized 5472 adults with a CD4 count above 350 cells/mm3 to continuous cART or to CD4 count-guided interruptions. During follow-up a major cardiovascular condition developed in 79 people, with more than a 50% higher risk in the interruption arm (hazard ratio 1.57, 95% CI 1.00 to 2.46, P = 0.05). With an expanded definition of major cardiovascular events, that risk reached statistical significance (hazard ratio 1.58, 95% CI 1.12 to 2.22, P = 0.009). But statistical analysis considering age, gender, use of anti-hypertensives, smoking, and total cholesterol found no evidence that being off cART at the time of the cardiovascular event or in the past 6 months made the event more likely. Every 10-fold higher most recent viral load also failed to implicate viral replication in incident cardiovascular events in several analyses.

People who interrupted cART during SMART had a worse total-to-high-density lipoprotein (HDL) cholesterol ratio than steadily treated people.3 That difference, the SMART team suggested, "could offer a partial explanation" for the higher cardiovascular event rate in the interruption arm. The researchers also acknowledged the hypothesis that inflammation kicked off by a "sudden burst of high level HIV replication" when a drug break began could trigger an MI or a stroke. Perhaps, the authors surmised, the trial did not measure viral load often enough to capture these viremic flares. Or maybe viral load simply is not a good way to measure inflammation and immune activation. The SMART team also suggested their analysis may have suffered from low statistical power because of the scant cardiovascular events recorded (79).

DAD investigators focused only on MIs,20 as the French6 and Boston34 studies did. The DAD team counted 345 MIs in 23,437 HIV-positive people monitored through February 2005 for an incidence of 3.65 per 1000 person-years. Median age at last follow-up was 43 years overall and 49 in people who had an MI. Three quarters of cohort members (78%) were white, and almost everyone (94%) had taken antiretrovirals. The DAD team found no link between peak viral load and MI risk (relative rate for each 10-fold higher peak viral load 1.06, 95% CI 0.95 to 1.18) or between CD4 count and MI risk (relative rate for each 50-cell higher count 0.98, 95% CI 0.95 to 1.01). But the investigators did not reckon the impact of other viral load measures on MI risk -- such as latest viral load or viral load at MI.

Why would a low viral load cut chances of cardiovascular disease or death? Uncontrolled viral replication means prodigious inflammation and immune activation, both of which threaten the heart and its vascular tributaries. But a subgroup analysis of an ACTG antiretroviral trial suggests another reason: endothelial function improves when cART curbs viral replication.11 ACTG investigators found that starting either a standard PI or NNRTI regimen -- or lopinavir plus efavirenz without NRTIs -- swiftly improved endothelial function reckoned as brachial artery flow-mediated dilation. This 82-person analysis charted significant improvements in flow-mediated dilation with all three regimens (Figure 4).

cART Quickly Improved Endothelial Function in a 3-Arm ACTG Trial
Figure 4. cART Quickly Improved Endothelial Function in a 3-Arm ACTG Trial Combination antiretroviral therapy (cART) may lower the risk of cardiovascular disease by improving arterial endothelial function, measured as flow-mediated dilation, according to results of an ACTG trial substudy.11 Arteries must be able to dilate readily to generate an anticoagulant surface.36 (Illustration from Servier Medical Art.)

Study participants were young (median 35 years, interquartile range 30 to 40), and 91% were men.11 As in many HIV populations, a high percentage, 44%, smoked. Median body mass index just crossed the overweight threshold (25.1 kg/m2), but group blood pressure was good (119/74 mm Hg). Except for low HDL cholesterol, other lipid, glucose, and insulin values were within the normal range. The group had a median pre-cART CD4 count of 245 cells/mm3 and a median viral load around 66,000 copies/mL.

After 24 weeks of treatment, 67% of study participants had a viral load below 50 copies/mL and another 18% had between 50 and 100 copies/mL. After only 4 weeks of treatment, flow-mediated dilation rose (improved) 0.74% overall (IQR (-0.62% to +2.74%, P = 0.003), with no difference between study arms.

After 24 weeks of treatment flow-mediated dilation continued to improve in every study arm, increasing by 1.48% overall (IQR -0.20% to +4.30%, P < 0.001). Change in flow-mediated dilation from baseline to week 24 correlated with only two factors -- adiponectin and change in viral load. (Adiponectin is an adipocyte-specific protein that may play a role in insulin resistance and atherosclerosis.) The correlation with viral load was inverse (-0.30, P = 0.017), meaning the bigger the drop in viral load, the greater the improvement in endothelial function. The ACTG team suggests their findings support the hypothesis that controlling HIV replication improves endothelial function, but they note that follow-up was short and the study group young.

Disentwining Diverse CD4 Impacts on Cardiovascular Risk

Two DAD Study analyses,2,20 a EuroSIDA review,1 the SMART cardiovascular endpoint dissection,3 and a combined analysis of the ESPRIT and SILCAAT interleukin 2 (IL-2) trials37 discerned no link between CD4 count (measured various ways) and risk of cardiovascular disease (also measured various ways) (Table 1). But since the last of those reports in 2010,1,37 five other cohort studies,6,34,38-40 including a new DAD analysis,40 did yoke CD4 count to cardiovascular endpoints (Table 2). Two other studies tied lower CD4 counts to subclinical signals of arterial disease,41,42 and one linked lower CD4 nadir to sustained hypertension.43

Table 1. Studies Finding No Association Between CD4 Measures and Cardiovascular Endpoints
Study (Years)n (% Male)Age, CD4 CountStudy GroupCD4 MeasureCVD EndpointMain Results
EuroSIDA1 (2001-2009)12,844 (73.2%)BL 39 y

Nadir 178, BL 403
Prospective cohort from Europe, Israel, ArgentinaDoubling of current CD4 countMI, stroke, CABG, coronary angioplasty, carotid endarterectomy (n = 384)Every doubling of current CD4 count predicted lower risk of 4 non-AIDS illnesses, but not CVD
DAD2 (1999-2002)23,468 (75.9%)BL 39 y

Nadir 226, BL 418
Prospective cohort from Europe, United States, AustraliaPer 50-cell higher nadir CD4 count and BL CD4 count (at enrollment)MI (n = 126)Every 50-cell higher nadir CD4 count or BL CD4 count had no impact on MI risk
DAD20 (1999-2005)23,437 (75.9%)BL 39 y, last FU 43 y

BL 420, last FU 461
Prospective cohort from Europe, United States, AustraliaPer 50-cell higher nadir CD4 countMI (n = 345)Every 50-cell higher nadir CD4 count had no impact on MI risk
SMART3 (2002-2006)5742 (73%)BL 44 y

BL 267
People randomized to continuous or interrupted cARTPer 100-cell higher current CD4 countClinical or silent MI, nonfatal stroke, CAD requiring surgery or invasive procedure (n = 79)Every 100-cell higher current CD4 count had marginal impact in interruption arm (P = 0.08) and no impact in combined trial arms
ESPRIT and SILCAAT37 (NR)3012 (82.3%)BL 41 y

Nadir 167, BL 400
People randomized to standard cART (no IL-2) in two IL-2 trialsPer doubling of latest CD4 countMI, stroke, CAD requiring procedure, other fatal heart/vascular events, sudden death (n = 95)Every doubling of latest CD4 count had no impact on rate of fatal or nonfatal CVD events

BL, baseline; CABG, coronary artery bypass graft; CAD, coronary artery disease; CVD, cardiovascular disease; FU, follow-up; IL-2, interleukin 2; MI, myocardial infarction; NR, not reported.

Table 2. Studies Finding an Association Between CD4 Measures and Cardiovascular Endpoints
Study (Years)n (% Male)Age, CD4 CountStudy GroupCD4 MeasureCVD EndpointMain Results
FHDH6 (2000-2006)289 cases, 884 controls (89%)BL 47 y

Nadir 135, BL 427
Case-control study of HIV+ with first MI or no MIDoubling of nadir CD4, highest tertile CD8MI (n = 289)Every doubling of CD4 nadir cut MI risk 10%; highest (vs lowest) CD8 tertile raised risk 48%
Boston34 (1998-2008)6517 (69.4%)46 y

26% <200
HIV+ in two large Boston hospitalsCurrent CD4 count <200, every 50-cell higher CD4 countMI (n = 273)Current CD4 count <200 raised MI odds 74%; every 50-cell higher current CD4 lowered MI odds 7%
HOPS38 (2002-2009)2005 (76%)42 y

Nadir 197, BL 395
Prospective cohort from 10 US centersBL CD4 count <350 vs >500, every 100-cell lower BL CD4 countMI, stroke, CAD, angina, PAD (n = 148)BL CD4 count <350 raised CVD risk 58%; every 100-cell lower BL CD4 count raised risk 8%
ATHENA39 (2000-2009)3068 (83.4%)41 y

Nadir 170, BL 360
Prospective cohort on cART in NetherlandsCD4 count <200, 200-350, 351-500, >500 2 years after starting cARTtMI, CABG, coronary stenting and/or angioplasty, cerebrovascular attack (n = 57)CD4 count of 200-350 (vs <200) 2 years after starting cART cut CVD risk 66%
DAD40 (1999-2008)33,308 (74.1%)BL 39 y

BL 408
Prospective cohort from Europe, United States, AustraliaEvery 50-cell higher current CD4 countCardiovascular death (n = 289)Every 50-cell higher current CD4 count cut CVD death risk 3%

BL, baseline; CABG, coronary artery bypass graft; CAD, coronary artery disease; CVD, cardiovascular disease; FHDH, French Hospital Database on HIV; HOPS, HIV Outpatient Study; MI, myocardial infarction; PAD, peripheral arterial disease.

In a EuroSIDA analysis of 12,844 HIV-positive people (Table 1), every doubling of current CD4 count independently predicted a lower incidence of AIDS, all non-AIDS events combined, non-AIDS malignancies, end-stage renal disease, pancreatitis, and liver-related events, but not cardiovascular events (incidence rate ratio 0.98, 95% CI 0.85 to 1.12, P = 0.78).1 The EuroSIDA team proposed that cardiovascular disease risk depends less on CD4 status than on lipid changes, lifestyle, and inflammation.

Two DAD analyses (Table 1) determined that every 50-cell higher nadir CD4 count2,20 or every 50-cell higher baseline CD4 count2 had no impact on MI risk in prospective follow-up of more than 23,000 people. But the DAD team acknowledged "the possibility that other unmeasured immunologic effects may exert an influence on the development of cardiovascular disease."20 For example, time-updated CD4 count, CD4 count at MI diagnosis, or CD4/CD8 ratio could have an impact in this population. In an updated DAD analysis involving 33,308 people (Table 2), higher latest CD4 count did predict a slightly lower risk of cardiovascular death (see below).40

In the SMART analysis of 79 cardiovascular events in that trial, every 100-cell higher current CD4 count marginally boosted chances of cardiovascular disease (adjusted hazard ratio 1.11 per 100 cells, 95% CI 0.99 to 1.25, P = 0.08) in the cART interruption arm.3 But that CD4 yardstick had no impact on cardiovascular risk in the combined study arms (adjusted hazard ratio 0.99, 95% CI 0.90 to 1.07, P = 0.74).

ESPRIT and SILCAAT randomized antiretroviral-naive adults to standard cART or to standard cART plus IL-2 (Table 1).37 To analyze the impact of various CD4 metrics on AIDS and non-AIDS endpoints, the investigators focused on 3012 people randomized to the standard-cART control arms. Every doubling of the latest CD4 count had no impact on the rate of fatal or nonfatal cardiovascular events (adjusted hazard ratio 1.05, 95% CI 0.77 to 1.43). Among all these analyses (Tables 1 and 2), the ESPRIT/SILCAAT study involved the smallest number of HIV-positive people and only 95 cardiovascular endpoints. The investigators surmised higher rates of cardiovascular and other non-AIDS diseases in people with HIV could reflect a "subtle ongoing inflammatory process stimulated by residual viral replication or the treatment" and "subclinical inflammation may not be best reflected by latest CD4+ count."37

French Hospital Database on HIV investigators planned a case-control study specifically to scrutinize the impact of viral load and CD4 count on risk of first myocardial infarction (Table 2).6 The French team matched 289 HIV-positive people who had a first MI in 2000-2006 to 3 controls of the same age, sex, and clinical center who had not had an MI. A current viral load above 50 copies/mL (versus below) independently raised chances of a new MI 51% (adjusted odds ratio 1.51, 95% CI 1.09 to 2.10).

In the same analysis, which adjusted for antiretroviral exposure and classic risk factors, two T-cell variables swayed MI risk: every doubling of CD4-cell nadir trimmed MI risk 10% (adjusted odds ratio 0.90, 95% CI 0.83 to 0.97), and being in the highest current CD8 count tertile (above 1150 cells/mm3) versus the lowest tertile (at or below 760 cells/mm3) hoisted MI odds almost 50% (adjusted odds ratio 1.48, 95% CI 1.01 to 2.18). Current CD4 count did not predict MI, but CD4 nadir/CD8 ratio did. The CD4/CD8 ratio was significantly lower (worse) in cases than controls (0.42 vs 0.50, P < 0.001). A higher CD8 count reflects ongoing immune activation to control HIV, indicated in this study by the parallel link between a detectable viral load and heightened odds of myocardial infarction. Previous studies have also linked MI or cardiovascular disease markers to CD4/CD8 ratio44 or other immune activation markers.44-46

The Boston study described in the preceding section found ties between several viral load measures and acute MI when the analysis excluded CD4 count.34 A multivariate regression model adjusted for viral load, age, gender, race, hypertension, diabetes, dyslipidemia, chronic kidney disease, smoking, years since first cART use, and antiretroviral medications individually associated with MI determined that a current CD4 count under 200 cells/mm3 boosted MI odds almost 75% (adjusted odds ratio 1.74, 95% CI 1.07 to 2.81, P = 0.02). Further analysis determined that every 50-cell higher current CD4 count cut MI risk 7% (adjusted odds ratio 0.93, 95% CI 0.89 to 0.97, P = 0.002). Every 50-cell higher CD4 nadir pared MI risk 5%, but that association did not reach statistical significance (adjusted odds ratio 0.95, 95% CI 0.89 to 1.01, P = 0.09). Thus in this analysis, the impact of CD4 measures on MI appeared to outweigh the impact of viral control, because viral load associations proved significant only when statistical models did not include CD4 count.

An HIV Outpatient Study (HOPS) analysis of 2005 HIV-positive people counted 148 new cardiovascular diagnoses (defined in Table 2) from 2002 through 2009.38 A multivariate model accounting for traditional risk factors determined that a baseline CD4 count below 350 cells/mm3 (versus at or above 500) boosted cardiovascular disease risk 58% (adjusted hazard ratio 1.58, 95% CI 1.09 to 2.31, P = 0.017). Every 100-cell lower baseline CD4 count upped the cardiovascular event risk 8% (adjusted hazard ratio 1.08, 95% CI 1.01 to 1.14). Additional adjustment for baseline injection drug use, frequency of alcohol use, and baseline viral load did not change these results. The HOPS team also calculated that about 20% of cardiovascular disease risk could be attributed to a sub-350 baseline CD4 count (versus above 499), an attributable risk similar to those seen with several classic risk factors (Figure 5).

CVD Risk Attributable to CD4 Count and Other Factors
Figure 5. CVD Risk Attributable to CD4 Count and Other Factors A 2005-person HIV Outpatient Study (HOPS) analysis figured that cardiovascular disease risk attributable to a baseline CD4 count below 350 cells/mm3 (versus above 499) was similar to risk attributable to classic risk factors like current or former smoking, high LDL or non-HDL cholesterol, low HDL cholesterol, and male gender.38 The HOPS team cautioned that this analysis may not apply to HIV populations with cardiovascular risk factor rates unlike those in these HOPS cohort members.

In the Netherlands ATHENA cohort investigators focused on 3068 people who had taken cART for at least 2 years and reached a viral load below 500 copies/mL, dividing them into 2-year CD4 brackets of below 200, 200 to 350, 351 to 500, and over 500.39 A multivariable model to pinpoint predictors of a new cardiovascular diagnosis (see Table 2) adjusted for age, gender, family history of heart disease, cardiovascular event before baseline, smoking, and alcohol abuse. Compared with a CD4 count below 200 cells/mm3 after 2 years of cART, a count of 200 to 350 cut the cardiovascular event risk by two thirds (adjusted hazard ratio 0.34, 95% CI 0.14 to 0.86, P= 0.02). People who reached a CD4 count above 500 had almost a 50% lower risk of reaching a composite endpoint including death, AIDS, malignancies, liver cirrhosis, and cardiovascular events (adjusted hazard ratio 0.54, 95% CI 0.33 to 0.87, P = 0.01). Because older age, a lower nadir CD4 count, and a higher precART viral load independently predicted poor CD4 recovery, the ATHENA team suggested that "starting HAART at higher CD4 cell counts, especially in older aged patients, may be beneficial."39 DAD investigators offered the biggest study to address CD4 impact on the ultimate cardiovascular endpoint -- death.40 They considered latest CD4 count in six brackets, under 50, 50 to 59, 100 to 199, 200 to 349, 350 to 499, and 500 or higher. Cardiovascular death rates were 3.11 per 1000 person-years for people in the lowest CD4 bracket and 1.16 per 1000 for those in the highest bracket. Every 50-cell higher current CD4 count trimmed the risk of cardiovascular death 3% (adjusted relative rate 0.97, 95% CI 0.95 to 0.99).

What CD4 and Viral Load Measure -- and What They Don't

Why do some cohort studies find no link between CD4 measures and cardiovascular disease (Table 1) while others do (Table 2)? Comparing features of the two groups of studies in these tables yields no easy answer, but perhaps some hints. The no-association studies were generally bigger, but the biggest analysis, the 1999-2008 DAD study,40 did find an link between latest CD4 count and cardiovascular disease, and it had the sternest endpoint -- death. The only case-control study, from the French Hospital Database on HIV, identified ties between both CD4 and CD8 counts and MI risk.6 Median or mean age tended to be older in the positive-association studies (Table 2). Older age would yield more cardiovascular endpoints and so may beef up the statistical power needed to show an association. But two no-association studies, EuroSIDA1 and DAD,20 had the highest number of endpoints.

DAD investigators who worked on the no-association 1999-2005 analysis20 noted that a CD4 metric other than the one they used could have found a link between CD4s and cardiovascular events. Both of the DAD studies that found no link between CD4 count and cardiovascular trouble used a 50-cell higher CD4-nadir measure,2,20 and one of those studies used cohort baseline CD4 count.2 The French study that found a CD4-heart link was the only other analysis to use nadir CD4 count, and they used a doubling of CD4 nadir.6 The 1999-2008 DAD analysis that found a CD4 association with cardiovascular death used latest CD4 count as the yardstick. Perhaps every 50-cell higher CD4 nadir is too fine a gauge to identify an association.

The French team suggested another reason why results of these 10 big studies differ. Their case-control probe showed that a viral load above 50 copies/mL upped the MI risk by half.6 Two other studies reviewed above tied a lower viral load to a lower heart disease or death risk.34,35 In the studies evaluating CD4 impact on cardiovascular disease, the French investigators proposed, "the differences between the studies could be explained by differences in the proportion of patients with controlled viral load."6

Why would lower or higher CD4 count affect risk of cardiovascular disease? On an elementary level, a higher CD4 count indicates better overall health, and healthier people are less likely to get diagnosed with any number of non-AIDS diseases, including heart disease. But there are probably more precise mechanisms. A climbing CD4 count typically mirrors falling numbers of CD8s -- the T cells recruited to kill infected cells and tumor cells. Fewer CD8s in circulation mean less HIV in circulation, in other words, less inflammation and immune activation.

This is not just airy hypothesis. A handful of studies address this issue in one way or another. During untreated HIV infection, CD4s wane and CD8s surge, an immunologic seesaw that sends a normal CD4/CD8 ratio (about 2) into the abnormal range (under 1). A study of 78 HIV-positive men with an average age of 46.5 and no history of coronary artery disease used computed tomography coronary angiography to assess indicators of atherosclerosis.44 Lower (worse) CD4/CD8 ratio was significantly associated with both number of plaque-bearing coronary artery segments and plaque volume. Notably, the relationship between CD4/CD8 ratio and plaque volume proved stronger than the association seen with CD4 count or viral load and plaque volume -- a result suggesting CD4 count and viral load may be relatively blunt instruments for assessing cardiovascular risk.

Two studies in San Francisco found higher levels of CD4 and CD8 activation with lower CD4 counts.47,48 The first study correlated CD8-cell activation with poor CD4 gains despite good virologic control with antiretroviral therapy.47 The 99 adults evaluated kept their viral load at or below 1000 copies/mL for a median of 21 months on cART. Although they had lower levels of CD8-cell activation than untreated HIV-positive people, they had higher levels than HIV-negative controls. Every 5% higher proportion of activated (CD38+/HLA-DR+) CD8 cells meant a 35-cell lower CD4 gain during therapy. The same researchers analyzed CD4- and CD8-cell activation in a cross-sectional study of 30 elite controllers -- people who maintain an undetectable viral load without cART.48 This study linked lower CD4 counts to higher levels of activated CD4s and CD8s (rho = -0.52, P = 0.003 for activated CD4s and rho = -0.37, P = 0.047 for activated CD8s).

HOPS investigators38 noted that lower CD4 counts mean higher activated CD4 numbers, and activated CD4 cells turn up in atherosclerotic lesions in the general population.49 The HOPS team also observed that the chronic inflammation seen in advanced HIV infection is driven by "the same inflammatory cells and proinflammatory cytokines that destabilize atherosclerotic plaques,"38 resulting in plaque rupture and coronary artery thrombosis.49-51

Does cART Prevent Heart Disease?

Does the evidence tying higher viral loads and lower CD4 counts to a bigger heart disease risk mean clinicians should consider cART a component of cardiovascular disease prevention? That one might even ponder such a proposition is remarkable. Just over a decade ago, SMART trial investigators planned that seminal study to test the hypothesis that avoiding cART for planned intervals would ease the burden of major cardiovascular, kidney, or liver disease.52 In other words, plenty of HIV luminaries thought cART should be shunned when possible to trim the risk of heart disease -- and lots of clinicians felt the same way. SMART demolished that strategy. But do the data reviewed above mean clinicians should start cART earlier -- as soon as possible, US guidelines say7 -- not only to thwart AIDS but also to ward off cardiovascular disease and other portentous afflictions? Some of the researchers who ran the studies reviewed here think so:

Summing up their 6500-person study of CD4 and viral load impact on MI, Steven Grinspoon, Paul Sax, and other Boston researchers wrote that "treatment of HIV infection to improve immunologic function is likely to be an important component of cardiovascular prevention for HIV patients" and that "cardiovascular risk reduction might therefore be an additional benefit of earlier initiation of ART."34

The HOPS team believes their findings "support prior observations that HIV infection in itself is a risk factor for cardiovascular disease not dissimilar in magnitude to some traditional risk factors for cardiovascular disease events," and they call for "randomized controlled trials to assess whether earlier initiation of antiretrovirals and avoidance of treatment interruptions will reduce the incidence of cardiovascular events."38

In February 2013 antiretroviral guidelines, US experts maintained that "increased risk of cardiovascular events with treatment interruption, the effects of ART on markers of inflammation and endothelial dysfunction, and the association between cardiovascular disease and CD4 cell depletion suggest that early control of HIV replication with ART can be used as a strategy to reduce risk of cardiovascular disease, particularly if drugs with potential cardiovascular toxicity are avoided."7 But they stress that research has yet to prove that cART prevents cardiovascular disease, and "therefore, a role for early ART in preventing cardiovascular disease remains to be established."


  1. Mocroft A, Reiss P, Gasiorowski J, et al. Serious fatal and nonfatal non-AIDS-defining illnesses in Europe. J Acquir Immune Defic Syndr. 2010;55:262-270.
  2. Friis-Møller N, Sabin CA, Weber R, et al; Data Collection on Adverse Events of Anti-HIV Drugs (DAD) Study Group. Combination antiretroviral therapy and the risk of myocardial infarction. N Engl J Med. 2003;349:1993-2003. Erratum in N Engl J Med. 2004;350:955.
  3. Phillips AN, Carr A, Neuhaus J, et al. Interruption of antiretroviral therapy and risk of cardiovascular disease in persons with HIV-1 infection: exploratory analyses from the SMART trial. Antivir Ther. 2008;13:177-187.
  4. Currier JS, Taylor A, Boyd F, et al. Coronary heart disease in HIV-infected individuals. J Acquir Immune Defic Syndr. 2003;33:506-512.
  5. Mary-Krause M, Cotte L, Simon A, Partisani M, Costagliola D; Clinical Epidemiology Group from the French Hospital Database. Increased risk of myocardial infarction with duration of protease inhibitor therapy in HIV-infected men. AIDS. 2003;17:2479-2486.
  6. Lang S, Mary-Krause M, Simon A, et al; French Hospital Database on HIV (FHDH)-ANRS CO4. HIV replication and immune status are independent predictors of the risk of myocardial infarction in HIV-infected individuals. Clin Infect Dis. 2012;55:600-607.
  7. Department of Health and Human Services. Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. February 12, 2013.
  8. Lang S, Mary-Krause M, Cotte L, et al. Impact of individual antiretroviral drugs on the risk of myocardial infarction in human immunodeficiency virus-infected patients: a case-control study nested within the French Hospital Database on HIV ANRS cohort CO4. Arch Intern Med. 2010;170:1228-1238.
  9. Worm SW, Sabin C, Weber R, et al. Risk of myocardial infarction in patients with HIV infection exposed to specific individual antiretroviral drugs from the 3 major drug classes: the data collection on Adverse Events of Anti-HIV Drugs (D:A:D) study. J Infect Dis. 2010;201:318-330.
  10. Boccara F, Lang S, Meuleman C, et al. HIV and coronary heart disease: time for a better understanding. J Am Coll Cardiol. 2013;61:511-523.
  11. Torriani FJ, Komarow L, Parker RA, et al. Endothelial function in human immunodeficiency virus-infected antiretroviralnaive subjects before and after starting potent antiretroviral therapy: the ACTG (AIDS Clinical Trials Group) Study 5152s. J Am Coll Cardiol. 2008;52:569-576.
  12. Coplan PM, Nikas A, Japour A, et al. Incidence of myocardial infarction in randomized clinical trials of protease inhibitorbased antiretroviral therapy: an analysis of four different protease inhibitors. AIDS Res Hum Retroviruses. 2003;19:449-455.
  13. Ding X, Andraca-Carrera E, Cooper C, et al. No association of abacavir use with myocardial infarction: findings of an FDA meta-analysis. J Acquir Immune Defic Syndr. 2012;61:441-447.
  14. Ribaudo JH, Benson CA, Zheng Y, et al. No risk of myocardial infarction associated with initial antiretroviral treatment containing abacavir: short and long-term results from ACTG A5001/ALLRT. Clin Infect Dis. 2011;52:929-940.
  15. Brothers CH, Hernandez JE, Cutrell AG, et al. Risk of myocardial infarction and abacavir therapy: no increased risk across 52 GlaxoSmithKline-sponsored clinical trials in adult subjects. J Acquir Immune Defic Syndr. 2009;51:20-28.
  16. D:A:D Study Group, Sabin CA, Worm SW, Weber R, et al. Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients enrolled in the D:A:D study: a multi-cohort collaboration. Lancet. 2008;371:1417-1426.
  17. Bavinger C, Bendavid E, Niehaus K, et al. Risk of cardiovascular disease from antiretroviral therapy for HIV: a systematic review. PLoS One. 2013;8:e59551.
  18. Obel N, Farkas DK, Kronborg G, et al. Abacavir and risk of myocardial infarction in HIV-infected patients on highly active antiretroviral therapy: a population-based nationwide cohort study. HIV Med. 2010;11:130-136.
  19. Lundgren JD, Neuhaus J, Babiker A, et al. Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients. AIDS. 2008;22:F17-F24.
  20. Friis-Møller N, Reiss P, Sabin CA, et al. Class of antiretroviral drugs and the risk of myocardial infarction. N Engl J Med. 2007;356:1723-1735.
  21. Daftary M, Dutta AP, Xue Z, et al. Women's Intragency HIV Study (WIHS): cardiovascular outcomes in women on PI therapy. XV International AIDS Conference. July 11-16, 2004. Bangkok. Abstract MoPeC3592.
  22. Holmberg SD, Moorman AC, Tong TC, et al. Protease inhibitor drug use and adverse cardiovascular events in ambulatory HIV-infected patients. The XIV International AIDS Conference. July 7-12, 2002. Barcelona. Abstract TuPeB4494.
  23. Rickerts V, Brodt H, Staszewski S, Stille W. Incidence of myocardial infarctions in HIV-infected patients between 1983 and 1998: the Frankfurt HIV cohort study. Eur J Med Res. 2000;5:329-333.
  24. Kwong GP, Ghani AC, Rode RA, et al. Comparison of the risks of atherosclerotic events versus death from other causes associated with antiretroviral use. AIDS. 2006;20:1941-1950.
  25. Iloeje UH, Yuan Y, L'Italien G, et al. Protease inhibitor exposure and increased risk of cardiovascular disease in HIV-infected patients. HIV Med. 2005;6:37-44.
  26. Barbaro G, Di Lorenzo G, Cirelli A, et al. An open-label, prospective, observational study of the incidence of coronary artery disease in patients with HIV infection receiving highly active antiretroviral therapy. Clin Ther. 2003;25:2405-2418.
  27. Islam FM, Wu J, Jansson J, Wilson DP. Relative risk of cardiovascular disease among people living with HIV: a systematic review and meta-analysis. HIV Med. 2012;13:453-468.
  28. Monforte A, Reiss P, Ryom L, et al. Atazanavir is not associated with an increased risk of cardio or cerebrovascular disease events. AIDS. 2013;27:407-415.
  29. Hill A, Sawyer W, Gazzard B. Effects of first-line use of nucleoside analogues, efavirenz, and ritonavir-boosted protease inhibitors on lipid levels. HIV Clin Trials. 2009;10:1-12.
  30. Ucciferri C, Falasca K, Vignale F, Di Nicola M, Pizzigallo E, Vecchiet J. Improved metabolic profile after switch to darunavir/ritonavir in HIV positive patients previously on protease inhibitor therapy. J Med Virol. 2013;85:755-759.
  31. Sax PE, Tierney C, Collier AC, et al. Abacavir-lamivudine versus tenofovir-emtricitabine for initial HIV-1 therapy. N Engl J Med. 2009;361:2230-2240.
  32. Post FA, Moyle GJ, Stellbrink HJ, et al. Randomized comparison of renal effects, efficacy, and safety with once-daily abacavir/lamivudine versus tenofovir/emtricitabine, administered with efavirenz, in antiretroviral-naive, HIV-1-infected adults: 48-week results from the ASSERT study. J Acquir Immune Defic Syndr. 2010;55:49-57.
  33. Currier JS, Lundgren JD, Carr A, et al. Epidemiological evidence for cardiovascular disease in HIV-infected patients and relationship to highly active antiretroviral therapy. Circulation. 2008;118:e29-e35. Erratum in Circulation. 2008;118:e108.
  34. Triant VA, Regan S, Lee H, Sax PE, Meigs JB, Grinspoon SK. Association of immunologic and virologic factors with myocardial infarction rates in a US healthcare system. J Acquir Immune Defic Syndr. 2010;55:615-619.
  35. Marin B, Thiébaut R, Bucher HC, et al. Non-AIDS-defining deaths and immunodeficiency in the era of combination antiretroviral therapy. AIDS. 2009;23:1743-1753.
  36. Stein JH. Cardiovascular risk and dyslipidemia management in HIV-infected patients. Top Antivir Med. 2012;20:129-133.
  37. Achhra AC, Amin J, Law MG, et al; INSIGHT ESPRIT & SILCAAT study groups. Immunodeficiency and the risk of serious clinical endpoints in a well studied cohort of treated HIV-infected patients. AIDS. 2010;24:1877-1886.
  38. Lichtenstein KA, Armon C, Buchacz K, et al; HIV Outpatient Study (HOPS) Investigators. Low CD4+ T cell count is a risk factor for cardiovascular disease events in the HIV outpatient study. Clin Infect Dis. 2010;51:435-447.
  39. van Lelyveld SF, Gras L, Kesselring A, et al; ATHENA national observational cohort study. Long-term complications in patients with poor immunological recovery despite virological successful HAART in Dutch ATHENA cohort. AIDS. 2012;26:465-474.
  40. Data Collection on Adverse Events of Anti-HIV drugs (D:A:D) Study Group, Smith C, Sabin CA, Lundgren JD, et al. Factors associated with specific causes of death amongst HIV-positive individuals in the D:A:D Study. AIDS. 2012;24:1537-1548.
  41. Kaplan RC, Kingsley LA, Gange SJ, et al. Low CD4+ T-cell count as a major atherosclerosis risk factor in HIV-infected women and men. AIDS. 2008;22:1615-1624.
  42. Ho JE, Deeks SG, Hecht FM, et al. Initiation of antiretroviral therapy at higher nadir CD4+ T-cell counts is associated with reduced arterial stiffness in HIV-infected individuals. AIDS. 2010;24:1897-1905.
  43. Manner IW, Trøseid M, Oektedalen O, Baekken M, Os I. Low nadir CD4 count predicts sustained hypertension in HIV-infected individuals. J Clin Hypertens (Greenwich). 2013;15:101-106.
  44. Lo J, Abbara S, Shturman L, et al. Increased prevalence of subclinical coronary atherosclerosis detected by coronary computed tomography angiography in HIV-infected men. AIDS. 2010;24:243-253.
  45. Hsue P, Hunt PW, Sinclair E, et al. Increased carotid intima-media thickness in HIV patients is associated with increased cytomegalovirus-specific T-cell responses. AIDS. 2006;20:2275-2283.
  46. Kaplan RC, Sinclair E, Landay AL, et al. T-cell activation and senescence predict subclinical carotid artery disease in HIV-infected women. J Infect Dis. 2011;203:452-463.
  47. Hunt PW, Martin JN, Sinclair E, et al. T cell activation is associated with lower CD4+ T cell gains in human immunodeficiency virusinfected patients with sustained viral suppression during antiretroviral therapy. J Infect Dis. 2003;187:1534-1543.
  48. Hunt PW, Brenchley J, Sinclair E, et al. Relationship between T cell activation and CD4+ T cell count in HIV-seropositive individuals with undetectable plasma HIV RNA levels in the absence of therapy. J Infect Dis. 2008;197:126-133.
  49. Hansson HK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med. 2005;352:1685-1695.
  50. Ross R. Atherosclerosis -- an inflammatory disease. N Engl J Med. 1999;340:115-123.
  51. Bursill CA, Channon KM, Greaves DR. The role of chemokines in atherosclerosis: recent evidence from experimental models and population genetics. Curr Opin Lipidol. 2004;15:145-149.
  52. Strategies for Management of Antiretroviral Therapy (SMART) Study Group, El-Sadr WM, Lundgren J, Neaton JD, et al. CD4+ countguided interruption of antiretroviral treatment. N Engl J Med. 2006;355:2283-2296.