Increased cardiovascular risk in HIV infection: drugs, virus and immunity

Abstract

1625–1627Recent studies have linked the use of specific anti-retroviraldrugsorclassesofdrugswithanincreasedriskinsubclinical atherosclerosis or actual cardiovascular events.The protease inhibitors were the first such class of‘culprits’ identified as increasing such a risk; however,results are conflicting [1–4]. Most recently, the nucleo-side reverse transcriptase inhibitors abacavir and didano-sine, originally not considered to have adverse con-sequences on the cardiovascular system, were reported tobe associated with a respective 90 and 49% increase inmyocardial infarctions (MIs) in a large cohort with over150000 person-years of follow-up [5]. Abacavir has apeculiar history. Although it is considered to be a‘metabolicallyfriendly’drug,asearlyas2005,theUppsalaMonitoring Centre, which analyses adverse reactionsreported in the WHO database, noted that there was asuspiciously high number of MIs in patients takingabacavir with and without protease inhibitors; however,in the sponsor’s trials of more than 9600 patients, noincrease in the riskof MI was observed [6]. Regardless ofthe contradictory data, it is difficult to ignore any factorsassociated with a 90% increase in a potentially morbidcomplication, but should we?Tocomplicatematters,HIVitselfhasbeenassociatedwithnearly a two-fold increase in risk of acute MI in HIV-infected individuals compared with HIV-uninfectedcontrols[7].Furthermore, recent datafromthe Strategiesfor Management of Antiretoviral Therapy (SMART)study, where patients who strategically interruptedantiretroviral treatment in order to limit their supposeddrug-related adverse effects, found themselves in theawkward position of having significantly more cardio-vascular events, so many in fact, this study wasprematurelyterminated[8].So,dothedrugshelporhurt?The contradictory findings from the studies that havereported clinical cardiovascular endpoints have limita-tions: many are observational with inherent biases,follow-up is relatively short, specific antiretroviral usagehistory is unclear and most have no HIV-uninfectedcontrols. In order to answer the very important questionofwhatexactlyisthecardiovascular riskforHIV-infectedindividuals taking antiretroviral therapy for the rest oftheir lives, an enormous study with well documented‘hard’ clinical endpoints would be required. The DataCollection on Adverse events of Anti-HIV Drugs (DAD)Study, although enormous, is not arandomized study butan analysis of data from a collaboration of observationalcohorts,eachwithitsownpotentialbiases.Inthatreport,517patientshadanMI[5];treatmentwasnotrandomizedand selection bias, particularly for a drug like abacavirwith no previous well known link to cardiovasculardisease, is a real possibility. A study with the power andcomplexity to answer the treatment-risk questionrigorously is probably not feasible. So what are thealternatives?The alternatives are to accept the data available fromstudies that have well documented clinical endpointswhile continuing to investigate subclinical atherosclerosiswith tests that have strong predictive values forcardiovascular disease, such as brachial artery flow-mediated dilation (FMD), carotid artery intima–media