icon-    folder.gif   Conference Reports for NATAP  
  17th CROI
Conference on Retroviruses
and Opportunistic Infections
San Francisco CA
February 16-19, 2010
Back grey_arrow_rt.gif
Heart Risk Factors Tied to Neurocognitive Skills in SMART Substudy
  17th Conference on Retroviruses and Opportunistic Infections, February 16-19, 2010, San Francisco
Mark Mascolini
A SMART trial neurology substudy underlined the complexities in evaluating and interpreting comorbidities and disease markers when it linked lower baseline neurocognitive performance to cardiovascular disease and two cardiovascular markers--but not to standard HIV dementia risk factors [1].
Edwina Wright and SMART coworkers surmised their findings suggest that "in HIV-infected persons with high CD4 cell counts, cardiovascular risk factors and disease may be more detrimental to neurological functioning than those [factors] more directly attributed to HIV, itself." But in their report the investigators did not postulate which mechanisms may bridge cardiovascular disease and neurocognitive impairment.
SMART, the watershed trial that tied intermittent antiretroviral therapy to higher rates of AIDS, major non-AIDS diseases, and death [2], included a neurology substudy that planned to compare neurocognitive function in people assigned to intermittent therapy and those assigned to steady treatment. Substudy participants completed five neuropsychological tests, and the investigators standardized raw test scores to z scores using US norms.
The substudy's primary neurocognitive yardstick was the aggregate quantitative neurocognitive performance z score, labeled QNPZ-5 to reflect the five tests. The SMART team defined neurocognitive impairment as a z score below -2 in two or more cognitive ability domains. The investigators also collected standard demographic and HIV clinical data as well as cardiovascular risk variables and data on depression and alcohol and drug use.
Wright and colleagues enrolled 292 people in the substudy, 145 from Thailand, 102 from Australia and North America, and 45 from Brazil. They stopped well short of the substudy recruitment goal of 600 because SMART ended sooner than almost anyone anticipated. Median age of substudy enrollees stood at 40 years, 42% were women, and 46% had more than 12 years of education. Median CD4 count at SMART entry measured 536 (interquartile range [IQR] 437 to 693) and median nadir count 225 (IQR 158 to 296). A large majority, 88%, had a viral load below 400 copies.
While 11% of substudy participants had a major ECG abnormality, 4% had diabetes, and 3% had prior cardiovascular disease (myocardial infarction, coronary heart disease, congestive heart failure, or peripheral vascular disease). Median total cholesterol was 196 mg/dL (IQR 168 to 222) and median low-density lipoprotein cholesterol 113 mg/dL (IQR 94 to 140). Almost 9% of participant were taking lipid-lowering drugs, and 11% were taking antihypertensives.
The group averaged a QNPZ-5 score of -0.7, denoting significantly below-average neurocognition compared with the general population (P < 0.001). A multivariate model weighed the impact of numerous variables on neurocognitive performance, including age, gender, race/ethnicity, education, location, CD4 count, prior cardiovascular disease, antihypertensives, and total cholesterol. Prior AIDS, baseline CD4 count, nadir CD4 count, a viral load below 400 copies, and antiretroviral central nervous system penetration effectiveness (CPE) score did not correlate with QNPZ-5 scores, but prior cardiovascular disease, antihypertensive use, and higher total cholesterol did:
· Prior cardiovascular disease: coefficient -0.7, P = 0.007 (meaning people with prior heart disease had a QNPZ-5 score an average 0.7 lower than people without prior heart disease)
· Antihypertensive use: coefficient -0.4, P = 0.02
· Every 10 mg/mL higher total cholesterol: coefficient -0.03, P = 0.01
People with prior cardiovascular disease had a 6 times higher risk of neurocognitive impairment on entering SMART (P = 0.02). Further analysis of 258 people who repeated the neurocognitive tests after 6 months discerned no difference between SMART arms in QNPZ-5. However, 84% of substudy participants randomized to intermittent therapy had restarted antiretrovirals by month 6 because of the protocol change that came with early unveiling of SMART results.
Wright and colleagues pointed out several limitations of their analysis--the relatively small sample, neurocognitive testing by study-trained staff instead of neuropsychologists, no clinical confirmation of neuropsychological impairment, use of US neuropsychological test norms in Thailand and Brazil, and the possibility of false-positive associations. Finally, the lack of association between cardiovascular risk factors and 6-month change in QNPZ-5 scores could reflect the short follow-up.
1. Wright E, Grund B, Robertson K, et al. CVD and CVD risk factors are associated with lower baseline neurocognitive performance in the SMART neurology substudy. 17th Conference on Retroviruses and Opportunistic Infections. February 16-19, 2010. San Francisco. Abstract 415.
2. Strategies for Management of Antiretroviral Therapy (SMART) Study Group, El-Sadr WM, Lundgren JD, Neaton JD, et al. CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med. 2006;355:2283-2296.