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
 
 
 
Plasma HIV Load Strongest Predictor of CSF Load in CHARTER Cohort
 
 
  17th Conference on Retroviruses and Opportunistic Infections, February 16-19, 2010, San Francisco
 
Mark Mascolini
 
- Higher CSF loads alone did not predict worse neuropsychological performance on standard tests. But a CSF load equal to or higher than a plasma load did predict worse neuropsychological performance after adjustment for age, AIDS diagnosis, current CD4 count, CD4 nadir, HCV serostatus, comorbidity status, and other factors. -

 
Plasma viremia strongly predicted cerebrospinal fluid (CSF) HIV load in antiretroviral-treated and untreated members of the US-based CHARTER cohort [1]. In antiretroviral-naive people, older age independently raised the risk of a detectable HIV RNA in CSF. In treated people, a lower central nervous system penetration effectiveness (CPE) score predicted a detectable CSF load.
 
Scott Letendre (University of California, San Diego) and CHARTER coworkers explored demographic and clinical correlates of CSF load in 1221 people from six centers who gave plasma and CSF samples, including 379 (31%) not taking antiretrovirals. In the treatment-naive group, 207 people (55%) had "minimal comorbidities."
 
The study group had a median age of 43 years (interquartile range [IQR] 38 to 48), a median HIV duration of 11.8 years (IQR 6.2 to 17.2), a median current CD4 count of 414 (IQR 257 to 594), a median lowest-ever (nadir) CD4 of 175 (IQR 50 to 300), and a median viral load of 2.3 log (about 200 copies, IQR 1.7 to 4.1). Nearly two thirds of the cohort (62%) had an AIDS diagnosis, and 90% had better than 95% adherence measured over the preceding 4 days.
 
Among the 842 people taking antiretrovirals, 84% had an undetectable CSF load, while only 24% of the 379 people naive to therapy had an undetectable CSF load. Among the people with no antiretroviral experience, four factors correlated with higher CSF loads in multivariate analysis:
 
· Lower current CD4 count >> higher CSF load, P < 0.001
· Lower CD4 nadir >> higher CSF load, P < 0.001
· Older age >> higher CSF load, P < 0.001
· Higher plasma load >> higher CSF load, P = 0.002
 
In the 842 people taking antiretrovirals, eight variables predicted a detectable CSF load in univariate analysis:
· Higher plasma load >> detectable CSF load, P < 0.001
· Current CD4 count under 200 >> detectable CSF load, P < 0.001
· Nonwhite ethnicity >> detectable CSF load, P < 0.001
· Less than 95% adherence >> detectable CSF load, P < 0.001
· Shorter duration of current regimen >> detectable CSF load, P < 0.001
· Higher number of past antiretrovirals >> detectable CSF load, P = 0.003
· Lower 2010 CPE score >> detectable CSF load, P = 0.007
· Older age >> detectable CSF load, P = 0.03
 
The 2010 update of the CPE score ranks three antiretrovirals as the best penetrators with a score of 4, none of them widely used in developed countries (zidovudine, nevirapine, and indinavir/ritonavir), followed by 10 antiretrovirals with a score of 3 (abacavir, emtricitabine, delavirdine, efavirenz, darunavir/ritonavir, fosamprenavir/ritonavir, unboosted indinavir, lopinavir/ritonavir, maraviroc, and raltegravir), seven with a score of 2 (didanosine, lamivudine, stavudine, etravirine, unboosted atazanavir, atazanavir/ritonavir, fosamprenavir/ritonavir), and eight with a low score of 1 (tenofovir, zalcitabine, nelfinavir, ritonavir, unboosted saquinavir, saquinavir/ritonavir, tipranavir/ritonavir, enfuvirtide).
 
The updated CPE score yielded a better (nearly stepwise) correlation with detectable CSF load than the 2008 score. Whereas 39% of cohort members with a score of 3 or lower had an detectable CSF load, 18% with a score of 6, 10% with a score of 8, and 15% with a score of 9 or higher had detectable virus in CSF.
 
Among people not taking antiretrovirals, 49 had a CSF load as high as or higher than their plasma load, while 330 had a CSF load lower than their plasma load (13% versus 87%, P = 0.01). Among people with minimal comorbidities while taking antiretrovirals, 29 had a CSF load at least as high as their plasma load and 178 had a CSF load lower than their plasma load (14% versus 86%, P = 0.001). Higher CSF loads alone did not predict worse neuropsychological performance on standard tests. But a CSF load equal to or higher than a plasma load did predict worse neuropsychological performance after adjustment for age, AIDS diagnosis, current CD4 count, CD4 nadir, HCV serostatus, comorbidity status, and other factors.
 
Letendre noted that the mechanisms explaining why age and nonwhite ethnicity affect CSF load remain undetermined, but he had some ideas: Both age and nonwhite ethnicity correlated with a longer duration of HIV infection and with more advanced HIV disease; and nonwhite ethnicity correlated with HCV coinfection.
 
Reference
 
1. Letendre S, FitzSimons C, Ellis R, et al, and the CHARTER Group. Correlates of CSF viral loads in 1221 volunteers of the CHARTER cohort. 17th Conference on Retroviruses and Opportunistic Infections. February 16-19, 2010. San Francisco. Abstract 172.