Relation between Viral Load and CD4 Percentage in Children on Non-suppressive HAART Regimens
Andrew Steenhoff*1,2, R Rutstein1, K Propert2, S Douglas2, K Gebo3, G Siberry3, A Gaur4, R Warford5, S Spector6, and R Gross2
1Children`s Hosp Philadelphia, PA, US; 2Univ of Pennsylvania, Philadelphia, US; 3Johns Hopkins Univ, Baltimore, MD, US; 4St Jude Children`s Res Hosp, Memphis, TN, US; 5St Luke`s Roosevelt Hosp Ctr, New York, NY, US; and 6Univ of California, San Diego, US
Background: At least 25% of children prescribed HAART
have a persistently detectable viral load, and many of these have resistant
virus, which may differ in virulence from wild type virus. When to change
therapy in these children has not been explored in pediatrics where the number
of drugs available in pediatric formulations is limited and regimen
modifications carry additional logistic burdens compared to adults.
Methods: We assessed the relationship between viral
load and CD4 percentage using the multi-center HIV research network, a
prospective observational cohort study conducted between 2000 and 2005.
Eligible participants were perinatally HIV-infected children prescribed HAART
who were followed longitudinally at participating sites. Random effects cross-sectional
regression models elucidated the effect of viral load on CD4 percentage. Log viral
load was used as the predictor and the effect of other explanatory variables
including age, age at HAART initiation and years on HAART were studied.
Results: A total of 91 perinatally infected patients
on HAART provided 1270 paired CD4 percentage and viral load observations. The
sample was 64% female, 87% black, 11% Hispanic, with a median age of HAART
initiation of 12.4 years (range, 2 months to 22 years). The median CD4
percentage was 28 (range 0 to 66) and 56% had an HIV-1 RNA ≤400 copies/mL
(range 2.6 to 7.0). Cross-sectional regression demonstrated that a viral load of
2 log10 copies/mL was associated with a CD4 percentage of 30.3 and viral
load of 4 log10 copies/mL with a CD4 percentage of 24.6%. In
multivariable analysis, initiation of HAART at a younger age (p <0.001)
was associated with a higher CD4 percentage, whereas duration of HAART was not
significant (p >0.05). There was however interaction between duration
on HAART and the association between CD4 percentage and viral load (p <0.001)—longer
duration on HAART associated with lower CD4 percentage.
Conclusions: In this U.S. cohort of perinatally
HIV-infected children prescribed HAART, a viral load of 10,000 copies/mL was
associated with a CD4 percentage in the moderate risk area for
immunosuppression of <25%. This represents a possible viral load cut-point
for therapeutic intervention. Future studies will need to further explore the
dynamic relationship between viral load and CD4 percentage change, as well as
the effects of age at initiation of HAART and duration of HAART on immune
status in children.