685   Comparison of HIV-1 Proviral DNA and Plasma RNA Genotyping and Its Utility in the Selection of Antiretroviral Therapy in Children.

E. R. Cooper*¹, D. Christiansen¹, R. M. Lloyd Jr.², H. J. Cabral³, S. Theodore¹, A. M. Regan¹, L. Nelson¹, and S. I. Pelton¹.
¹Boston Univ. Sch. of Med., Boston, MA;²Applied Sci. Inc., Suwanee, GA; and³Boston Univ. Sch. of Publ. Health, Boston, MA.

Background:Empiric selection of antiretroviral (ARV) regimens for experienced HIV-infected children has resulted in limited success. We evaluated whether knowledge of viral mutations (genotype) for selection of ARV improved outcome.

Methods:Genotyping for presence of mutations associated with resistance was done for 33 children. Plasma and proviral specimens were evaluated concurrently. New ARV regimens were selected for 13 children based on results of genotyping. A successful intervention was defined as a>1 log decline in viral load. Major mutations (MM) were defined as those substitutions known to confer resistance by themselves to specific ARV drugs.

Results:No significant differences were identified between plasma and proviral samples Wilcoxon signed-rank test (p = 0.21). For 23 of 28 paired samples, the number of MM was concordant (within 1). The number of MM did not correlate with increasing viral load (r = 0.004, p = 0.98) or declining CD4 count (r = 0.26, p = 0.17). In 13 patients, an ARV regimen of>3 drugs for which no MM were present could be identified. Six of the 13 (46%) demonstrated>1 log decline for>12 weeks. In 5 of the 7 children (71%) who failed to demonstrate a virologic response, compliance concerns were present.

Conclusions:Mutation patterns of resting (proviral) and replicating (plasma) population were highly correlated. Neither viral load nor CD4 count was correlated with number of MM. Selection of ARV therapy in conjunction with viral genotype for experienced children was successful when adherence to the new regimen was achieved.