433   A Prospective, Randomized Study on the Usefulness of Genotypic Resistance Testing and the Assessment of Patient-Reported Adherence in Unselected Patients Failing Potent HIV Therapy (ARGENTA): Final 6- Month Results.

A. De Luca¹, A. Antinori², A. Cingolani¹, M. G. Rizzo¹, R. Murri¹, A. Ammassari¹, F. Baldini¹, S. Di Giambenedetto¹, P. Marconi¹, B. Ciancio¹, and R. Cauda¹.
¹Inst. of Clin. Infectious Diseases, Catholic Univ. of Sacred Heart and²Natl. Inst. for Infectious Diseases "L. Spallanzani" IRCCS, Rome, Italy.

Background:The relative importance of resistance-guided treatment decisions and of patient adherence assessment on virologic and immunologic outcome is not fully established.

Methods:Consecutive patients failing HAART, with HIV RNA (VL) >2,000 c/mL (bDNA, limit 50 c/mL), were prospectively randomized (1:1) to receive standard of care (SOC) or additional genotypic resistance information (G) (Visible Genetics). Treatment decisions were panel-discussed in both arms. Adherence was measured by a self-administered questionnaire. T-test was used to compare continuous and exact test for categorical variables. Logistic regression was used to analyse associations of variables with virologic endpoints.

Results:174 patients were randomized (89 to SOC, 85 to G). At baseline, median CD4 = 425, VL = 4.3 log, 25% had failed>3 HAART regimens, 41% experienced 3 drug classes, median resistance mutations = 7 (range 0—17); these were unbalanced between arms (SOC = 7, G = 8, p = 0.03). 123 pts (71%) filled out an adherence questionnaire: 43% of these reported last missed dose during last week (<95% estimated adherence = non-ADH;>95% adherence = ADH). By intent-to-treat (174/174 pts), pts with VL <500 c/mL at 3 mo were 12% in SOC and 27% in G (p = 0.02); at 6 mo 17% in SOC and 21% in G (p = ns). Among pts with baseline VL<4 log c/mL, those with VL<500 c/mL at 3 mo were 11% in SOC and 48% in G (P = 0.0008); at 6 mo 21% in SOC and 41% in G (P = 0.10). At 3 mo, pts with VL<500 were 7% in non-ADH with SOC, 19% in non-ADH with G, 25% in ADH with SOC and 32% in ADH with G. CD4 changes at 3 and 6 mo did not differ between randomization groups. Mean CD4 changes at 3 mo were +50/mL in ADH and -12 in non-ADH (p < 0.01), at 6-mo +62 in ADH and -13 in non-ADH (p < 0.01). Independently predictive of VL<500 c/mL at 3 mo were being assigned to G (OR 2.6; 1.1—6.0), a pre-randomization history of VL<500 (OR 2.9; 1.2—6.8), failing the 1st or 2nd HAART (OR 3.2; 1.0—10.0) and <95% adherence (OR 0.4; 0.1—0.9); at 6 mo independently predictive of VL<500 c/mL were baseline VL (OR 0.4; 0.2—1.0 for each log more), failing the 1st or 2nd HAART (OR 5.6; 1.2—25.9) and a pre-randomization history of VL<500 (OR 3.0; 1.2—7.5).

Conclusions:Genotype-guided treatment decisions benefit also heavily pre-exposed pts but in a time-limited fashion. Patient-reported adherence strongly influences virologic responses, and its assessment is more useful than genotyping in predicting CD4 responses. Maximal virologic benefit from genotype guidance is obtained in adherent pts and in those with lower VL at failure.