434   Utility of HIV Genotyping and Clinical Expert Advice—The Havana Trial.

C. Tural¹, L. Ruiz*¹, C. Holtzer², P. Viciana³, J. González⁴, E. Ferrer¹, J. Martínez-Picado¹, I. Ruiz⁵, D. Dalmau⁶, P. Domingo⁷, C. A. B. Boucher⁸, J. Schapiro⁹, J. Romeu¹, G. Sirera¹, B. Clotet¹, and the Havana Study Group.
¹HIV Clin. Unit and IrsiCaixa Retrovirology Lab., Hosp. Univ. Germans Trias i Pujol, Badalona, Spain;²Visible Genetics, Evry, Paris, France;³Hosp. Virgen del Rocío, Sevilla, Spain;⁴Hosp. La Paz, Madrid, Spain;⁵Hosp. Vall d'Hebró, Barcelona, Spain;⁶Hosp. Mútua de Terrassa, Barcelona, Spain;⁷Hosp. Sant Pau, Barcelona, Spain;⁸Univ. of Utrecht, The Netherlands; and⁹Stanford Univ., CA.

Background:We attempt here to demonstrate the potential roles of genotyping (Geno) and expert advice (EA) interventions may have in an ART-experienced patient population.

Methods:Final 24-week analysis of this prospective, multicenter, randomised trial. 274 ART-experienced patients were randomised to a change in ART based on genotyping (Geno⁺) or no genotyping (Geno^-) and to receive EA from a committee of virologists and clinicians or no EA. All genotyping was done using TRUGENE HIV-1 Genotyping kits (VGI) with results interpreted by a software program (Retrogram v.1.0, Virology Networks). A repeat-measure ANOVA test was used for treatment comparisons, and Scheffe's test was used for pair-wise analysis.

Results:At baseline subjects had mean CD4 = 388 cells/mm³and VL = 4.04 logs. At week 24 there was a significant difference in % BLD (<400 copies) between the Geno⁺and Geno^-arms (57.5% vs 42.4%, p = 0.01), the EA⁺and EA^-arms (59.1% vs 41.1%, p =0.003) and the EA⁺Geno⁺arm and EA^-Geno^-arm (69.2% vs 36.4%, p = 0.001), but not between the other arms. There was a significant difference in VL reduction between the Geno⁺and Geno^-arms (-1.1 log vs -0.8 log, p = 0.02) and the EA⁺and EA^-arms (-1 log vs -0.9 log, p = 0.03) In the analysis of + Geno arms, the 3rd + failure group had a significant VL reduction (-1.1 log vs. -0.7 log, p = 0.04), but the 1st and 2nd failure groups did not.

Conclusions:At the 6-month endpoint, we found that both genotyping with interpretation and EA improved virologic outcomes; there was a suggestion that EA may add to the benefit of genotyping and the value of genotyping in very heavily ART-experienced patients.