![]() |
9TH EUROPEAN AIDS CONFERENCE (EACS) 1st EACS RESISTANCE & PHARMACOLOGY WORKSHOP October 25 - 29, 2003 Warsaw, Poland |
| 3.4 Resistance Testing in Clinical Practice F6/7 - THE PROGNOSTIC VALUE TO PREDICT VIROLOGICAL OUTCOMES OF 14 DISTINCT SYSTEMS USED TO INTERPRET THE RESULTS OF GENOTYPIC HIV-1 DRUG RESISTANCE TESTING IN UNTREATED PATIENTS STARTING THEIR FIRST HAART |
| (1) Catholic University, Rome, Italy,2 Royal Free and University College Medical School, London, UK,3 University Tor Vergata, Rome, Italy,4 University of Milan, Italy,5 Hospital of Lecco, Lecco, Italy,6 University of Modena, Modena, Italy,7 Hospital of Grosseto, Grosseto, Italy |
Objective: To determine ability of systems to interpret baseline genotypes from antiretroviral naïve individuals to predict virological outcomes of the initial HAART.
Methods: Baseline isolates were retrospectively genotyped (ABI) and interpreted by 14 systems: the deduced genotypic susceptibility scores (GSS) of the subsequent initial HAART were related to the virologic outcomes. Multivariable Cox's and hierarchical linear models were used adjusting for baseline HIV RNA, mode of HIV transmission and use of SQVhg.
Results: 417 patients from the I.Co.N.A. cohort were analyzed. Median baseline CD4 and HIV-RNA were 263 cells/ml and 4.91 log10 c/ml. Prevalence of mutations (IAS-USA 2002) associated with NRTI resistance was 9.4%, NNRTI and major PI mutations were <1%. Using Cox's models, GSS from 5 systems predicted the time to achieve HIV RNA<500 copies/ml (hivresistanceWeb v3, ANRS AC-11 v2002, Rega 5.5, CHLuxemburg v4.0, decision trees from Geno2Pheno, with adjusted mean increases of the hazard of success ranging from 78% to 112% for 1 unit higher GSS). The time to the virologic rebound after week 24 was predicted by a single system (support vector machines regression from Geno2Pheno: for 1 unit higher GSS, 36% mean hazard reduction). Using multivariable linear models, GSS from 4 systems predicted the virologic changes (ANRS AC-11 v2002, Rega 5.5, support vector machines with or without regression): for 1 unit higher GSS, mean changes from baseline ranged from #0.18 to #0.26 log10 copies/ml.
Conclusions: Genotyping coupled with adequate interpretation in chronic naïve patients can usefully predict virologic outcomes of the initial HAART regimens.
Presenting Author: DR SIMONA DI GIAMBENEDETTO, Catholic University, Rome, Italy, Largo Francesco Vito 1, 00168, ROME, Italy, Phone: +390630154945
031025
31