New antiretroviral agents have greatly increased the likelihood of achieving and maintaining suppression of HIV replication, even for patients highly experienced with antiretroviral therapy. For all patients, the goal is to maintain HIV RNA <50 copies/mL. However, not all patients are able to maintain this level of viral suppression on antiretroviral (ARV) therapy;if viral rebound occurs, HIV resistance emerges. Since resistance impairs viral fitness (compared to wild-type virus) at least in the short term, even non-suppressive therapy is beneficial. However, with continued evolution of HIV during non-suppressive therapy, in the longer term, viral replication increases (fitness improves) and the balance between resistance and clinical outcome may change. The clinical objective of this project is to evaluate the clinical consequences of drug resistance (defined by a new AIDS defining illness or death) after accounting for other contributing factors such as sequential therapies and partial responses to those treatments. This proposal will utilize data from the CNICS (CFAR Network of Integrated Clinical Systems) cohort, the first electronic medical record-based database repository. As CNICS is a clinic-based research network, it directly reflects the outcomes of clinical decisions and management options made daily in the care of HIV- infected individuals at seven large HIV clinics across the U.S. CNICS captures a broader range of information associated with the rapidly changing course of HIV disease management through collection of data at the point-of-care. HIV resistance data, collected in the most elemental form (codon sequences) are incorporated into CNICS by direct data transmission from the testing laboratories. With data from the CNICS, the specific aims of this project are: """""""" To determine the prevalence of antiretroviral regimen failure at seven clinical sites in the U.S. and define if the prevalence is stable or changing with time. """""""" To determine the prevalence of HIV drug resistance at seven clinical sites in the U.S. and define if the prevalence is stable or changing with time. """""""" To evaluate the relationships between antiretroviral drug failure, resistance, and clinical disease progression. We propose targeted analytical strategies, including novel techniques, to approach the multidimensional aspect of regimen sequence and genotypic and phenotypic HIV resistance data. Specific methods to analyze regimen data include nonparametric Bayesian models, Markov switching models and kernel-based methods. HIV resistance data will be summarized using kernel-tree and score-based metrics, as well as with Dynamic Bayesian Networks (DBN). Clinical outcome will be modeled with four approaches (Cox proportional hazards regression, structural, kernel, and DBN).
This project will use data from seven large HIV specialty clinics in the U.S. to determine how many patients have failed antiretroviral regimens and to determine how much HIV drug resistance is present in the patients being treated for HIV. We will use specialized statistical techniques to define the relationship between HIV drug resistance and progression of HIV clinical disease, defined as a new AIDS infection or cancer or death.
| Tilghman, M W; Pérez-Santiago, J; Osorio, G et al. (2014) Community HIV-1 drug resistance is associated with transmitted drug resistance. HIV Med 15:339-46 |
| Aldous, Jeannette L; Pond, Sergei Kosakovsky; Poon, Art et al. (2012) Characterizing HIV transmission networks across the United States. Clin Infect Dis 55:1135-43 |
| Poon, Art F Y; Aldous, Jeannette L; Mathews, W Christopher et al. (2011) Transmitted drug resistance in the CFAR network of integrated clinical systems cohort: prevalence and effects on pre-therapy CD4 and viral load. PLoS One 6:e21189 |
| May, Susanne; Gamst, Anthony; Haubrich, Richard et al. (2010) Pooled nucleic acid testing to identify antiretroviral treatment failure during HIV infection. J Acquir Immune Defic Syndr 53:194-201 |
| Aldous, Jeannette L; Haubrich, Richard H (2009) Defining treatment failure in resource-rich settings. Curr Opin HIV AIDS 4:459-66 |
| Poon, Art F Y; Brouwer, Kimberly C; Strathdee, Steffanie A et al. (2009) Parsing social network survey data from hidden populations using stochastic context-free grammars. PLoS One 4:e6777 |
