CROI 2015 Program and Abstracts

Abstract Listing

Oral Abstracts

8

The Clinical Pharmacology of Medication Adherence Terrence Blaschke Stanford University, Stanford, CA, US

Recognizing that suboptimal adherence is the major cause of treatment failure and lack of effectiveness of ARVs for PrEP, scalable interventions to improve adherence are badly needed. However, there are multiple patterns of suboptimal adherence and an understanding of the definitions and taxonomy of adherence is essential in designing interventions (Vrijens et al., Br J Clin Pharmacol. 2012; 73:691-705) Many interventions have been proposed, but to date none have been scalable and sustainable (Medication Adherence Interventions, Evidence Report No. 208. AHRQ Publication No.12-E010-EF, 2012). A limitation of most studies is the absence of reliable measures of adherence before and after an intervention. Moreover, the duration of benefit in those studies showing some increase in adherence dissipates over several months. What is needed is a scalable approach to identifying suboptimal adherence, monitoring those more likely to continue or become poorly adherent due to known predisposing factors, then focusing interventions on that cohort of patients. Due to the correspondence of drug exposure to loss of efficacy or protection from infection, sparse sampling of dosing information is insufficient, and detailed dosing information itself, shared with the patient and the provider, can significantly improve adherence. Modern technology allows detailed dosing histories to be obtained unobtrusively and collected centrally at a point in time when interventions can be applied. Combined with approaches such as Managed Problem Solving (Gross et al., JAMA Intern Med. 2013; 173:300-306) and Lifetime HIV Antiretroviral Therapy Adherence Intervention: Timing Is Everything (Bangsberg and Haberer, JAMA Intern Med. 2013; 173:306-7), progress towards scaling interventions towards populations at high risk for suboptimal adherence is now within reach. 9 Epidemiological and Biostatistical Issues in Studying Rare Events in HIV Stephen J. Gange Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, US Studies of rare events have advanced scientific understanding of HIV prevention, treatment, and pathogenesis. While the definition of what is considered ‘rare’ is subjective, examples include individuals who: Remain free of HIV despite persistent high exposure; Acquire HIV in prevention studies & populations with low incidence; Exhibit extremes of disease progression: rapid disease progression, long-term survivors, elite suppressors; and/or Develop particular symptoms or disease conditions that might be the result of HIV infection, therapy, and/or comorbidities; In this presentation, I will discuss how rare events impacts and motivates innovative methods in (1) study design options for observational and interventional investigations; (2) measurement, including methods for protecting against error and bias; and (3) causal and statistical inference, including the trend away from probabilistic-based inference and the rise of machine-learning techniques. Vaccine Research Center, NIAID, NIH, Bethesda, MD, US The immune system is comprised of incredibly diverse sets of cells, each programmed to carry out overlapping sets of effector functions. Quantifying any one function provides an incomplete view of the immune response, as information about what other responses are generated is absent. Quantifying multiple responses is far superior, but when carried out on a bulk level, loses information about cellular heterogeneity, gene programs, and a myriad of interactions that may occur at the single-cell level. Since individual cells are the atomic unit of immune function, the maximum information content is achievable only by measuring these functions independently and simultaneously on a cell-by-cell basis. For this reason, flow cytometry is a powerful technology to assess immune function in settings like vaccination and pathogenesis. Taking advantage of the ability of this technology to sort individual cells while preserving viability and nucleic acids, we can extend the multiplexing of gene expression measurements. Recent advances in fluorescent probe technology now provides us with the capability to simultaneously and independently quantify more than 30 cell-associated proteins with unheralded sensitivity. Integrating this technology with single cell transcriptomics provides a unique view of cellular functions. We choose to quantify lymphocyte-centric genes, including those encoding transcription factors, signaling molecules, effector molecules, and regulatory molecules. On a single-cell basis, we can correlate protein expression with gene expression; e.g., discordant results for the same gene reveal post-transcriptional regulatory mechanisms. We have identified gene signatures associated with vaccine-elicited T cells as well as with productively SIV-infected cells in vivo. This technology gives us an unprecedented view into the complexity and range of immunological functions expressed by vaccine or virus-specific immune cells. Using this approach, we can search for correlates of clinical outcome based on either: quantitative gene expression; and/or cell subset representation, enumerated by groups of cells sharing gene expression profiles. These analyses give us new insights into functional immune states in pathogenesis, treatment, and vaccination. 11 Studying Heterogeneity With Single Cell RNA-Sequencing Simon Quenneville Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland Heterogeneity is a complicated problem to study. In many contexts, researchers use purified or cultured cells, treating them as a uniform population, while often a certain degree of heterogeneity is present. Treated in bulk, the large amount of individual will allow the buffering of extreme phenotypes. Heterogeneity is often suspected, but the means of exploring the phenomena were lacking. Recently, RNA-seq technology as been optimized for single-cell analysis. This allows us to look at expression levels of a large number of individual cells and genes to describe subpopulations. We have used this method to investigate specific problems in HIV infection, I will use our example to describe the huge possibilities of this technique. We and others have observed heterogeneity: activated CD4 T cells are not all permissive to HIV infection. For example, infecting activated cells with increasing doses of HIV particles leads to a maximum level of transduction. This maximum is also variable between individuals, ranging from a few percent cells being susceptible up to more than half. Purifying infected cells followed by a comparison with uninfected cells is interesting, but the infectious process itself is altering transcription. Single cell RNA-seq allowed us to investigate the heterogeneity in CD4 T cell population coming from permissive and non-permissive donors. We have been looking for subpopulations with differential permissivity, but also for cell markers that would allow us to identify and purify the “permissive sub-population”. This presentation will overview the process, flaws and limitation of this technique and describes the pipeline we are developing. SessionW3Workshop Room 6D 2:30 pm– 4:30 pm Frontiers in Laboratory Science 10 Measuring Immunity 1 Cell at a Time Mario Roederer

Oral Abstracts

86

CROI 2015