CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

TUESDAY, FEBRUARY 24, 2015 Session P-Q6 Poster Session

Poster Hall

2:30 pm– 4:00 pm Aging: Frailty, Telomeres, and mtDNA 783 Frailty and Cause-Specific Hospitalization Among Persons AgingWith HIV and Drug Use Damani A. Piggott 1 ; Abimereki D. Muzaale 1 ; Shruti H. Mehta 1 ; Ryan P.Westergaard 2 ;ToddT. Brown 1 ; KushangV. Patel 3 ; Sean X. Leng 1 ; Gregory D. Kirk 1 1 Johns Hopkins University, Baltimore, MD, US; 2 University of Wisconsin, Madison, WI, US; 3 University of Washington, Seattle, WA, US

Background: Hospitalization events exact a substantial economic and clinical burden for aging HIV-infected populations. Frailty is a key aging-related syndrome, predictive of major adverse clinical outcomes, including all-cause hospitalization among older HIV-uninfected adults. We have previously reported the association of frailty with advanced HIV and mortality; however, limited data exist on the relationship of frailty to hospitalizations due to infectious or non-infectious causes among HIV-infected persons or their uninfected counterparts. Methods: Frailty was ascertained in the ALIVE cohort of persons with prior or current injection drug use based on the 5 Fried phenotype criteria. Hospitalization events were ascertained from 2005-2012 and categorized using Agency for Healthcare Research and Quality clinical classification software into: chronic disease, infectious disease, and non-chronic non-infectious conditions. Cox proportional hazards models were used to estimate the risk (hazard ratios [HR] with 95% confidence intervals [CI]) for time to first hospitalization for each category. Results: Among 1303 participants with a median age of 48 years, 32%were HIV infected, and 12%were frail. In multivariable models adjusting for age, sociodemographics, comorbidity, substance use, and HIV/AIDS status, frailty was significantly associated with chronic disease (aHR 2.03; 95% CI, 1.40, 2.96), and infectious disease (aHR 2.41; 95% CI, 1.54, 3.76) hospitalization; but not with non-chronic non-infectious hospitalization (aHR 1.07; 95% CI, 0.72, 1.59). A prior AIDS diagnosis was associated with increased hospitalization risk in all 3 categories. Among HIV-infected persons, independent of CD4 count, HIV viral load, or prior AIDS, frailty was significantly associated with increased AIDS hospitalization risk (aHR 6.30; 95% CI, 1.20, 33.1). Frailty was also independently associated with non-AIDS infectious disease hospitalization risk (aHR 2.21; 95% CI, 1.40, 3.50). Conclusions: The frailty phenotype selectively predicts vulnerability to chronic disease and infectious disease related hospitalization, independent of comorbidity, degree of HIV immunosuppression and virologic control. Frail persons are susceptible to increased hospitalization for both AIDS and Non-AIDS infection. Further elucidation of frailty pathways may facilitate targeted interventions to reduce health care utilization and improve clinical management for aging HIV-infected persons and their high risk counterparts. 784 Association of HIV Viral Load and Shorter Telomere Length Shawn Gogia ; Jue Lin;Yifei Ma; Rebecca Scherzer; Elizabeth Blackburn; Ramin Farzaneh-Far; Steven Deeks; Priscilla Hsue University of California San Francisco, San Francisco, CA, US Background: Telomeres are DNA sequences on the ends of chromosomes that protect genomic integrity and shorten as cells age. Telomerase counteracts telomere shortening and prevents cellular aging. Telomerase may be inhibited by reverse transcriptase inhibitors and telomeres in T cells may be shortened by excess HIV-associated proliferation. Given association of HIV with CVD and other conditions associated with aging, we initiated systematic studies of telomere length (TL), telomerase activity (TA) and vascular function in a diverse cohort of HIV-infected and uninfected adults. Methods: We measured TL, TA, and endothelial function in 65 HIV-infected and 24 uninfected controls. Mean PBMC TL was measured from DNA using qPCR and TA was assessed using the Gel-TRAP assay. Endothelial function was measured using flow-mediated vasodilation of the brachial artery (FMD). Generalized linear regression models with log link function were used to compare levels of TL and TA and to identify associated factors. Results: HIV+ and control subjects had similar demographic and traditional CVD risk factors. Median TL was 0.95 in treated and virally suppressed patients and 0.93 in untreated and non-suppressed, compared with 1.07 in uninfected controls (p=0.046), while median TL was 1.02 in untreated, virally suppressed patients (p=0.72 vs. controls). TL remained 12% shorter (p=0.004) in untreated individuals with detectable viremia compared with controls after adjustment for age, gender, and race. Among HIV patients, age (10% shorter per decade, p<0.001), viral load >10K copies/mL (21% shorter vs VL<2K, p<0.001), and higher sCD163 (9% shorter per doubling, p=0.018) were independently associated with shorter TL. Correlations with FMD were weak for both TL (r=0.12, p=0.34) and TA (r=0.11, p=0.46). Conclusions: HIV+ individuals have shorter TL compared to controls, a finding that appears to be driven by HIV replication. In particular, we found that having an HIV RNA level >10,000 copies/ml was associated with a larger reduction in TL than a decade of aging. Higher sCD163, a marker of monocyte/macrophage activation, was also independently associated with shortened TL. These data support the concept that HIV compromises telomere maintenance, perhaps with partial reversal by ART. Because TL and TA may also be influenced by monocyte/macrophage activation and other processes, larger studies are necessary to define the role of TL and telomerase in HIV+ persons. 785 Novel Mechanisms of Nucleoside Analog Associated Mitochondrial DNA Mutation Kristian Gardner; Patrick F. Chinnery; Brendan A. Payne Newcastle University, Newcastle-upon-Tyne, United Kingdom Background: Mitochondrial DNA (mtDNA) mutation is a key feature of human aging. NRTI therapy may lead to accelerated accumulation of mtDNA mutations, but the mechanisms are unknown. Methods: We used a transmitochondrial cybrid cell line containing a single 7.5kb mtDNA deletion mutation, and fibroblasts from elderly individuals containing the m.414T>G point mutation. Cells were exposed to NRTIs for 32 days. MtDNA content was determined by multiplex real-time PCR. Deletion mutations were analysed by multiplex real-time PCR and long-range PCR. Point mutations were analysed by pyrosequencing (Qiagen) and massively parallel resequencing (Illumina MiSeq). Results: In ddI-treated cybrids, heteroplasmy level of the deletion mutation increased from 75% to 96% (SD 1.6%, p<0.01). This was due to selective depletion of wild-type mtDNA content (to 14% of baseline, SD 3%, p<0.001) but preservation of mutant mtDNA. Cybrids treated with physiological doses of d4T, AZT or TDF showed no significant changes in deletion heteroplasmy, however 10x physiological dose d4T did increase heteroplasmy to 86% (SD 4%). There was no evidence of new deletion mutation formation with any NRTI at normal or 10x dosing. In fibroblasts there was no significant change in the heteroplasmy level (~50%) of the m.414T>G point mutation with any of the NRTI exposures. However deep resequencing revealed multiple additional low-level (<5% heteroplasmy) point mutations. When subjected to mtDNA depletion (due to ddI), followed by repopulation of mtDNA content (i.e. a molecular bottleneck), significant shifts in low-level mutations were seen (ddI, mean shift 6.3%, SD 2.8%; untreated, mean 2.7%, SD 1.2%; p=0.01). This effect was accentuated with 10x physiological dose ddI exposure. No significant heteroplasmy shifts were seen with d4T, AZT or TDF. There was no new point mutation formation.

Poster Abstracts

479

CROI 2015