CROI 2019 Abstract eBook

Abstract eBook

Oral Abstracts

inadequate disease control and/or exacerbations can contribute to supply- demand mismatch, and COPD increases risk for pneumonia, a common cause of sepsis. Further investigation is required to understand mechanisms for this association and to optimize preventative and therapeutic strategies.

Oral Abstracts

32 HIV POST SCD STUDY: 80% HIGHER RATE OF AUTOPSY-DEFINED SUDDEN ARRHYTHMIC DEATH IN HIV Zian H. Tseng , Ellen Moffat, Eric VIttinghoff, Annie Bedigian, Joseph K. Wong,

Philip Ursell, Andrew Connolly, Jeffrey Olgin, Priscilla Hsue University of California San Francisco, San Francisco, CA, USA

33 SUDDEN CARDIAC DEATH AMONG HIV-INFECTED AND -UNINFECTED VETERANS Matthew Freiberg 1 , Meredith S. Duncan 1 , Suman Kundu 1 , Asri Mumpuni 1 , Emily Epstein 1 , Annie Bedigian 2 , Amy C. Justice 3 , Eric VIttinghoff 2 , Zian H. Tseng 2 , for the Veterans Aging Cohort Study Investigators 1 Vanderbilt University, Nashville, TN, USA, 2 University of California San Francisco, San Francisco, CA, USA, 3 Yale University, New Haven, CT, USA Background: We have reported HIV infection as a risk factor for sudden cardiac death (SCD) in San Francisco County, but to date this association has not been examined in larger populations using chart-reviewed events. Here we examine the association between HIV infection and SCD in a large, national, cohort of HIV infected (HIV+) and uninfected veterans Methods: We analyzed data on 144,362 Veterans (30% HIV+) from the Veterans Aging Cohort Study, a prospective study of HIV+ veterans and age, sex, race/ethnicity and clinical site matched uninfected veterans. We followed veterans from their first clinical encounter on or after 4/1/2003 until SCD, non-SCD death, or censoring on 12/31/2014. Sudden cardiac death was determined using death certificates and manual review of the VA electronic health record (EHR). To meet our SCD definition, participants had to have cardiac cause of death on their death certificate. SCD was excluded for deaths in a hospital, hospice, or nursing home, or due to accidents, overdose, suicide, or homicide. SCD was also ruled out if in the year prior to death, EHR review revealed metastatic cancer or active treatment for cancer, use of high flow oxygen or dialysis, an AIDS defining illness, CD4<50 cells/mm3 within 6 months before death, DNR/DNI, a new significant health condition one month before death, or a life altering event within one year if this event resulted in an end stage disease or severe disability. We calculated rates of SCD by HIV status and used Cox proportional hazards regression to model the association between HIV infection and SCD, adjusting for demographics, prevalent cardiovascular disease, SCD risk factors, and possible confounders. In secondary analyses we compared SCD incidence in HIV+ subgroups defined by time-updated viral load and CD4 cell count to HIV uninfected veterans. Results: Participants had a mean age of (50±10.7 years), were mostly male (97.2%) and African American (47.3%) and were followed for a median of 9.0 years. After adjustment for demographics, prevalent cardiovascular disease, SCD risk factors, and other possible confounders, HIV+ veterans had a 14% higher risk of SCD (hazard ratio=1.14, 95% confidence interval 1.04-1.25) compared to uninfected veterans. The risk was highest among those with sustained high HIV viral loads or low CD4 cell counts (Table). Conclusion: HIV infected people have an increased risk of sudden cardiac death compared to uninfected people when they have sustained unsuppressed HIV viremia or low CD4 cell counts.

Background: Persons living with HIV have higher rates of CVD including acute MI, heart failure, and our group first reported high rates of out-of-hospital presumed sudden cardiac death (SCD) using World Health Organization (WHO) criteria. However, the precise incidence of actual sudden arrhythmic deaths (SAD) in HIV remains unknown. Methods: Between 2011 to 2016, we prospectively identified all incident deaths attributable to out-of-hospital cardiac arrest among individuals with and without HIV aged 18-90 in SF County for medical record review and comprehensive autopsy, toxicology, and histology via medical examiner surveillance of consecutive out-of-hospital deaths. Autopsy-defined SAD had no extracardiac cause of death or acute heart failure. Final cause was adjudicated by a committee of pathologists, cardiologists, HIV clinicians, and electrophysiologists. Results: 126 out-of-hospital HIV-infected deaths were identified, and 47 of these met WHO SCD criteria. The mean age was 65.6 years, 94%male, and 57% white. Compared to uninfected WHO-defined (presumed) SCDs (N=505), SCDs with HIV were more likely to have a history of MI, psychiatric disorder, cigarette smoking, and substance abuse. Similar to the general population, about half of WHO-defined SCDs in HIV were autopsy-defined SADs; the remainder were non-cardiac and included 16 due to occult overdose. Presumed SCDs with HIV were more likely to be due to occult overdose (13% vs 34%, p<0.0001) and renal failure (1% vs. 6%, p=0.003) as compared to uninfected presumed SCDs. Adjusted incidence ratios for WHO (presumed) SCD and autopsy-defined SAD were both significantly higher in HIV (IRR 1.82, 95%CI 1.4-2.4, p<0.0005 and IRR 1.83, 95%CI 1.2-2.8, P=0.006, see Figure). After adjustment for age, gender, heart disease and CAD, SCDs with HIV had 60% higher interstitial fibrosis by myocardial trichrome staining compared to uninfected SCDs. Conclusion: In this countywide postmortem study, 1/3 of apparent SCDs in HIV over a 5-year period were due to occult overdose. However, adjusted rates of both presumed SCDs and autopsy-defined SAD were 82% and 83% higher respectively in HIV compared to the uninfected population. Higher levels of cardiac fibrosis in HIV, a known substrate for SAD in the general population, may underlie the mechanism by which HIV increases risk for SAD. Development of criteria and evaluation for implantable defibrillators should be carefully considered in HIV as a means to prevent SAD in this high-risk population.

14

CROI 2019