CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

665 ALCOHOL USE AND RISK OF MYOCARDIAL INFARCTION (MI): DOES MI TYPE MATTER? Robin M. Nance 1 , Bridget M. Whitney 1 , Matthew Budoff 2 , Kristina Crothers 1 , W. C. Mathews 3 , Elvin Geng 4 , Michelle Floris-Moore 5 , Geetanjali Chander 6 , Bryan Lau 6 , Mary McCaul 6 , Richard D. Moore 6 , Matthew Feinstein 7 , Mari Kitahata 1 , Heidi M. Crane 1 , for the Centers for AIDS Research Network of Clinical Information Systems 1 University of Washington, Seattle, WA, USA, 2 University of California Los Angeles, Los Angeles, CA, USA, 3 University of California San Diego, San Diego, CA, USA, 4 University of California San Francisco, San Francisco, CA, USA, 5 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 6 Johns Hopkins University, Baltimore, MD, USA, 7 Northwestern University, Chicago, IL, USA Background: People living with HIV (PLWH) are at increased MI risk. MIs are classified into type 1 (T1MI) due to atherothrombotic coronary plaque rupture and type 2 (T2MI) from supply-demand mismatch such as with sepsis. Data on alcohol and MIs in HIV are limited, conflicting and do not distinguished MI types. Understanding the relationship between alcohol use and MI by type may clarify differences in prior study findings. Methods: PLWH in care from 6 sites completed tablet-based assessments ~every 6 months including alcohol use (AUDIT-C). Alcohol severity was defined by AUDIT-C score (0-12 points); alcohol and binge frequency were defined as number of drinking and binge days/month. Alcohol categories were none, mild and hazardous (AUDIT-C score of ≥5 for men, ≥4 for women). MIs were centrally adjudicated and categorized by type. Alcohol associations were examined using Cox models, adjusted for age, sex, race/ethnicity, hepatitis C, smoking, diabetes, hypertension, dyslipidemia, and kidney disease. All models adjusted for CD4 cell count and viral load as time-varying variables. We repeated models using time- updated alcohol use but as sensitivity models given the potential moderating effects. Due to prior studies identifying a high prevalence of “sick quitters” among non-drinkers, we repeated analyses among those with some alcohol use. Results: Among 12,800 PLWH, 64% drank alcohol, and there were 134 T1MI and 112 T2MI during follow-up. In adjusted analyses, those reporting higher baseline alcohol scores and frequency of alcohol use had lower T1MI risk; this association was not seen for binge drinking frequency or T2MI (Table). In analyses limited to those reporting alcohol use, associations were non-significant. In analyses of alcohol use categories, a significant protective association was seen for mild alcohol use vs. no use (0.54, 95% confidence interval 0.32-0.89, 0.02) and T1MI, but not for hazardous alcohol use. Sensitivity analyses with time-updated alcohol use showed similar results for alcohol severity although alcohol frequency was no longer significant for T1MI. Conclusion: These findings suggest a J-shaped curve for alcohol associations with T1MI for PLWH with some protective association seen for mild alcohol use although potentially driven by “sick-quitters”. They highlight the influence of different alcohol definitions and the importance of carefully considering the impact of “sick quitters”. These same associations were not seen for T2MI highlighting the benefits of adjudicating MIs.

664 DIFFERENCES IN TYPES OF MYOCARDIAL INFARCTIONS AMONG PATIENTS AGING WITH HIV Bridget M. Whitney 1 , Robin M. Nance 1 , Joseph Delaney 1 , Susan Heckbert 1 , Matthew Budoff 2 , Kevin High 3 , Alan Landay 4 , Matthew Feinstein 5 , Richard D. Moore 6 , W. C. Mathews 7 , Elvin Geng 8 , Michael Saag 9 , Mari Kitahata 1 , Heidi M. Crane 1 , for the Centers for AIDS Research Network of Clinical Information SystemsA 1 University of Washington, Seattle, WA, USA, 2 University of California Los Angeles, Los Angeles, CA, USA, 3 Wake Forest University, Winston-Salem, NC, USA, 4 Rush University, Chicago, IL, USA, 5 Northwestern University, Chicago, IL, USA, 6 Johns Hopkins University, Baltimore, MD, USA, 7 University of California San Diego, La Jolla, CA, USA, 8 University of California San Francisco, San Francisco, CA, USA, 9 University of Alabama at Birmingham, Birmingham, AL, USA Background: The Universal Definition classifies MI by type according to the mechanism of myocardial ischemia. Type 1 MI (T1MI) result spontaneously from atherosclerotic plaque instability. Type 2 MI (T2MI) are secondary to other causes such as sepsis and cocaine-induced vasospasm resulting in oxygen demand-supply mismatch. We previously demonstrated that, in contrast to the general population, almost half of MIs among people living with HIV (PLWH) are T2MI. We conducted this study to compare MI rates by type and age among PLWH. We hypothesized that increases in rates with older age would differ by MI type, and that in contrast to the general population, T2MI would be more common in younger individuals, but there would be a measurable rate of T1MI even among 18-30 year-old PLWH. Methods: Potential MI events were identified in the centralized data repository at 6 CNICS sites. Case identification criteria included MI diagnoses and cardiac biomarkers to optimize ascertainment sensitivity. For each potential MI, sites assembled de-identified packets with physician notes, ECGs, procedure results, and lab results. Two experts reviewed each packet followed by a 3rd if discrepancies occurred. Reviewers categorized each MI by type and identified causes for T2MI. By decade of age, we calculated T1 and T2MI rates and confidence intervals (CI) per 1000 person-years of follow-up. Rate ratios were calculated for rates of T2MI vs. T1MI per decade of age. Results: We included 564 T1MI (54%) and 483 T2MI (46%). T1MI rates increased with older age although T1MI occurred in all decades including young adults (Table). T2MI rates were significantly higher than T1MI rates for PLWH under 40 and increased with age among those over 40 (Table). T1MI rates were similar or higher than T2MI rates among those over 40 (significantly higher for those 61-70 years of age). Of note, there were differences in causes of T2MI among those at younger vs. older ages with cocaine-induced vasospasmmore common in younger PLWH while causes such as hypertensive urgency and arrhythmia were more common in older PLWH. Conclusion: We found that among PLWH rates of T2MI were higher than T1MI until age 40 differing fromwhat is seen in the general population, but rates of both were very high among older PLWH. Causes of T2MI differed by age with substance use prominent at younger ages and cardiovascular-related risk factors common at older ages. These results highlight the importance of evaluating MI by type among PLWH.

Poster Abstracts

CROI 2019 253

Made with FlippingBook - Online Brochure Maker