CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

viremia within 8 months of ART was associated with high risk of morbidity and mortality. Further exploration of reasons for the development/persistence of CMV viremia following ART is warranted.

754 RISKS OF OPPORTUNISTIC INFECTIONS FOR HIV +/- VETERANS UNDERGOING CANCER CHEMOTHERAPY

Alain Makinson 1 , Lesley S. Park 2 , Kimberly Stone 3 , Maria Rodriguez-Barradas 4 , Sheldon T. Brown 5 , Roxanne Wadia 6 , Kristina Crothers 7 , Cynthia L. Gibert 8 , Roger Bedimo 9 , Matthew B. Goetz 10 , Fatma Shebl 11 , Jacques Reynes 1 , Vincent Le Moing 1 , Keith M. Sigel 3 1 University Hospital Montpellier, Montpellier, France, 2 Stanford University, Stanford, CA, USA, 3 Icahn School of Medicine at Mt Sinai, New York, NY, USA, 4 Baylor College of Medicine, Houston, TX, USA, 5 James J. Peters VA Medical Center, Bronx, NY, USA, 6 VA Connecticut Healthcare System, West Haven, CT, USA, 7 University of Washington, Seattle, WA, USA, 8 Washington DC VA Medical Center, Washington, DC, USA, 9 VA North Texas Health Care Center, Dallas, TX, USA, 10 VA Greater Los Angeles Health Care System, Los Angeles, CA, USA, 11 Massachusetts General Hospital, Boston, MA, USA Background: Persons living with HIV (PLHIV) treated for cancer may be at increased risk of opportunistic infections (OIs) compared with uninfected patients. Methods: Using the Veterans Aging Cohort Study we evaluated OI incidence in 5,289 patients with malignancies diagnosed 1996-2018, and treated with chemotherapy. We identified zoster, cytomegalovirus (CMV), tuberculosis, Candida esophagitis, pneumocystis pneumonia (PCP), toxoplasmosis, Cryptococcus, atypical Mycobacterium, Salmonella bacteremia, histoplasmosis, coccidiomycosis or Progressive Multifocal leukopathy using ICD diagnosis codes, chart review confirmed, within 6 months of chemotherapy initiation. We fitted Poisson models to evaluate the association between HIV and OI risk overall, stratified by hematological and non-hematological cancers. We used inverse probability weighting (IPW) of HIV status to control for differences between groups and adjusted for prophylaxis use. Models including only PLHIV were fitted to evaluate risk factors for OI incidence. Results: Amongst 2,237 PLHIV, a greater proportion of malignancies were hematologic (29%) than in uninfected Veterans (16%). Median age was 58 years, 98%were males, 81%were current/ever smokers, 49%were African-American. PCP prophylaxis was used more frequently in PLHIV (42% vs. 5%; p<0.001). We confirmed 107 OIs in 101 subjects: Candida esophagitis (n=43), zoster (n=31), CMV (n=11), PCP (n=11), Cryptococcus (n=3), atypical Mycobacterium (n=6), Salmonella bacteremia (n=1), histoplasmosis (n=1). HIV was an independent risk factor for OIs (incidence rate ratio [IRR] 4.3; 95% CI: 2.3-8.1) after accounting for IPW and prophylaxis use. 83% of OIs in PLHIV occurred in the setting of a viral load >500 copies/mL and/or CD4 count <200/mm 3 . There were no definitive cases of PCP in non-hematologic tumors among PLHIV with CD4>200/mm 3 . In multivariable analyses of PLHIV only, HIV viremia (IRR 9.9; 95% CI: 5.9-16.5) and comorbidity burden (IRR 1.1; 95% CI: 1.05-1.3 for 1 point Charlson increase) were independently associated with OI risk. Conclusion: OIs were significantly increased in PLHIV with malignancies undergoing chemotherapy. However, our study does not support systematic PCP prophylaxis in PLHIV with non-hematological malignancies and controlled HIV-disease.

753 CMV VIREMIA IN PATIENTS WITH ADVANCED HIV INFECTION: A 48-WEEK FOLLOW-UP STUDY Paula Suanzes 1 , Adaia Albasanz 2 , Juliana Esperalba 1 , Candela Fernández 1 , Júlia Sellarés 3 , Ariadna Torrella Domingo 4 , Bibiana Planas 4 , Antonio Segura 3 , Mario Martin 4 , Jordi Navarro 3 , Joaquin Burgos 3 , Adrià Curran 3 , Esteban Ribera 3 , Vicenç Falcó 3 1 Hospital Universitario de la Vall d’Hebron, Barcelona, Spain, 2 Hospital Universitario de Bellvitge, Barcelona, Spain, 3 Hospital Universitario de la Vall d’Hebron, Barcelona, Spain, 4 Vall d'Hebron Research Institute, Barcelona, Spain Background: Nowadays, the incidence of CMV end-organ disease (EOD) is very low even though the prevalence of CMV viremia is around 30% in patients with HIV infection and ≤100 CD4 T-lymphocytes (TL). We hypothesize that immune reconstitution after initiation of antiretroviral therapy (ART), rather than anti-CMV specific treatment, is the best strategy to clear CMV viremia in patients without EOD. We aim to study the dynamics of CMV viral replication and the recovery of specific immune response against CMV after the initiation of ART. A pre-planned interim analysis of this study was presented at CROI 2017. Here we present the final results of a 48-week prospective study. Methods: A prospective observational study including patients with HIV infection and <100 CD4 TL was performed between September 2015 and July 2018. We determined HIV viral load (VL), CD4-TL and CMV VL by quantitative PCR at baseline, 4, 12, 24 and 48 weeks. We determined specific immune response against CMV (QuantiFERON-CMV®) at baseline and at 48 weeks. ART was started for all patients but only patients with CMV EOD received anti-CMV treatment. Statistical analysis: Friedman's (quantitative) and chi-square (qualitative) tests were used to assess the evolution over time. Results: Fifty-two patients were included, 19 (36.5%) were women, median age (IQR) was 43.8 (36.5-53.1) years. At baseline median (IQR) CD4-TL count was 30/µL (20-60) and median (IQR) HIV VL was 451,500 copies/mL (179,750- 1,285,000). Sixteen (30.8%) patients had detectable CMV viremia at baseline, 3 (7.7%) at 12 weeks and none at 48 weeks. Only 1 patient developed EOD (stomatitis) during follow-up. Six (11.5%) patients were lost to follow-up and 7 (13.4%) died, none of them related to CMV infection. Thirty-seven (71.2%) patients had specific CMV immune response at baseline compared to 27 (69.2%) at 48 weeks. The specific CMV IFN-γ response increased from baseline (median: 1.25, IQR: 0.12-5.24) to 48 weeks (median: 2.5, IQR: 0.1-7.425) in the 39 patients that completed the follow-up (p=0.07). Conclusion: The prevalence of CMV viremia in patients with advanced HIV infection is high but the incidence of CMV-EOD is low nonetheless. CMV viremia gets suppressed after starting ART without specific anti-CMV treatment.

Poster Abstracts

CROI 2020 278