CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

756 URINE-BASED TB SCREENING WITH TB-LAM AND ULTRA IN HIV+ UGANDANS WITH MENINGITIS Jayne Ellis 1 , Enock Kagimu 2 , Ananta Bangdiwala 3 , Michael Okirwoth 4 , Gerald Mugumya 4 , Vincent Wadda 4 , David R. Boulware 3 , Nathan C. Bahr 5 , Fiona V. Cresswell 4 1 University College London, London, United Kingdom, 2 Infectious Diseases Institute, Kampala, Uganda, 3 University of Minnesota, Minneapolis, MN, USA, 4 Infectious Disease Institute, Kampala, Uganda, 5 University of Kansas Medical Center, Kansas City, KS, USA Background: Tuberculosis (TB) is a common cause of HIV-related death, yet diagnosis is often missed, particularly with concurrent illness such as meningitis. In one study, use of urine TB-lipoarabinomannan lateral flow assay (LAM) reduced missed TB diagnoses and mortality in HIV+ inpatients with CD4 <100 or suspected TB. The utility of the novel Xpert MTB/Rif Ultra (Ultra) assay on urine has not been evaluated. We sought to determine the prevalence of disseminated TB by testing urine with LAM and Ultra in hospitalized adults with meningitis in Uganda. Methods: We prospectively enrolled HIV+ adults with meningitis in Kampala or Mbarara, Uganda. Participants were tested for meningitis etiologies using a stepwise algorithm. In parallel, participants underwent systematic urine-based screening for TB using the LAM (Alere) and Ultra (Cepheid). 60 μL of urine was tested with the LAM. All remaining urine was centrifuged and the cell pellet resuspended in 2mL of urine for Ultra testing. Results were reported to clinicians in real-time. Results: From Jan 2018 to Sept 2019, we enrolled 251 HIV+ inpatients. Table 1 shows baseline characteristics (median CD4 = 37 cells/mcL; IQR 12-85). The majority had cryptococcal meningitis (59%, 148/251), and 15% (38/251) had definite/probable TB meningitis (Table 1). Overall, 25% (63/251) had evidence of disseminated TB by either urine assay. In cryptococcal subjects, 20% (29/145) had evidence of disseminated TB by LAM and 5% (5/96) by Ultra. In definite/ probable TB meningitis, 32% (12/37) had a positive urine LAM and 33% (12/36) had a positive Ultra (Table 1). 178 participants had both urine LAM and Ultra results: 18% (32/178) were LAM positive, 11% (20/178) by Ultra, and 4% (8/178) positive by both assays. Mortality was higher in patients with evidence of disseminated TB by either urine assay (table 1). Conclusion: In hospitalized Ugandans with advanced HIV disease and suspected meningitis, systematic screening with urine LAM and Ultra found a high prevalence of disseminated TB (25%). Cryptococcosis and TB co-infection was common (20%). Given the overlap in symptoms, TB may be missed in this setting without systematic testing. In those with TB meningitis, urine tests were positive in one-third; these tests may represent rapid, non-invasive adjunctive tests for TBM diagnoses. There was little concordance of Ultra and LAM, the reason for which warrants further investigation.

755 CHARACTERISATION OF SOUTH AFRICA’S XPERT MTB/RIF ULTRA "TRACE" LABORATORY RESULTS Lesley Scott 1 , Pedro Da Silva 2 , Kyle Fyvie 1 , Gabriel D. Eisenberg 1 , Silence Ndlovu 2 , Puleng S. Marokane 2 , Wendy Stevens 1 1 University of the Witwatersrand, Johannesburg, South Africa, 2 National Health Laboratory Service, Johannesburg, South Africa Background: South Africa introduced Xpert-MTB/RIF Ultra (Ultra) assay into their national TB program in October 2017. Increased sensitivity of the Ultra over the previous Xpert MTB/RIF assay is attributed to the inclusion of IS6110/ IS1081, improved chemistry, and larger PCR reaction volume. The lower limit of detection of Ultra is 15.6cfu/ml, and a new semi-quantitative category “trace” identifies paucibacillary specimens that are IS6110/IS1081 positive but rpoB negative. The complexities of “trace” was explored. Methods: Exact demographic matching was applied to NHLS’s centralised laboratory test result data between Oct17-Nov18. This generated a cohort of uniquely identified patients (UIDs) with an initial “trace” result and at least one subsequent laboratory follow-up test (Ultra, smear, culture). Results: Overall, Ultra “trace” test results contributed an additional ~2% over the national ~10% positivity rate during the review period (see Figure). A total of 35623 “trace” UIDs were identified with 48.7% (n=17342) reflecting ≥1 additional laboratory follow-up test within the cohort time. Ultra was requested in 49.9% (n=8648/17342); culture 57.5% (n=9964/17342) and smear 64.2% (n=11133/17342) of cases. Follow-up occurred within 14 days of the first ultra “trace” result for 81.9% (n=14208/17342) of the cohort. Cases with a positive follow-up test were reported in 40.0% (n=6934) of cases: 52.9% (n=4575/8648) Ultra; 33.8% (n=3364/9964) culture [with rifampicin resistance confirmation]; 6.7% (n=750/11133) smear. 60.0% of (n=10408/17342) UIDs generated negative follow-up results: 47.1% (n=4073/8648) by Ultra; 66.2% (n=6600/9964) culture; 93.3% (n=10383/11133) smear. Follow-up Ultra generated the highest proportion of positive test results. Conclusion: Ultra’s “trace” category likely indicated TB disease in 40.0% of cases, which would have been undiagnosed by Xpert MTB/RIF testing, and at least 33.8% available for confirmation of rifampicin susceptibility by culture. Laboratory diagnostic algorithms can be refined to reduce testing costs and suggests clinical cohort studies are required to further explore “trace” for patient management.

Poster Abstracts

757 APPLICABILITY OF URINE LAM TEST IN ADVANCED HIV-INFECTED ADULTS IN UKRAINE

Marta Vasylyev 1 , Svitlana Antonyak 2 , Sergii Antoniak 2 , Natasha Rybak 3 , Vira Buhiichyk 1 , Oleksandra Sluzhynska 4 , Maryana Sluzhynska 5 , Viktor Tretiakov 6 , Tetiana Ismagilova 6 , Iryna Dizha 6 , Yana Sheremeta 6 , Dmytro Skirgiko 6 , Anastasiia Mazurenko 6 , Justyna D. Kowalska 7 1 Lviv Regional Public Health Center, Lviv, Ukraine, 2 Clinic of the Gromashevsky Institute of Epidemiology and Infectious Diseases, Lviv, Ukraine, 3 Brown University, Providence, RI, USA, 4 SALUS Charitable Foundation, Lviv, Ukraine, 5 Lviv Regional

CROI 2020 279