CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

completed treatment, those who discontinued were more likely PLWH (32% vs 14% p<0.01), younger (median age 34 vs 36, p=0.06), and male (80% vs 64%, p<0.01). Overall, the reasons for treatment discontinuation were death (28, 17%), treatment abandonment/loss to follow up (92, 57%), treatment toxicity (9, 6%), resistance detection (7, 4%), or other/unknown reasons (18, 11%). Compared to those without HIV, PLWH were more likely to discontinue due to toxicity (15% vs 1%) and death (33% vs 10%), but less likely to discontinue due to abandonment/loss to follow up (46% vs 62%) and resistance (0% vs 6%). PLWH who discontinued treatment were less likely to be on ART at baseline than PLWH who completed treatment within 9 months (29% vs 50%, p=0.01). CD4 count and viral load were similar among PLWH in all three groups. Patients who received treatment for >9 months, were similar to those who completed treatment within 9 months regarding sex, but were older (median age 43 vs 36, p<0.01) and more likely PLWH (27% vs 14%, p=0.02) – rates of ART at baseline were similar between the treatment completion groups. Conclusion: This study highlights suboptimal TB treatment completion rates, due primarily to death and loss to follow-up. This reinforces the need for early ART initiation and better-tolerated TB treatment regimens, particularly in PLWH.

742 HIV, TUBERCULOSIS, AND CHRONIC LUNG DISEASE AMONG KENYAN ADULTS Jerry S. Zifodya 1 , Tecla M. Temu 1 , Sarah Masyuko 1 , George Nyale 2 , Jerusha Nyabiage 1 , Dickens Onyango 3 , John Kinuthia 2 , Stephanie Page 1 , Sylvia LaCourse 1 , Carey Farquhar 1 , Kristina Crothers 1 1 University of Washington, Seattle, WA, USA, 2 Kenyatta National Hospital, Nairobi, Kenya, 3 Ministry of Health, Nairobi, Kenya Background: People living with HIV (PLWH) are at increased risk for non- communicable diseases such as chronic lung disease (CLD) and also remain at risk for pulmonary opportunistic infections including tuberculosis (TB) that may contribute to the development of CLD. We hypothesized that prior active TB would independently increase the risk for CLD and sought to determine whether HIV modifies the relationship between TB and CLD. Methods: This is a cross-sectional, interim analysis of a cohort of PLWH with well controlled HIV and uninfected adults in Kisumu, Kenya, enrolled from 12/2018 through 10/2019. All participants underwent standardized spirometry and a validated respiratory specific questionnaire. Prior active TB and pneumonia were based on self-report. Multivariable logistic regression was used to evaluate cofactors of CLD as defined by spirometry: obstructive, restrictive, and impaired (a composite of obstructive and/or restrictive patterns). Effect modification was evaluated by an interaction term between HIV and impaired lung function. Results: We have enrolled 474 participants (79% of target sample size), 246 PLWH and 228 uninfected. PLWH are more likely to report prior active TB (28% vs 4%, p<0.0001) and pneumonia (28% vs 19%, p=0.008). Impaired lung function is more common in PLWH compared to uninfected participants (21% vs 14%, p=0.07). Of those reporting prior TB, a similar proportion of PLWH and uninfected adults had impaired lung function (36% vs 30%, respectively; p>0.05). In multivariable analyses, prior TB was associated with impaired lung function among all participants (aOR 3.1, 95% CI 1.7-5.7, p=0.004) (Table). In separate multivariable analyses, prior TB was also associated with obstructive spirometry (aOR 2.7, 95% CI 1.3-5.8, p=0.009). An interaction between HIV and impaired spirometry was not statistically significant. Conclusion: In adjusted models, impaired lung function was not significantly associated with HIV but was strongly associated with prior active TB, which was more common in PLWH. Prior TB was also more likely to be associated with an obstructive as opposed to restrictive pattern. TB may potentially account for a greater prevalence of CLD in PLWH, but we did not find evidence for an interaction between HIV and TB in the risk for impaired spirometry in this ongoing cohort. Future work should investigate mechanisms and potential management strategies that could mitigate the risk of impaired lung function in those with prior TB.

Poster Abstracts

741 N-ACETYLCYSTEINE FOR ANTI-TB DRUG-INDUCED LIVER INJURY: A RANDOMISED CONTROLLED TRIAL Muhammed S. Moosa 1 , Gary Maartens 1 , Hannah Gunter 1 , Shaazia Allie 1 , Mashiko Setshedi 1 , Mohamed F. Chughlay 1 , Nicole Kramer 1 , Annemie Stewart 1 , Background: First-line anti-tuberculosis (TB) therapy can cause liver injury. N-Acetylcysteine (NAC) is widely used in patients with paracetamol toxicity and there is limited evidence of benefit in liver injury due to other causes. Methods: We conducted a randomised, double-blind, placebo-controlled trial to assess whether intravenous NAC improves liver recovery in adult patients admitted to hospital with first-line anti-TB drug induced liver injury (AT-DILI), diagnosed by alanine transaminase (ALT) ≥3 times upper limit of normal (ULN) with hepatitis symptoms or ALT ≥5 times ULN if asymptomatic. NAC was dosed as per paracetamol toxicity guidelines. The primary endpoint was the time for ALT to fall below 100 U/L. Secondary endpoints included duration of hospitalisation, in-hospital mortality and adverse events. We compared time to ALT<100 U/L and time to discharge from hospital using Kaplan-Meier analyses and log-rank tests. We included all participants who commenced NAC/placebo infusion in the analysis. Results: Fifty- three participants received NAC and 49 placebo. Mean age was 38 years (SD±10), 58 (57%) were female and 89 (87%) were HIV positive, 40 (45%) of whomwere on antiretroviral therapy. Median serum ALT and total bilirubin at presentation were 462 U/L (IQR 266-790) and 56 mmol/L (IQR 25-100) respectively. There was no difference in the time to ALT <100 U/L (figure 1A), with a median of 7.5 days (IQR 5.5 -11) and 8 days (IQR 5 -13) in the NAC and placebo arms respectively. Hospital stay was shorter in participants who received NAC (figure 1B), log rank p=0.0093; median hospital stay was 9 days (IQR 6-15) in the NAC arm and 18 days (IQR 10-25) in the placebo arm. Mortality was 14% and did not differ by study arm. The infusion was stopped early due to an adverse reaction in 5 participants, all of whomwere receiving NAC (nausea and vomiting in 3, anaphylactoid reaction in 1 drip site pain in 1). Conclusion: NAC did not shorten time to ALT< 100 U/L in participants with AT-DILI. However, NAC significantly reduced duration of hospital stay. NAC may reduce morbidity and hospitalisation costs in patients hospitalized with AT-DILI. Larger clinical trials are needed to confirm this finding. Wendy Spearman 1 , Mark Sonderup 1 , Karen Cohen 1 1 University of Cape Town, Cape Town, South Africa

CROI 2020 273