CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

4.0-5.6) log 10 CFU/mL CSF; p <.001). Of those receiving ART, 50% (151/300) had initiated ART ≤ 4 months, and 16% (48/300) had initiated ART ≤ 14 days. Persons starting ART ≤ 4 months prior were more likely to present with CSF pleocytosis (47% vs 30%; p =.003) compared to those initiating ART > 4 months prior to diagnosis. Among persons receiving ART for > 4 months, 82% had HIV viral loads > 1000 copies/mL. Two-week mortality did not differ by overall ART status (27% vs 26%; p =.86). However, 50% (24/48) of those receiving ART for ≤ 14 days died within 2-weeks compared with 18% (19/103) of those receiving ART for 15-122 days and 23% (35/149) of those receiving ART for > 4 months ( p <.001). Hazard ratio for mortality decreased as the duration from ART initiation to development of CM increased from 7 to 28 days ( Figure ). Conclusion: Cryptococcosis after ART initiation is common in Africa. Patients initiating ART who unmask cryptococcal meningitis are at a high risk of death. Immune recovery in the setting of CNS infection is detrimental, and management of this population requires further study. Implementing pre-ART cryptococcal antigen screening is urgently needed to prevent CM after ART initiation.

Persons receiving abbreviated durations of amphotericin had shorter median days of hospitalization [8 (IQR 7-14), 11 (10-14) or 15 (14-18)] days by respective duration group; p<.001), a lower incidence of phlebitis (0%, 8%, 20%; p=.04), and better 18-day CSF sterilization rates (72% in intermediate group vs. 50% in standard group; p=<.01). The incidence of CM relapse and grade ≥3 adverse events were similar across groups. Among 2-week survivors, there was no significant difference in 18-week mortality across customization groups (46%, 32% and 26% for short, intermediate, and standard groups; p=0.31). Conclusion: Patients with CM present with varying CSF fungal burdens. We have demonstrated the ability to attain adequate CSF sterility with shortened courses of amphotericin by following a customized approach to therapy based on initial fungal burden. While this approach may be challenging to follow in resource-limited settings, it could maximize therapeutic efficiency while minimizing adverse events and hospitalization costs. Maria D. Niembro-Ortega 1 , Rosa A. Martínez-Gamboa 1 , Brenda Crabtree- Ramírez 1 , Pedro Martínez-Ayala 2 , Andrea Rangel-Cordero 1 , Nancy G. Velázquez-Zavala 1 , Janet Santiago-Cruz 1 , Paulette G. Díaz-Lomelí 1 , Armando Gamboa-Domínguez 1 , Edgardo F. Reyes-Gutiérrez 1 , Pedro Torres-González 1 , José Sifuentes-Osornio 1 , Luis A. Ponce de Leon-Garduño 1 1 Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico City, Mexico, 2 Hospital Civil Fray Antonio Alcalde, Guadalajara, Mexico Background: Disseminated histoplasmosis (DH) is an AIDS-defining infection and the most common endemic fungal infection in North America. The current diagnostic tests for DH are time-consuming, expensive and not widely available. Recently a nested PCR targeting a 100-kDa protein (Hc100) from H. capsulatum was described for DH diagnosis. Additionally an urine antigen for DH diagnosis (IMMY alpha) is recently available for in-site use. We aim to determine the diagnostic performance of Hc100 PCR and IMMY alpha urine antigen in comparison with culture for DH in HIV infected patients. Methods: We conducted a double-blind, multicenter, prospective, diagnostic tests study from December 2015 to June 2017. Adult HIV-positive patients presenting with fever and at least two of the following criteria: lymphadenopathy, hepatomegaly, splenomegaly, mucosal ulcers, dermatologic lesions, gastrointestinal bleeding; cytopenias, elevation of AST, LDH or ferritin levels; or radiographic findings suggestive of extrapulmonary infection were included from 10 hospitals in Mexico. At least a BACTEC MYCO/F Lytic blood culture and a blood sample for Hc100 amplification were required for inclusion. Additional samples (bone marrow, urine, tissue) for each case were obtained according with the in-site physician´s criteria. Fungal culture and Hc100 were performed to every additional sample. All clinical specimens were analyzed at a central laboratory. Results: 270 patients with probable DH were included in the study. Fifty-four patients resulted in an alternative diagnosis (50% had mycobacterial infection). Seventy-seven cases were confirmed with DH (67 of them through culture isolation and 10 by histopathology only); 218 blood cultures, 171 bone marrow cultures and 53 cultures of tissue were compared with its corresponding Hc100 nested PCR, and 246 urine samples for the IMMY ALPHA Antigen EIA. (Table 1). The positivity rate of Hc100 PCR varied depending on the sample. The highest sensitivity (90%) and negative predictive value (96.7%) were observed from tissue samples. Specificity of the IMMY ALPHA Histoplasma antigen EIA when compared to any positive culture was 95% (95% CI 91.8-97.3). Conclusion: In HIV patients with suspected DH, Hc100 PCR and IMMY ALPHA Histoplasma antigen EIA had a good diagnostic performance in terms of sensitivity and negative predictive value. Both tests were rapid and accessible to implement in endemic areas were mycology laboratories are not available.

793 PCR AND URINE ANTIGEN FOR DIAGNOSIS OF DISSEMINATED HISTOPLASMOSIS IN AIDS PATIENTS

Poster Abstracts

792 CUSTOMIZED AMPHOTERICIN DURATION FOR CRYPTOCOCCAL MENINGITIS BASED ON FUNGAL BURDEN

Darlisha A. Williams 1 , Kathy Huppler Hullsiek 2 , Andrew Akampurira 1 , Tonny Luggya 1 , Edwin Nuwagira 3 , Lillian Tugume 1 , Edward Mpoza 1 , Reuben Kiggundu 1 , Kenneth Ssebambulidde 1 , Mahsa Abassi 2 , Conrad Muzoora 3 , David Meya 1 , David R. Boulware 2 , Joshua Rhein 2 1 Infectious Disease Institute, Kampala, Uganda, 2 University of Minnesota, Minneapolis, MN, USA, 3 Mbarara University of Science and Technology, Mbarara, Uganda Background: Successful induction therapy for cryptococcal meningitis (CM) relies on finding a balance between achieving fungal clearance and minimizing severe and potentially life-threatening amphotericin toxicity. In the ASTRO-CM trial, a customized approach to amphotericin induction therapy was employed based on initial fungal burden in cerebral spinal fluid (CSF). Here we report the feasibility and clinical outcomes of a customized approach to CM induction therapy. Methods: We prospectively enrolled 460 HIV-infected CM patients in Uganda fromMarch 2015 to May 2017. Cryptococcal cultures were obtained at baseline. Amphotericin was discontinued after 7 days if there was no fungal growth after 7 days of culture, or discontinued after 10 days if cultures were < 10,000 CFU/mL after 10 days of culture. Participants otherwise received 14 days of amphotericin. Participants were classified as receiving short (7-8 days; sterile culture), intermediate (9-12 days; <10,000 CFU/mL) or standard (13-14 days; ≥10,000 CFU/mL) durations of amphotericin. Clinical outcomes were compared. Results: A total of 319 (69%) persons survived to 2 weeks. Of survivors, 69% (219/319) received the intended duration of amphotericin per the outlined algorithm. In those following the intended algorithm, 5% (11/219) received short duration, 23% (50/219) received intermediate duration, and 72% (158/219) received standard duration of amphotericin induction therapy.

CROI 2018 298