Our HIV clinic population is multiracial and international, with a high proportion of patients originating from Sub-Saharan Africa and significant numbers presenting

late with advanced HIV at diagnosis. We also sought to compare baseline characteristics of patients with and without cryptococcal antigenemia, in order to establish whether screening should be targeted at any specific groups. This was a retrospective cohort study conducted between April and October 2011 at Croydon University (previously Mayday) Hospital and St George’s Hospital in London. Newly diagnosed patients were identified from clinic and laboratory databases using the inclusion criteria: i) age ≥18 years; ii) new confirmed positive HIV see more serology diagnosed for the first time between January 2004 to October 2010, with stored serum or plasma available for testing; iii) CD4 count < 100 cells/μL; iv) not yet on ART at time of stored blood sample. The study was approved by the UK National Research Ethics committee and the Research and Development Office of St George's Hospital NHS Trust. St George's Hospital Virology laboratory stores serum for 2 years and plasma (HIV viral loads) for up to 10 years. Given the

use of retrospective stored samples, plus a requirement for samples to be at least 6 months old prior to testing (to allow patients to have become established on ART, such that any retrospective positive result would not impact current clinical Birinapant care), the requirement for informed consent was waived. Stored serum or plasma samples from time of initial HIV diagnosis were anonymised prior to testing. CRAG testing was performed on serum or plasma using the Cryptococcal Latex Agglutination test (Immuno-Mycologics Inc, USA), an antibody-agglutination reaction detecting the capsular polysaccharide antigen of C. neoformans with a specificity and sensitivity of >95%. Samples were incubated with Pronase(Roche) at 56 °C for 15 min and analysed according to manufacturers’

instructions. All samples were screened undiluted and at a 1:100 dilution. Any samples with a titre of ≥1:2 were defined as positive, and serially diluted twofold to determine the CRAG titre. Demographic and clinical data, including CD4 count at HIV diagnosis, age, sex, ethnic group, country Cyclin-dependent kinase 3 of origin and sexual orientation, were obtained from clinic databases by clinicians independent from the laboratory researchers. For any patients with cryptococcal antigenemia detected on retrospective testing of stored serum or plasma, clinical presentation at HIV diagnosis, results of relevant investigations, antifungal treatment, time to start of ART and development of incident or relapsed CM in the first 6 months on ART were obtained from medical notes and laboratory results review. Data were analysed using GraphPad Prism v5 (GraphPad Software, USA), using the t-test to compare continuous variables and the Fisher’s exact test for categorical variables.