P449 Semi-quantitative cryptococcal antigen rapid test (CryptoPS, Biosynex®) for cryptococcal meningitis in patients living with HIV in Sub-Saharan Africa: prospective multicenter diagnostic accuracy study (DREAMM)

Abstract

Poster session 3, September 23, 2022, 12:30 PM - 1:30 PMBackgroundCryptococcal meningitis (CM) remains a leading cause of HIV-related meningoencephalitis in African low- and middle-income countries (LMICs), causing 15%-20% of HIV-related deaths. Rapid Diagnostic Tests (RDTs) are powerful tools and key to speeding-up the diagnosis at the bedside, allowing for rapid and targeted treatment, especially in LMICs. For the past 10 years, Cryptococcal Antigen (CrAg) RDTs have a major role in CM management.Driving Reduced AIDS Meningo-Encephalitis Mortality (DREAMM) was a multicenter implementation science study and a capacity-building project to reduce the mortality of HIV-related central nervous system infections (CNS). One of the main DREAMM approaches was to improve the diagnosis of CNS infections at the bedside and in parallel in local laboratories. Within DREAMM, HIV-infected, adult people living with HIV (>18 years old) with suspected CNS infections were recruited in five hospital sites in Cameroon, Malawi, and Tanzania.ObjectivesOur objective was to evaluate the implementation of CrAg CryptoPS (Biosynex, Illkirch Graffenstaden, France), a new semi-quantitative RDT, in routine care settings in Sub-Saharan Africa.MethodsAll CrAg CryptoPS performed were compared to the reference CrAg lateral flow assay (Immy®). The evaluation was done by the local research teams in four DREAMM laboratories sites. CrAg CryptoPS's implementation was evaluated in 301 plasma samples and 258 cerebrospinal fluid (CSF) samples from 320 participants (patients diagnosed with cerebral toxoplasmosis did not have a lumbar puncture). In this analysis, the results will be considered in a binary way (positive/negative).ResultsBetween January 2018 and March 2021, 356 participants were prospectively enrolled with suspected HIV-related CNS infections, including CM, tuberculous meningitis, cerebral toxoplasmosis, and bacterial meningitis cases. Cryptococcal meningitis was the leading cause of CNS infections in Malawi and Tanzania with 66.3% (53/80) and 59.6% (59/99) cases respectively, and the second cause in Cameroon with 40.0% (39/90) cases after cerebral toxoplasmosis.In plasma, CryptoPS's sensitivity was 99.23% (95% CI, 0.98-1.01) and specificity was 94.15% (95% CI, 0.91-0.98); positive and negative predictive values were 92.8% and 99.4%, respectively. In CSF, the sensitivity and specificity of CryptoPS were 100% (95% CI, 0.0-0.0), and 99.26% (95% CI, 0.98-1.01), respectively; positive and negative predictive values were both 100%. A low number of false-positives were observed (<4% in plasma and <0.5% in CSF).ConclusionCryptoPS was evaluated in a context of hospitalized patients within a project including all causes of HIV-related CNS infection, not only CM. The sensitivity and specificity of CryptoPS calculated in these preliminary results are promising. Semi-quantitative CryptoPS has the potential to be used to tailor antifungal therapy but further optimizations need to be done prior to large-scale implementation in African LMICs. In addition, future work to determine CrAg antigen titres is planned, in the perspective to optimize treatment of CrAg positive cases who decline lumbar puncture.