Headache as a Predictor of Cryptococcal Meningitis in Ambulatory Patients With Symptomatic HIV-associated Cryptococcal Antigenemia

Le résumé fourni par la source

Background: People with HIV-associated cryptococcal antigenemia are at high risk for meningitis and death. In resource-limited settings, lumbar puncture to evaluate for meningitis is infrequently performed in asymptomatic individuals; therefore, symptomatic individuals are prioritized. We evaluated the predictive value of meningeal symptoms for cryptococcal meningitis in people with cryptococcal antigenemia. Methods: We conducted a secondary analysis from 3 randomized clinical trials conducted in Uganda between 2017 and 2024. We included adults with meningeal symptoms who had cerebrospinal fluid cryptococcal antigen (CrAg) testing performed. Logistic regressions and classification trees were used to determine the associations between meningeal symptoms and meningitis. Area under the receiver operating characteristic curve (AUROC) and misclassification rate were used to evaluate model performance. Results: Among 344 participants, median CrAg titer was 1:1280 (interquartile ratio, 1:100-1:2560). Overall, 285 (83%) presented with headache and 205 (60%) participants had meningitis. Presence of headache was associated with meningitis (adjusted odds ratio 11.7; 95% confidence interval [CI], 5.23-28.50). The AUROC of headache alone to predict meningitis was 0.65 (95% CI, 0.61-0.69), with 95% sensitivity and 35% specificity. The AUROC improved to 0.86 (95% CI, 0.82-0.90) when stiff neck, photophobia, confusion, sex, and CrAg titer ≥1:160 were added to the logistic regression model. In the final classification tree, headache combined with CrAg titer ≥1:160 demonstrated the highest probability (79%) of meningitis. Conclusions: Headache and high CrAg titer is the most reliable predictor for meningitis. In the absence of plasma CrAg titration, symptoms of headache, stiff neck, photophobia, and confusion predict meningitis for individuals with cryptococcal antigenemia.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.