Cryptococcal Antigenemia and the Implications of Viral Load–Directed Cryptococcal Antigen Screening in Antiretroviral Therapy–Experienced Patients With Human Immunodeficiency Virus

To the Editor

The World Health Organization (WHO) currently supports baseline CD4 testing for people living with human immunodeficiency virus (PLHIV) who are initiating or re-initiating antiretroviral therapy (ART) and cryptococcal antigen (CrAg) screening for persons with a CD4 count less than 100 cells/mL and up to 200 cells/mL within an advanced-disease package of care [1, 2]. Some PLHIV taking ART will experience viral load (VL) nonsuppression (VL ≥1000 copies/mL) and may not have CD4 results available to inform CrAg screening. Therefore, we read with interest the study of VL-directed CrAg screening to prevent cryptococcal meningitis (CM) and mortality by Mpoza et al [3]. While we applaud the authors for investigating this important and multifaceted topic, we disagree with several study assumptions and interpretations.

This cross-sectional study retrospectively conducted CrAg screening from a random sample of Ugandan adults taking ART with 1 VL ≥1000 copies/mL. Of 1186 samples tested for CrAg, 35 were CrAg-positive, yielding a 3.0% prevalence of CrAg and 4.2% among those with VL ≥5000 copies/mL. Post–positive CrAg screening, one would anticipate an intervention to improve outcomes. Six-month meningitis-free survival was reported as 61% (14/23; 95% confidence interval [CI], 35–76%), but only 42% (8/19; 95% CI, 20–67%) for those not receiving fluconazole (n = 19). However, to reason this is an improved outcome is not possible for several reasons. First, the number of CrAg-positive patients was small (n = 35), outcomes were limited (available for 25 patients, of whom 2 were lost to follow-up, 1 was already receiving fluconazole as secondary prophylaxis for CM, and 5 received fluconazole preceding CrAg testing by this study [8/25 patients already had routine, facility CrAg testing]). Second, individuals who did not receive fluconazole were included in both meningitis-free survival groups (61% vs 42%), and wide CIs make the groups difficult to distinguish. Third, not fully presented is whether or why not fluconazole was implemented for all those in need. Fourth, what the survival status was for 17 of 25 CrAg-positive patients receiving their first-ever CrAg test was not fully presented. Fifth, separate survival analysis was done for those lost to follow-up taking ART by extrapolating data from a different group, an ART-naive CrAg-positive cohort [4]. Finally, the cost-benefit analyses reported included testing costs alone but failed to consider cost of pre-emptive fluconazole therapy.

The benefits of CrAg screening are well described among those not taking ART and we acknowledge that ART-experienced persons with VL nonsuppression are at risk for cryptococcosis [5]. Yet, the authors failed to present data that a novel, VL-directed screening approach with an appropriate intervention clearly improved outcomes. We believe the evaluation of VL-directed CrAg screening and its impact on morbidity and mortality should be addressed prospectively. Also, operational matters need attention before changing guidelines (eg, VL results turnaround time) and ensuring anyone at risk of cryptococcal disease has rapid CD4 access, including those on ART with virological failure.

We commend the authors for their description of CrAg prevalence in those on ART with VL nonsuppression, however, more data would be needed to justify a recommendation that the WHO should change CrAg screening in PLHIV with an isolated high VL result.