Immune reconstitution inflammatory syndrome in the central nervous system: Limitations for diagnosis in resource limited settings

Abstract Introduction The diagnosis of IRIS is based on evidence of clinical worsening and immune reconstitution in the setting of combined antiretroviral therapy (cART) initiation. While central nervous system IRIS (CNS IRIS) is thought to be prevalent in resource limited settings (RLS), its identification is constrained by limited data on pre-treatment HIV disease and diagnostic testing. A diagnosis can be improved with neuroimaging and cerebrospinal fluid (CSF) studies, which are not universally available in RLS. This study evaluated whether diagnoses of CNS IRIS could be achieved in a resource limited setting based on established criteria. Methods A retrospective chart review of HIV+ individuals, on ARVs at the time of presentation in two cohort studies of suspected CNS opportunistic infection or tuberculous (TB) meningitis who were admitted to a tertiary care facility in Lusaka, Zambia. Results Using currently validated criteria, none of the 254 participants evaluated could be diagnosed with CNS IRIS, as there was no information on post-treatment trajectory of HIV viral loads or CD4 counts. Only one participant had a definitive, non-IRIS infectious diagnosis based on comprehensive testing. Of the remaining 253 patients, 68 (27%) had an identified potential CNS pathogen, 92 (36%) had inflammatory CSF in the absence of a pathogen, and 94 (37%) had normal CSF despite presenting with CNS symptoms. Conclusion The absence of HIV disease trajectory data, and lack of comprehensive diagnostic testing, compounded by a high prevalence of infectious pathogens, substantially limits the ability to diagnose CNS IRIS in RLS.