Abstract Body

Integrase inhibitors(II) are associated with an accelerated HIV-RNA decline and enhanced CD4 recovery. In late-presenters, these factors are associated with the immune reconstitution inflammatory syndrome (IRIS), a pathological immune reaction against antigens of opportunistic infections (OI). Whether the use of II containing cART is a risk factor for IRIS is unknown as phase-III trials of licensed II included few late presenters.

Observational study within the ATHENA cohort. Case finding by full chart review was done in all treatment-naïve patients initiating cART from 2009 onwards who were at increased risk for IRIS: those with CD4 ≤200cells/mm3, who were diagnosed with PCP, toxoplasmosis, Kaposi’s sarcoma, CMV disease, cryptococcosis, mycobacterial disease or PML and/or initiated corticosteroids ≤12months after cART-initiation and/or died ≤12months after cART-initiation. 2 definitions of IRIS were used: IRIS criteria by French (=French IRIS) and IRIS diagnosed by the treating physician(=clinical IRIS). Patient charts were reviewed for both definitions using a standardized CRF. The 2 primary outcomes were French IRIS and combined clinical or French IRIS. Cox regression was used to compare the risk of IRIS in II and non-II users, while controlling for potential confounders. Patients were censored when switching from INI to non-INI or vice versa.

369 of 3250 patients initiating first-line cART fulfilled in- and exclusion criteria for chart review with a mean viral load and CD4 count of 275423c/ml and 38cells/mm3. Most prevalent OI were PCP (N=172), Candidiasis (N=143), Mycobacterial infections (N=51) and Kaposi’s sarcoma (N=38). Any form of IRIS was observed in 26/69 (38%) of II-users compared to 47/300 (16%) in the non-II users (OR 3.2, 95%CI 1.8-5.8) (Table). Cox regression showed that use of II was independently associated with French as well as any form of IRIS (HR 2.6, 95%CI 1.3-5.1, p=0.004 and HR 2.6, 95%CI 1.6-4.4, p=0.0001).

Patients diagnosed with an OI and a CD4-count ≤200cells/mm3 initiating II-based cART had a more than doubled incidence of IRIS. If confirmed in future studies, initiating II-based cART in late-presenters with OI may have to be revisited, especially in resource limited-settings.