Abstract Body

Immune Reconstitution Inflammatory Syndrome (IRIS) may complicate antiretroviral therapy (ART).  Whereas ART promotes HIV viral suppression and CD4+ T cell increases on a population level, the utility of these measurements as objective diagnostic criteria for IRIS remains unclear for an individual. We determined whether CD4 and plasma HIV RNA measurements are diagnostically useful to help differentiate paradoxical IRIS from culture-positive relapse.

In prospective cohorts with cryptococcal meningitis enrolled from 2006-2012, 70 HIV+ persons had 75 recurrent symptomatic meningitis episodes, due to paradoxical IRIS (n=62) or culture-positive relapse (n=13). IRIS was defined as per the INSHI consensus case definition, with CSF culture status being the primary determinant of the causative event. We compared CD4 count change, plasma HIV viral load change, and CSF cytokines (n=50) between IRIS vs relapse.

62 paradoxical cryptococcal-IRIS events did not consistently present with substantial reconstitution of CD4+ T cell counts (61% increased by >25 cells/mL), nor did these CD4 values distinguish IRIS from 13 culture-positive relapse events (54% with increases of >25 cells/µL).  All IRIS cases had excellent, appropriate virologic responses, yet 50% (6/12) with culture-positive relapse did not have HIV viral suppression (>40,000 copies/mL min). The median CSF WBC counts did not differ at time of cryptococcal-IRIS event (median 25; IQR, 5 to 85, max 240 cells/mcL) versus relapse (median 15; IQR, <5 to 40, max 1150 cells/mcL) (P=.53). Immune parameters which distinguished IRIS at event included 3-4 fold greater CSF levels of: Th1 cytokine IFN-g (P=.006), Th2 cytokine IL-4 (P=.011); pro-inflammatory IL-17(P=.038). Persons with culture-positive relapse did have markedly higher CSF IL-13 levels (relapse: median 253 (IQR, 63-471) pg/mL vs. IRIS: 6.9 (IQR, 2.4-17) pg/mL, P=.006) and non-statistically higher GM-CSF levels (P=.17).

CD4 counts, plasma HIV RNA, nor CSF WBCs at event onset did not consistently discriminated IRIS from relapse. The distinct immunologic signatures that did distinguish the two clinical scenarios, included 35-fold higher CSF IL-13 levels in relapse, a Th2 cytokine which is non-protective and associated with uncontrolled cryptococcal infection in murine models.  In IRIS, multiple T-cell cytokines were increased (IFN-g, IL-4, IL-17) at event. In resource-limited settings, those with CSF culture-positive relapse should be targeted for viral load testing.