Pre-ART profile differs for mortality and TB-IRIS in people starting ART

Early mortality and TB-IRIS remain frequent among people starting ART with a low CD4 count in areas with epidemic HIV and TB. Researchers in Botswana and at the University of Pennsylvania hypothesized that because the immunopathogenesis of early mortality and TB-IRIS differ, people who die early would have a pre-ART immunologic profile different from people in whom TB-IRIS develops.

To test that hypothesis, the investigators conducted a prospective comparison of 170 antiretroviral-naive adults with active TB who began ART with a CD4 count below 125. All had pulmonary TB, and all were taking a standard TB regimen. No one had evidence of drug resistance, and none were pregnant or taking steroids.

The researchers divided participants into three groups according to outcome within 6 months of starting ART: 120 people (71%) who survived without TB-IRIS (the control group), 33 (19%) in whom TB-IRIS developed, and 18 (11%) who died. One person with TB-IRIS died, but it was unclear whether TB-IRIS was the immediate cause of death. The researchers recorded clinical factors potentially related to mortality or TB-IRIS, and they used the Luminex assay to measure levels of 29 soluble biomarkers before ART began. They analyzed Luminex values by logistic regression to identify predictors of early mortality or TB-IRIS.

The study group averaged 37 years in age (range 22 to 73), and 74 (44%) were women. Median HIV load when treatment began stood at 5.5 log (over 300,000 copies) and median CD4 count at 60 (interquartile range [IQR] 31 to 90).

Median time to death in the early-mortality group measured 49 days (IQR 31 to 84). In unadjusted analysis, four factors differed significantly between the early-death group and controls: female sex (67% versus 41%, P = 0.04), nevirapine-based therapy (33% versus 10%, P = 0.01), a non-TB opportunistic infection at baseline (44% versus 22%, P = 0.04), and median CD4 count (35 versus 64, P = 0.03). Only one factor differed significantly between the TB-IRIS group and controls, the proportion with body mass index below 19 kg/m2: 31% versus 51% (P = 0.05). Compared with controls, the TB-IRIS group tended to take nevirapine more frequently (21% versus 10%, P = 0.08).

Multivariate analysis adjusted for CD4 count, female sex, and presence of non-TB opportunistic infections identified higher pre-ART levels of MCP-1, IL-10, and IL-6 as independent predictors of early mortality, at the following adjusted odds ratios (aOR) (and 95% confidence intervals):

-- MCP-1: aOR 10.9 per 10-fold higher (1.3 to 94.0)
-- IL-10: aOR 3.9 per 10-fold higher (1.1 to 14.0)
-- IL-6: aOR 3.1 per 10-fold higher (1.0 to 9.3)

Most pre-ART cytokine levels were lower in people in whom TB-IRIS developed than in controls. In an analysis adjusted for baseline body mass index, pre-ART levels of six cytokines were independently associated with TB-IRIS:

-- GM-CSF: aOR 0.2 per 10-fold higher (0.04 to 0.5)
-- IL-15: aOR 0.4 for 10-fold higher (0.2 to 0.8)
-- IL-12p70: aOR 0.3 per 10-fold higher (0.1 to 0.7)
-- IL-12p40: aOR 0.3 per 10-fold higher (0.1 to 0.8)
-- IL-6: aOR 0.5 per 10-fold higher (0.2 to 1.0)
-- IL-17a: aOR 0.4 per 10-fold higher (0.2 to 0.8)

The researchers proposed that "pre-ART risk profiling of each outcome is possible and should be investigated in order to direct appropriate interventions to each patient population." They argued that studies of people with advanced TB and HIV should include both TB-IRIS and early mortality as outcomes.

Reference

1 Ravimohan S, Tamuhla N, Steenhoff AP, et al. Pre-ART immunologic profiles characterize risk of early mortality and TB-IRIS in HIV/TB co-infected adults. ICAAC 2014. September 5-9, 2014. Washington, DC. Abstract H-1199b.

Source: NATAP