Tuberculous meningitis (TBM) has a very high mortality, especially in HIV-infected patients, and there is an
urgent need to improve treatment. Intracerebral inflammation has long been recognized as an important
determinant of TBM outcome and adjunctive anti-inflammatory corticosteroid treatment have been shown to
prevent death, but not disability, in HIV-uninfected people with TBM. Corticosteroids represent the only host
directed therapy of proven benefit in tuberculosis treatment, yet their effect is modest, their adverse effects
substantial, and the mechanism by which they reduce mortality is unknown. Furthermore, there is evidence for
heterogeneity in their effect, dependent upon inter-individual variation in the intracerebral inflammatory
response; and whether those with HIV-infection benefit from corticosteroids remains uncertain.
Cellular metabolism is critical for the function of immune cells. We recently found that high concentrations of
cerebrospinal fluid (CSF) tryptophan strongly predicted mortality in Indonesian TBM patients. Using genome-
wide SNP analysis we identified 11 quantitative trait loci that were associated with both CSF tryptophan
concentrations and survival in a separate patient cohort. Many questions remain. How is tryptophan
metabolism altered during TBM? How does it correlate with inflammation, immunopathology, and response to
corticosteroids? How is tryptophan metabolism genetically regulated? Lastly, and very importantly, what is the
effect of HIV co-infection on tryptophan metabolism before and during TBM treatment?
Our aim is to address these questions by integrating data and specimens from large studies in Vietnam and
Indonesia with state-of-the-art omics technology and systems biology in the Netherlands and USA. Based on
our previous study findings ULTIMATE’s first aim is to define tryptophan metabolism in 1500 TBM patients and
300 controls with other brain infections from previous studies in Indonesia and Vietnam, using state-of-the-art
LC-MS platforms. Our second aim is to define the genetic regulation of tryptophan metabolism in TBM with a
combination of genome-wide SNP-typing of DNA from the same patient group and whole genome sequencing
of a subset (n=200). Our third aim is to identify biomarkers predicting the effect of corticosteroids, and discover
potential new targets for host-directed therapy by integrating clinical and neuroradiological data with CSF
transcriptomics, metabolomics, and host genotyping of 600 patients randomized to corticosteroids or placebo.
We will address these 3 aims in HIV-infected and uninfected patients because HIV-associated TBM has higher
mortality and different immunopathology.
Our strong preliminary data, unique access to two large bioarchives and on-going randomized controlled trials
in TBM, and expertise in integration of large-scale clinical and multi-layer ‘omics’ data promises to provide a
step-change in understanding TBM pathogenesis and discovery of new targets for future host-directed