Microbial translocation and immune activation – a substudy of the COSTOP trial

Lead Research Organisation: MRC/UVRI Uganda Research Unit on AIDS


In Uganda (and other countries in Africa), HIV positive persons are advised to take cotrimoxazole (commonly known as Septrin), every day, since studies have shown that this leads to a reduction in the frequency of common illnesses that affect persons with HIV and improves their health. The mechanism underlying this benefit of cotrimoxazole is not clear. Cotrimoxazole kills certain bacteria and has some effect on malaria parasites. However this benefit has also been found in people who have bacteria resistant to cotrimoxazole. One possible mechanism could be that cotrimoxazole reduces the number of bacteria in the intestines or limits the damage to the intestinal mucosa. In HIV infected persons the intestinal barrier is damaged and there is passage of bacterial products from the intestine into the body. This “microbial translocation” is thought to cause general activation of the immune system, which worsens the outcome of HIV infection. So in this study we are comparing patients who keep taking cotrimoxazole or stop, for evidence of bacterial products crossing from the intestines into the blood, damage to the intestinal wall and immune activation. Understanding how cotrimoxazole exerts its effect will help us to improve the care of HIV infected persons.

Technical Summary

HIV infection is associated with translocation of microbial products from the gut and general immune activation. These are correlated with progression to AIDS in US settings, but the correlation is not clear in African populations, possibly as a consequence of higher environmental exposure to intestinal infections.
In the main COSTOP non inferiority trial (ISRCTN44723643), HIV infected adults on ART and CTX prophylaxis were randomized into two groups, one continuing CTX prophylaxis and one receiving placebo. The trial concluded that CTX+ART reduces morbidity due to malaria and bacterial infections, compared to ART alone. We hypothesized that CTX might lower the bacterial burden in the GI tract, reducing microbial translocation and consequent immune activation and proposed to test the hypothesis that discontinuation of CTX prophylaxis could be associated with development of differences in microbial translocation and immune activation as well as clinical outcome.
Two groups of 80 HIV infected individuals on ART and CTX or placebo were enrolled and bled at baseline, 3, 6 and 12 months.
Plasma was collected to be tested for soluble markers of microbial translocation (plasma LPS and bacterial 16S rDNA), immune control of LPS (EndoCab, sCD14), enterocyte damage (I-FABP), inflammation (CRP) and activation (IL-6, TNF-a) using ELISA assays, LAL assays or PCR; an aliquot of whole blood was used to evaluate immune activation by flow cytometry as the proportion of CD8 T cells expressing HLA-DR and CD38. These tests are ongoing.
The primary statistical analysis will compare the mean LPS at 12 months between the two groups using an analysis of covariance model, with terms for acute malaria/parasitaemia and study arm. Other continuous markers of microbial translocation and immune activation will be analysed similarly. Linear mixed model for longitudinal data will be fitted for each continuous response using values at baseline, 3, 6 and 12 months to compare slopes over time in the two groups. Cox regression models will be fitted in exploratory analyses of the time to first event (CTX preventable infection or death) to see if the hazard of an event occurring correlates with LPS levels.
This research will increase our understanding of the effects and mechanism of CTX prophylaxis in HIV treatment of populations in regions with high burden of malaria and bacterial infections.
Title COSTOP Immunology dataset 
Description A dataset of inflammatory, bacterial translocation and T cell activation markers measured at baseline and 3, 6 and 12 months after cotrimoxazole discontinuation in HIV infected Ugandans stabilised on ART and cotrimoxazole. N=80 controls (continuing cotrimoxazole) and N=80 discontinuing. The markers are: CD8+ CD38+ HLA-DR+ frequency in peripheral blood; LPS (plasma); sCD14 (plasma); EndoCAb (antibodies anti LPS-core, in plasma); IL-6 (plasma); TNF-a (plasma); CRP (plasma); I-FABP (plasma). An interim analysis was performed in March 2017 with about 86% of LPS tests results available, for the purpose of reporting to Researchfish. A final analysis will be done after completion of LPS tests in May 2017. The main differences between placebo and CTX groups were seen in plasma levels of CRP (p=0.03), TNF-a (p=0.06), IL-6 (p=0.65) and I-FABP (p=0.24), which were all higher in the placebo group and in endotoxin core antibodies (p=0.12), which were lower in the placebo group. Placebo and CTX groups were not statistically significantly different in plasma levels of LPS (p=0.76), and sCD14 (p=0.60) and in CD8 T cell activation (p=0.74) although the placebo group had higher CD8+ CD38+ HLADR+ cell frequencies at months 6 and 12. Adjustment for malaria status and bacterial infections did not affect these comparisons. This interim analysis suggests a modest effect of cotrimoxazole cessation on markers of inflammation, immune activation and bacterial translocation in HIV infected individuals who had been stabilised on ART and cotrimoxazole. 
Type Of Material Database/Collection of data 
Provided To Others? No  
Impact None yet - the dataset is still being added to and unpublished.