Abstract
Introduction: Early clearance (EC) is the eradication of M. tuberculosis infection by innate immunity, before an adaptive immune response develops. Case contact studies consistently report a group of highly exposed but uninfected contacts that could be early clearers, and genetic loci are associated with the trait. Understanding how early clearers are protected could lead to vaccines or other therapies to prevent M. tuberculosis infection.
Methods: In Indonesia, contacts of smear and x-ray positive Tuberculosis (TB) cases were recruited within 2 weeks of the case starting treatment. A Quantiferon Gold In Tube (QFN-GIT) test was performed at baseline, and contacts with a positive test were classified as "baseline positive" (BP), and those with a negative test received a repeat QFN-GIT 14 weeks later. Contacts who continued to test QFN-GIT negative, were classed as "persistently negative" (PN) and those who tested positive were classified as "converters". Cell populations, cytokine responses to stimuli, serum 25-OH vitamin D and the expression of candidate innate immune genes were measured in samples collected at baseline. The primary comparison was of the PN group, putative early clearers, to converters. Measurements collected from the whole cohort (peripheral blood cell counts, 25-OH vitamin D) were modelled for their association with risk of conversion in multi-level models that adjusted for clustering and confounding variables. Sub-studies on cytokine responses, gene expression and M. tuberculosis lineage were performed in subsets of participants.
Results: A total of 941 case contacts of 317 index cases were recruited. One hundred and sixty-two (51%) of index cases had lung cavities on x-ray. Of the contacts 544 (57.9%) were BP, 223 (23.6%) were PN, 76 (8.0%) were converters, 18 had indeterminate QFN-GIT results and 22 (2.3%) had TB disease. A further 27 (2.9%) had unevaluated symptoms and 31 (3.1%) were lost to follow up. Risk of conversion was higher for contacts exposed to a 3+ smear grade case than those exposed to a grade 1+ or scanty case (RR=2.63; 95% CI 1.54 - 4.50; P<0.0001), and it was also higher for current smokers versus non smokers (RR=1.63; 95% CI 1.03 - 2.57; P=0.04). Risk of conversion was lower in those who been vaccinated with Bacillus Calmette-Guérin (BCG) compared to those who had not (RR=0.52; 95% CI 0.35 - 0.77; P=0.001). Risk of conversion increased with each additional mg/kg2 increase in body mass index (RR=1.06; 95% CI 1.01 - 1.10; P=0.01), and each 1000 cells per microliter decrease in immature granulocyte count (RR=0.72; 95% CI 0.60 - 0.90; P=0.004) and a neutrophil count in the lowest quartile, relative to the second quartile (RR 1.91 95%CI 1.05 - 3.50; P=0.04). There was no association with risk of conversion and vitamin D insufficiency (RR=1.10; 95% CI 0.65 - 1.85; P=0.7) or deficiency (RR=1.12; 95% CI 0.64 - 1.98; P=0.7).
Cytokine responses to whole blood stimulation were analysed for 97 PN contacts and 41 converters. Responses to M. tuberculosis stimulation were not significantly higher for the primary comparison, but the IL-6 response to Streptococcus (RR=1.41; 95% CI 1.09 - 1.84; P=0.01) and the IL-8 responses to E. coli (RR=1.38; 95% CI 1.04 - 1.84; P=0.03) were higher in PN than converters. Restricting the analysis to 111 BCG vaccinated contacts (81 PN, 30 converters) revealed PN had higher IL-1β and IL-6 responses to Mtb lysate (RR=1.43; 95% CI 1.00 - 2.03; P=0.047) and (RR=1.40 95% CI 1.04 - 1.88; P=0.027) respectively and higher IL-6 response to S. pneumoniae (RR=1.52; 95% CI 1.14 - 2.03; P=0.005) and E. coli (RR: 1.26; 95% CI 1.00 - 1.59; P=0.054).
Out of 45 candidate genes, the expression of nine genes differed at the P<0.05 threshold and of these only one (CTSD) was higher in PN than converters.
Data on the M. tuberculosis lineage of the index case was available for 271 contacts. In this sub-study exposure to an index case with a lineage two isolate was cultured was associated with a positive QFN-GIT at baseline or follow up (RR=1.22; 95% CI 1.02 - 1.46; P=0.03), relative to exposure to M. tuberculosis of another lineage.
Conclusions: Higher concentrations of immature granulocytes, neutrophils and pro-inflammatory innate cytokines in the PN group reveal early clearers have a pro-inflammatory phenotype. ECwas associated with BCG vaccination, and a higher innate cytokine responses in vaccinated contacts. This findings suggest BCG mediated protection occurs by trained innate immunity. Lineage two M. tuberculosis could evade EC more successfully than other lineages.