Transcriptomic Biomarkers for Tuberculosis: Evaluation of DOCK9. EPHA4, and NPC2 mRNA Expression in Peripheral Blood

Abstract

Lately, much effort has been made to find mRNA biomarkers for tuberculosis (TB) disease/infection with microarray-based approaches. In a pilot investigation, through RNA sequencing technology, we observed a prominent modulation of DOCK9, EPHA4, and NPC2 mRNA abundance in the blood of TB patients. To corroborate these findings, independent validations were performed in cohorts from different areas. Gene expression levels in blood were evaluated by quantitative real-time PCR (Brazil, n = 129) or reanalysis of public microarray data (UK: n = 96; South Africa: n = 51; Germany: n = 26; and UK/France: n = 63). In the Brazilian cohort, significant modulation of all target-genes was observed comparing TB vs. healthy recent close TB contacts (rCt). With a 92% specificity, NPC2 mRNA high expression (NPC2high) showed the highest sensitivity (85%, 95% CI 65%–96%; area under the ROC curve [AUROC] = 0.88), followed by EPHA4 (53%, 95% CI 33%–73%, AUROC = 0.73) and DOCK9 (19%, 95% CI 7%–40%; AUROC = 0.66). All the other reanalyzed cohorts corroborated the potential of NPC2high as a biomarker for TB (sensitivity: 82–100%; specificity: 94–97%). An NPC2high profile was also observed in 60% (29/48) of the tuberculin skin test positive rCt, and additional follow-up evaluation revealed changes in the expression levels of NPC2 during the different stages of Mycobacterium tuberculosis infection, suggesting that further studies are needed to evaluate modulation of this gene during latent TB and/or progression to active disease. Considering its high specificity, our data indicate, for the first time, that NPC2high might serve as an accurate single-gene biomarker for TB.

Performance of diagnostic and predictive host blood transcriptomic signatures for Tuberculosis disease: A systematic review and meta-analysis

“…Eight [ 23 – 25 , 38 , 44 , 45 , 49 , 56 , 59 , 60 ] of the 18 studies evaluated signatures in populations with ODs. Only one study [ 43 ] made reference to blinding of the index test readers. Bias in patient selection arose in most studies due to case-control design and non-reporting of sampling method, resulting in a non-representative spectrum of patients.…”

Concise whole blood transcriptional signatures for incipient tuberculosis: A systematic review and patient-level pooled meta-analysis

“…The differences in primary intended applications, which are reflected in the study populations used for biomarker discovery, may account for their inferior performance when evaluated solely for identification of incipient TB in a predominantly healthy, HIVnegative adult population. The signatures with inferior performance also included three discovered from panels of pre-selected candidate genes, rather than a genome-wide approach 15,28,33 , and four with only 6-24 TB cases in the discovery sets [32][33][34]38 . This suggests that using a genome-wide approach and including adequate numbers of diseased cases are important considerations during signature discovery.…”

“…The size of the nodes is proportional to the number of 26 downstream biomarkers associated with each regulator and the thickness of the edges is proportional to the -log10 P value for enrichment of each of the upstream 27 regulators. 28 38 space, stratified by months to disease. 39 Dashed lines represent positive predictive value planes, based on 2% pre-test probability.…”

Low Prevalence of Latent Tuberculosis Infection among Contacts of Smear-Positive Adults in Brazil

“…1 A study conducted by Brazilian authors with participants aged > 18 years identified progression of 25.0% during the 1-year followup period. 23 Considering that children represent a risk group for TB progression, 7 it is expected that the rate of progression in this group would be even higher among our participants.…”