This article is more than 5 years old. Information may no longer be current.

Blood-based signature predicts tuberculosis in healthy patients

Researchers in this study have developed a blood-based signature from following individuals with Mycobacterium tuberculosis that can predict active tuberculosis in healthy individuals, according to recent research.

“Our results, showing that blood-based signatures in healthy individuals can predict progression to active tuberculosis disease, pave the way for the establishment of diagnostic methods that are scalable and inexpensive,” Daniel E Zak, PhD, from the Center for Infectious Disease Research in Seattle, and colleagues wrote in their study. “An important first step would be to test whether the signature can predict tuberculosis disease in the general population, rather than the select populations included in this project; for example, the risk of tuberculosis disease in our populations was much higher than the lifetime risk of 10% encountered in the general population. The newly described signature holds potential for highly targeted preventive therapy, and therefore for interrupting the worldwide epidemic.”

Zak and colleagues followed a cohort of 6,363 South African adolescents between the ages of 12 years and 18 years who had been infected with M. tuberculosis for a minimum of 2 years and collected blood samples for a period of 6 months to follow the disease’s progression, according to the abstract. Using the blood samples, the researchers analyzed RNA sequencing data and created a prospective signature of risk that they then used to evaluate 4,466 South African and Gambian participants.

They found that the 16-gene signature of risk had a sensitivity of 66.1% (95% CI, 63.2% 68.9%) and a specificity of 80.6% (95% CI, 79.2%-82.0%) in predicting tuberculosis progression 12 months prior to diagnosis, according to the abstract. In the South African and Gambian adolescents, the risk signature was validated (P = 0.018, RNA sequencing; P = 0.0095, quantitative real-time PCR) and had a sensitivity of 53.7% (95% CI, 42.6%-64.3%) and a specificity of 82.8% (95% CI, 76.7%-86.0%). – by Jeff Craven

Disclosure: The researchers report no relevant financial disclosures.