Abstract
Tuberculosis (TB) remains a significant, yet under-recognized cause of death in the pediatric population, with a WHO estimate of 1 million new cases of childhood TB in 2016 resulting in 250,000 deaths. Diagnosis is notoriously difficult; manifestations are protean due to the high proportion of cases of extra-pulmonary TB in children, and logistical problems exist in obtaining suitable specimens. These issues are compounded by the paucibacillary nature of disease with the result that an estimated 96% of pediatric TB-associated mortality occurs prior to commencing anti-tuberculous treatment. Further development of sensitive, rapid diagnostic tests and their incorporation into diagnostic algorithms is vital in this population, and central to the WHO End-TB strategy. Initial gains were made with the expansion of nucleic acid amplification technology, particularly the introduction of the GeneXpert fully-automated PCR Xpert MTB/Rif assay in 2010, and more recently, the Xpert MTB/Rif Ultra (Ultra) assay in 2017. Ultra provides increased analytical sensitivity when compared with the initial Xpert assay in vitro; a finding now also supported by six clinical studies to date, two of which included pediatric samples. Here, we review the published evidence for the performance of Ultra in TB diagnosis in children, as well as studies in adults with paucibacillary disease providing results relevant to the pediatric population. Following on from this, we speculate upon future directions for Ultra, with focus on its potential use with alternative diagnostic specimens, which may be of particular utility in children.
Highlights
Tuberculosis (TB) is currently the leading cause of mortality worldwide from a single infectious agent, being responsible for an estimated 1.7 million deaths in 2016 (1)
When compared to the 19% improvement in Xpert sensitivity, these results suggest a greater additional diagnostic benefit for Ultra in the HIV-positive population, a result mirrored in the adult population (13)
We have reviewed the existing evidence regarding the accuracy of Ultra, with particular focus on the use of Ultra in children
Summary
Tuberculosis (TB) is currently the leading cause of mortality worldwide from a single infectious agent, being responsible for an estimated 1.7 million deaths in 2016 (1). More than 96% of TB-related deaths are estimated to occur in children not receiving anti-tuberculous treatment, highlighting significant challenges in diagnosis. The often non-specific nature of TB presentation in children has led to the development of a multitude of “scoring systems” based on clinical assessment and basic investigations, with most aimed at the diagnosis of pulmonary TB (2). Due to the poor specificity of diagnostic algorithms, estimates have shown that TB may be both over-diagnosed and over-treated in some settings (4); yet underdiagnosed in other settings. Mycobacteriological diagnostics used in adults remain the “gold standard” but demonstrate a lower sensitivity in children (5), both from the paucibacillary nature of TB in children and the problem obtaining adequate respiratory or non-respiratory specimens for bacteriological confirmation (as young children are frequently unable to voluntarily expectorate sputum) (6)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
