医学
肺结核
犬尿氨酸
内科学
结核分枝杆菌
疾病
胃肠病学
免疫学
病理
色氨酸
生物
生物化学
氨基酸
作者
Jeffrey M. Collins,Kidist Bobosha,Nandakumar S. Narayanan,Neel Gandhi,Cheryl L. Day,Jyothi Rengarajan,Russell R. Kempker,Max S. Y. Lau,Mary M. Nellis,Nestani Tukvadze,N. Sarita Shah,James C. M. Brust,Azhar Nizam,Nazir Ismail,Keith D. Kauffman,Shunsuke Sakai,Dean P. Jones,Daniel L. Barber,Thomas R. Ziegler,J. Ernst
标识
DOI:10.1093/infdis/jiaf240
摘要
Abstract Background High-resolution metabolomics has shown promise for identifying blood-based biomarkers of tuberculosis (TB). We sought to discover a metabolic signature to detect pulmonary TB disease and monitor treatment response. Methods Plasma from Ethiopian persons with pulmonary TB at diagnosis (n=82) was compared to household contacts with TB symptoms (n=104) and 2, 6, and 12 months after treatment initiation. Participants were divided into training and test sets for model building, with additional validation using independent cohorts from the countries of Georgia (n=89) and South Africa (n=85). Signatures were further evaluated in non-human primates infected with M. tuberculosis (Mtb). Results Among the metabolites that most significantly differed in concentration, tryptophan and retinol were significantly decreased in persons with TB disease (45.2 uM vs 62.5 uM and 4.1 uM vs 8.2 uM respectively), while kynurenine was significantly increased (2.1 uM vs 1.6 uM; q<0.0001 for all). A signature that included the kynurenine/tryptophan ratio and retinol showed excellent classification for TB disease (AUC=0.97). The signature had an AUC of 0.97 in HIV+ and 0.95 in HIV- persons with TB disease from South Africa and 0.93 in TB patients from Georgia. In Ethiopian participants, signature scores decreased after 2 (0.85 to 0.42) and 6 months of TB treatment (0.42 to 0.18; p<0.0001 for both) to similar levels as controls. Plasma retinol also declined in NHPs infected with Mtb 15-16 weeks after infection (5.9 uM vs 3.6 uM; p<0.001). Conclusions The plasma Kyn/Trp ratio and retinol represents a promising metabolic signature that could advance TB diagnostics.
科研通智能强力驱动
Strongly Powered by AbleSci AI