119. Prospective validation of a three-gene signature for tuberculosis diagnosis, predicting progression and evaluating treatment response
Session: Oral Abstract Session: Tuberculosis and other Mycobacterial Infections
Thursday, October 4, 2018: 8:45 AM
Room: W 2002

Background: The World Health Organization (WHO) has identified the need for a non-sputum based triage test for TB that can be used to identify those who need further testing to identify active disease. We investigated whether our previously described 3-gene TB score could identify individuals with active tuberculosis (ATB) prior to seeking care (Òactive case detectionÓ) and how the 3-gene TB score correlated with the timing of disease onset, disease severity, and response to treatment.

Methods: This study consisted of a prospective nested case-control trial, Brazil Active Screening Study (BASS; 2016), and re-analysis of data from two prospective cohort studies, the Adolescent Cohort Study (ACS; 2005-2007) and the Catalysis Treatment Response Cohort (CTRC; 2010-2013). The BASS case-control sub-cohort contained 81 adults (ages 20-72 years, 33 ATB, 48 controls). The ACS contained 153 adolescents (ages 12-18 years, 46 ATB, 107 LTBI). The CTRC contained 138 adults (ages 17-67 years, 100 ATB, 17 other lung disease patients, 21 healthy controls).

Results: The 3-gene TB score diagnosed ATB patients with high accuracy: BASS cohort AUC=0.87 (95%CI=0.82-0.91, Fig 1A), ACS cohort AUC=0.86 (95%CI=0.76-0.97, Fig 1B), and CTRC AUC=0.93 (95%CI=0.88_0.97). In the ACS, the 3-gene TB score predicted progression from LTBI to ATB 6 months prior to positive sputum test (AUC=0.86; 95%CI=0.79-0.92, Fig 1B).  In the CTRC, the 3-gene TB score correlated with glycolytic activity ratio of PET-CT at baseline (correlation=0.54, p=3.98e-08, Fig 1C) and at end of treatment (correlation=-0.408, p=3.72e-05). In the CTRC, the 3-gene TB score at baseline predicted the likelihood of prolonged sputum positivity following treatment initiation and treatment response at 6 months (p=3.60e-05). Collectively, across all cohorts, the 3-gene TB score identified ATB patients with 90% sensitivity and 70% specificity, and had 99% negative predictive value (NPV) at 5% prevalence.

Conclusion: Across three independent prospective cohorts, the 3-gene TB score closely approaches the WHO target product profile benchmarks for non-sputum-based triage test at high NPV. These performance characteristics make it a potential test for ruling out ATB and for monitoring disease status.

Hayley Warsinske, PhD1, Aditya Rao, BA2, Flora Martinez Figueira Moreira, BS3, Paulo Cesar Pereira Dos Santos, BS3, Andrew Liu, MS2, Madeleine Scott, BA2, Stephanus Malherbe, MD, PhD4, Katharina Ronacher, PhD5, Gerhard Walzl, PhD4, Jill Winter, PhD6, Timothy E. Sweeney, MD, PhD7, Julio Croda, MD PhD8, Jason R. Andrews, MD9 and Purvesh Khatri, PhD10, (1)Immunology, BMI, Stanford University, Stanford, CA, (2)Stanford University, Stanford, CA, (3)Federal University of Grande Dourados, Dourados, Brazil, (4)Stellenbosch University, Stellenbosch, South Africa, (5)The University of Queensland, Brisband, Australia, (6)Catalysis Foundation for Health, Emeryville, CA, (7)Institute for Immunity, Transplantation, and Infections and Division of Biomedical Informatics, Department of Medicine, Stanford University, Stanford, CA, (8)Faculty of Medicine, Federal University of Grande Dourados; Oswaldo Cruz Foundation, Dourados, Brazil, (9)Division of Infectious Diseases and Geographic Medicine, Stanford University School of Medicine, Stanford, CA, (10)2- Institute for Immunity, Transplantion and Infections, 3- and Biomedical Informatics Research, Dept. of Medicine, Stanford University, Palo Alto, CA


H. Warsinske, None

A. Rao, None

F. Martinez Figueira Moreira, None

P. Cesar Pereira Dos Santos, None

A. Liu, None

M. Scott, None

S. Malherbe, None

K. Ronacher, None

G. Walzl, None

J. Winter, None

T. E. Sweeney, Inflammatix, Inc.: Employee and Shareholder , Salary .

J. Croda, None

J. R. Andrews, None

P. Khatri, Inflammatix Inc: Board Member , Equity .

Previous Abstract | Next Abstract >>

Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 3rd with the exception of research findings presented at the IDWeek press conferences.