DIAGNOSTIC MICROBIOLOGY AND INFECTIOUS DISEASE. 2019:114937.

Brock M, Hanlon D, Zhao M and Pollock NR

Diagn Microbiol Infect Dis. 2019 Nov 13:114937. doi: 10.1016/j.diagmicrobio.2019.114937.

Abstract

Urinary detection of Mycobacterium tuberculosis lipoarabinomannan (LAM) for tuberculosis (TB) diagnosis is well characterized, but the utility of serum LAM detection remains unclear. We developed an assay for serum LAM detection using single-molecule array (Simoa), purified M. tuberculosis LAM, and anti-LAM monoclonal antibodies and evaluated performance on diluted/heat-treated serum samples from patients with and without active TB and/or HIV. The Simoa assay had a limit of detection of 0.35 pg/mL and lower limit of quantification of 0.942 pg/mL. Corrected serum LAM concentrations ranged from 0 to 132.0 pg/mL [median 1.71, interquartile range (IQR) 0.94-6.80] in 90 TB+ patients and from 0 to 2.29 pg/mL (median 1.03, IQR 0.47-1.69) in 55 TB- patients. Using a cutoff of 2.3 pg/mL for 100% specificity, assay sensitivity was 37% in all TB+ subjects (33/90; 95% CI 0.27-0.48), 47% in TB+/HIV+ subjects (26/55; 0.34-0.61), and 60% in TB+/HIV+/smear+ subjects (21/35; 0.42-0.76). Mycobacterial LAM is detectable in serum with high specificity and reasonable sensitivity using Simoa.