Epidemiology: Open Access
Open Access

Abstract

Introduction: In previous study we have reported the development of a serological test using Pichia pastoris recombinant
CFP32 antigen, a virulence factor of Mycobacterium tuberculosis (MTB), with both high sensitivity and specificity in the
Tunisian population. In this study, we aimed to assess the ability of the CFP32 serological test, to discriminate between TB
patients and non-TB among people of different ethnic origins.
Materials & Methods: Serum samples were obtained from 224 donors from different countries, divided into two groups: i.
samples from Tunisian cohort (intermediate endemicity) including TB patients and their contacts ii. Samples from WHO
banked sera (obtained from sub-Saharan African and Latin American countries with high endemicity) including TB patients
and TB suspected individuals. Receiver operating curve (ROC) was used to evaluate the performance of the CFP32 serological
test.
Results: CFP32 specific antibody levels in the TB patients was significantly higher than non-TB controls in Tunisian group
(p<10-4), Sub-Saharan African group (p<10-3) and Latin American group (p<10-4). Considering all groups together, the
area of the ROC curve was 0.77 (95%CI, 0.70-0.83). The cutoff level was set at 0.63, giving the optimal combination between
sensitivity (70%, CI95%: 60-79) and specificity (73%, CI95%: 63-81).
Conclusion: Although CFP32 single antigen serological test did not achieve excellent sensitivity and specificity, a combination
of CFP32 with other select antigens and particularly stage-specific infection antigens could be used to improve performance of
a serological test regardless of TB incidence and latent infection.

Biography

Jihene bettaieb is working in Laboratory of Medical Epidemiology, Pasteur Institute of Tunis, Tunisia.