Immune Mediators as Plasma Biomarkers for Identifying Household Contacts and Classifying Clinical Forms and Leprosy Reactions

Marcio Sobreira Silva Araujo ^1* Jairo Carvalho ^1,2 Marcelo Antônio Pascoal-Xavier ^1,3 Marcelo Grossi Araújo ³ Júlia Pereira Martins ¹ Andréa Teixeira-Carvalho ¹ Matheus de Souza Gomes ⁴ Laurence Rodrigues do Amaral ⁴ Vanessa Peruhype Pascoal ¹ Jordana Grazziela Coelho-dos-Reis ⁵ Olindo Assis Martins Filho ¹

• ¹ René Rachou Institute, Oswaldo Cruz Foundation (Fiocruz), Belo Horizonte, Brazil
• ² Fundação Hospitalar do Estado de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil
• ³ Faculdade de Medicina, Universidade Federal de Minas Gerais, Belo Horizonte, Mato Grosso, Brazil
• ⁴ Federal University of Uberlandia, Uberlândia, Minas Gerais, Brazil
• ⁵ Federal University of Minas Gerais, Belo Horizonte, Minas Gerais, Brazil

The present study intended to evaluate the performance of plasma immune mediators to classify Leprosy Patients [L(PB); L(MB)], Leprosy Reactions Patients [T1LR; T2LR], Household Contacts [HHC] and Non-Infected Controls [NI]. Quantitative measurements of immune mediators were carried out by high throughput microbead multiplex array. The results demonstrated that most plasma immune mediators were increased in all clinical groups as compared to NI. Higher frequencies but lower maximum magnitude of increase (fold change according to NI) was observed for T1LR-(63%; 6.1x) and T2LR-(63%; 9.7x) as compared to HHC-(48%; 68.5x), L(PB)-(56%; 8.5x) and L(MB)-(48%; 37.9x). Bi-dimensional scattering profile (magnitude order vs significance) identified a higher number of immune mediators in T2LR-(12/27) in comparison to HHC-(8/27), L(PB)-(7/27), L(MB)-(5/27) and T1LR-(5/27). CXCL8 was selected as a parameter with the highest accuracy and significance (AUC=0.98; p=0.0002) to classify NI vs HHC. CCL3 was the single analyte with moderate accuracy and significance (AUC=0.74; p=0.0422) to classify L(PB) vs L(MB). IL-9 was selected as an attribute with moderate accuracy and significance (AUC=0.77; p=0.0041) to classify T1LR vs T2LR. Decision tree algorithms confirmed the high accuracy (96%) of CXCL8 to classify NI vs HHC. The use of CCL3 followed by IFN- classified L(MB) vs L(PB) with high accuracy (93%). Moreover, the analysis of IL-9 followed by IL-6 and CXCL10 classified T1RL vs T2RL with high accuracy (96%). In general, combined stepwise algorithms showed enhanced classification accuracy as compared to single attribute analysis. Together, our findings supported the potential use of plasma immune mediators as complementary laboratorial biomarkers to identify Household Contacts and classify distinct clinical forms of Leprosy and Leprosy Reactions.