Diagnostic performance of two domains derived from a novel Trypanosoma cruzi protein to chronic Chagas disease in endemic areas of Brazil

Abstract

Chronic Chagas disease (CCD), a potentially life-threatening illness, is an infection caused by Trypanosoma cruzi parasite. In thisstage, the WHO advises using at least two distinct IgG-basedtests for a reliable diagnosis, due to the lack of a “gold-standard”laboratory technique. Recently, we demonstrated the diagnosisperformance of two domains (D3 and D6) derived from a hypothetical T. cruzi protein, named Tc323, which has no homologsin T. brucei or Leishmania genomes. For this, we used plasmaof patients with CCD and non-infected donors from Argentina,Bolivia, and Paraguay. Here, we extended our work by testingplasma of individuals with CCD or infected with unrelated diseases living in endemic regions of Brazil. After the optimizationof our indirect ELISA, recombinant D3 and D6 proteins were assessed by employing plasma of 405 T. cruzi-positive and 530T. cruzi-negative individuals from distinct areas of Brazil. Additionally, 748 samples from donors with unrelated diseases wereincluded to evaluate cross-reactivity. The receiver operatingcharacteristic (ROC) curve was used to define the cut-off valuewith the best diagnostic performance for each ELISA assay using pooled plasma samples from positive and negative donors.The results were expressed as reactivity index (RI), calculatedas the ratio between the optical density (OD) of the samples andthe OD of the cut-off, where RI >1.00 were considered positive.To evaluate the overall precision for each antigen, areas underthe ROC curve (AUC) were performed. The AUC values were92.89% and 92.11% for D3 and D6, respectively. Both domainsexhibited 97.92% specificity while D3 was 79.51% sensitive andD6 81.48%. The accuracy of D6 was 90.8% compared to the lower value for D3 (89.95%). Assuming an inconclusive range of RIvalues within 1.0 ± 0.10, 13 (3.95%), and 16 (4.86%) true positivesamples tested against D3 and D6, respectively, led to false results. By analyzing the true negative samples, 4 (0.75%) fell inside the inconclusive space using D3 and only 1 (0.18%) usingD6. The incidence of cross-reactivity was almost negligible: only4 out of 748 plasmas displayed seroreactivity using D3, while 5samples generated false positives using D6. In both cases, samples came from patients infected with Leptospira or Leishmania.However, no reactivity was detected with plasma from individuals with dengue, filariasis, HBV, HCV, HIV, rubella, schistosomiasis, syphilis, COVID, leprosy and toxoplasmosis.Our results demonstrated a noteworthy performance of D3 andD6 for the diagnosis of CCD individuals from endemic areas ofBrazil. Even more, both domains showed insignificant cross-reactivity with non-related infectious diseases.