Abstract
BACKGROUND: The performance of serological tests for Trypanosoma cruzi diagnosis in Mexico has not included discordant control sera nor has it evaluated the role of immune response specificities, patient infection history or clinical status. OBJECTIVES: The performance of commercial serological and molecular diagnostic tests and diagnostic algorithms was analysed in Mixtecan and Zapotecan ethnic populations having recent and long-term infection history. METHODS: An amplified global gold standard for T. cruzi infection included serological (≥ 2 conventional tests positive) and molecular (sequence identity of any of five genes using end point polymerase chain reaction (epPCR) or any positive using quantitative polymerase chain reaction (qPCR) diagnostic test results. FINDINGS: Only 81% of previously diagnosed untreated infections were reconfirmed using serology, while an additional 14% only using PCR. Serological diagnosis sensitivity (≥ 2 tests positive) in the primary diagnosis cohort was 8%, while specificity was 16%. Diagnosis sensitivity was similar using epPCR and qPCR only in primary diagnoses and all identified using the satellite (SAT) gene. The 18S ribosomal DNA identified T. cruzi and T. dionisii co-infections from Pacific coast sites. MAIN CONCLUSIONS: The current study provides evidence for inadequate diagnostic performance of conventional serological tests and the need to develop appropriate antigenic tools and use molecular testing of seronegatives to ascertain absence of infection.