What is the Significance of a High Cycle Threshold Positive IS481 PCR for Bordetella pertussis?

Abstract

To the Editors: We read with great interest the article by Waters et al regarding an outbreak of atypical pertussis diagnosed by polymerase chain reaction (PCR).1 It is well established that PCR targeting the IS481 gene, of which there are 50 to 200 copies per bacterial cell, is a highly sensitive method for detecting Bordetella pertussis.2 Indeed, with this assay, a positive result with a cycle threshold (CT) value >35 likely represents the detection of less than 1 organism per sample.2,3 Consequently, it remains unclear how to interpret a late-cycle positive PCR result. Does it represent detection of true pertussis disease (possibly modified by vaccination, partial natural immunity or antimicrobial therapy)? Or, does it represent transient colonization where B. pertussis was an innocent bystander with regards to the clinical syndrome that prompted diagnostic testing? Given that Walters et al used a CT of ≤40 to define a case, and given that their mean CT value was 38.4, further analysis is necessary to understand the clinical relevance of their results and to shed light on the dilemma of interpreting CT values >35. To do so, it would be invaluable to look for the existence of an association between higher CT values and the presence of coinfection with other respiratory pathogens (as tested by PCR on the study's pertussis-positive nasopharyngeal samples). Furthermore, a receiver operating characteristic curve would help elucidate how CT cut-off values affect the diagnostic assay's sensitivity and specificity when compared with a reference standard such as the clinical case definition of pertussis. Jesse Papenburg, MD, FRCPC Infectious Diseases Division, Department of Pediatrics and Department of Microbiology Montreal Children's Hospital, McGill University Health Centre Montreal, Canada Infectious Disease Research Centre Centre Hospitalier Université Laval Quebec City, Canada Patricia Fontela, MD, MSc Department of Epidemiology and Biostatistics McGill University Montreal, Canada