Association between pertactin-producing Bordetella pertussis and fulminant pertussis in infants: a multicentre study in France, 2008-2019.

Abstract

Objectives: Virulence factors of the causative agent, Bordetella pertussis, may be involved in fulminant pertussis, the most severe form of whooping cough (pertussis) in infants. We aimed to assess the association between fulminant pertussis and the status of pertactin (PRN) production of B. pertussis clinical isolates.

Methods: Symptomatic infants aged <6 months with a positive B. pertussis culture from 2008-2019 were included. B. pertussis isolates and clinical data were collected from French hospital laboratories through the national pertussis surveillance network. Fulminant pertussis was defined as a case with a leukocyte count >40 × 10⁹/L and at least one of the following criteria: respiratory failure, pulmonary hypertension, shock, or multiple organ failure. PRN production was assessed by western blotting. Baseline characteristics of infants and microbiological findings were compared between patients with and without fulminant pertussis. To identify patient and microbiological features associated with fulminant pertussis, a multivariable modified Poisson regression model was developed with confounders selected using a directed acyclic graph.

Results: We included 361 infants with pertussis (median age 63 days [interquartile range, 39-86]), of whom 32 (9%) progressed to fulminant pertussis. None of the mothers was vaccinated during pregnancy. Of the 361 implicated B. pertussis isolates, 294 (81%) produced PRN. Patients with fulminant pertussis were more often neonates (adjusted relative risk [aRR]: 3.62, 95% confidence interval [CI]: 1.76-7.44), infants with a history of prematurity (aRR: 7.08, 95% CI: 3.06-16.36), unvaccinated infants (aRR: 4.42, 95% CI: 1.02-19.24), and infants infected by PRN-producing isolates (aRR: 3.76, 95% CI: 1.02-13.83).

Discussion: PRN-producing B. pertussis was independently associated with an increased risk of fulminant pertussis. In a context where PRN-containing acellular pertussis vaccines favour the emergence of PRN-deficient isolates, our study suggests a positive role for such vaccines in driving the evolution of B. pertussis populations towards reduced virulence.