Advanced therapies for infantile malignant pertussis: A systematic review of hydroxyurea, leukapheresis, exchange transfusion, and extracorporeal membrane oxygenation

Objectives: The objectives of the study are to evaluate the effectiveness, safety, and survival outcomes of advanced interventions, such as hydroxyurea, leukapheresis, exchange transfusion (ET), and extracorporeal membrane oxygenation (ECMO), in infants with malignant pertussis. Methods: A Preferred Reporting Items for Systematic Reviews and Meta-Analyses-2020-compliant systematic review of PubMed and Europe PMC (January 2000–June 2025) identified studies of infants treated with these interventions. The risk of bias was assessed using risk of bias in non-randomized studies of interventions (ROBINS-I), Joanna Briggs Institute (JBI), and a measurement tool to assess systematic reviews 2 (AMSTAR-2) tools, and data were synthesized descriptively. Results: Among 830 records screened, 46 studies met the inclusion criteria. Hydroxyurea, typically administered at 20 mg/kg/day, was associated with gradual leukoreduction within 5–7 days and reported survival rates ranging from 70% to 90% across observational cohorts, comparable to those observed with ET and with fewer reported procedural complications. ET was consistently associated with rapid leukoreduction and short-term improvement in oxygenation, particularly when performed early. Leukapheresis demonstrated inconsistent benefit and was infrequently reported. ECMO alone was associated with lower survival rates; however, when used in conjunction with prior leukoreductive strategies, reported survival exceeded 70% in selected cohorts. Conclusion: Observational evidence suggests that early leukoreductive strategies may be associated with improved outcomes in infants with malignant pertussis, particularly when implemented before cardiovascular deterioration. ET and hydroxyurea were the most consistently reported interventions, while ECMO appeared to provide benefit primarily as adjunctive support following leukoreduction. Given the heterogeneity and non-randomized nature of the available data, these findings should be interpreted cautiously and considered hypothesis-generating.