Pertussis in the first 3 months of life is frequently severe and potentially fatal because of hyperleukocytosis, pulmonary hypertension and refractory hypoxemia,1 with an estimated 195,000 deaths worldwide among children younger than 5 years in 2013.2 Aggressive management strategies that include exchange transfusion, leukofiltration and extracorporeal membrane oxygenation (ECMO) could be useful in neonates with severe pertussis. However, there is a high risk of mortality in infants managed by ECMO. We report the successful outcome of a 17-day-old infant with severe pertussis managed with early initiation of ECMO and leukodepletion.
A 3200 g female baby born at 36 weeks of gestation was admitted to a local hospital at day 12 of life because of feeding difficulties. She started to cough 3 days before admission. She had been in close contact with her 3-year-old sister who had a cough illness that was not initially recognized as pertussis. Second, polymerase chain reaction in the nasopharyngeal aspirate was positive for Bordetella pertussis both for her mother and sister. Within 24 hours, the infant developed respiratory failure and apnea requiring endotracheal intubation and mechanical ventilation. Despite antibiotic treatment, conventional ventilation and later high frequency oscillation ventilation, the respiratory status continued to deteriorate with extensive bilateral pulmonary infiltrates. At that time, the echocardiography did not show pulmonary hypertension and the hemodynamic status remained stable without inotropic support. Eight days after the onset of the disease, she was transferred to our institution. On admission, laboratory studies revealed the presence of extreme leukocytosis [120,000 white blood cells (WBCs)/μL], and echocardiography still excluded pulmonary hypertension. A chest radiograph showed right upper lobe consolidation, and the arterial blood gasses, under appropriate mechanical ventilation (AI 22, PEEP 5, rate 40, FiO2 0,80), showed respiratory acidosis (pH: 7.22, PaO2: 60 mm Hg, and PaCO2: 63 mm Hg). Urgent double volume exchange was performed and the white blood cell (WBC) count dropped to 26,000/μL. During the following 24 hours, the WBC count increased to 45,000/μL, with worsening respiratory status. The echocardiography (performed 9 days after disease onset) demonstrated pulmonary hypertension and a dilated failing right ventricle, with a normal left ventricular function. Before using inotropic agents and pulmonary vasodilators, the use of ECMO was considered in anticipation of a rapid hemodynamic collapse that was reported in pertussis with severe pulmonary hypertension. Thus, our patient was placed on veno-arterial (VA) ECMO support with a pump flow of 300 to 500 mL/min and FiO2 between 50% and 90%. The arterial cannula was placed in the right common carotid artery and the venous cannula in the right internal jugular vein (see Figure, panel A, Supplemental Digital Content 1, http://links.lww.com/INF/C178). Leukofiltration using a WBC filter placed on the ECMO circuit was performed at the time of circuit priming (see Figure, panel B, Supplemental Digital Content 1, http://links.lww.com/INF/C178), and the WBC count decreased to 7000/μL. The infant was kept on VA ECMO for 7 days, with lung protective mechanical ventilation (AI 10, PEEP 12, rate 40, FiO2 0,40). After 24 hours of ECMO, inhaled nitric oxide treatment was initiated at 20 ppm and exogenous surfactant was administrated. Because it promoted lung recruitment and improved blood oxygenation, prone positioning for 4 hours, every 8 hours, was performed during 10 days. Progressive improvement in the overall condition was noted 36 hours after ECMO initiation. With decrease of pulmonary artery pressure, both the lung compliance and gas exchange improved (pH: 7.45, PaCO2: 44 mm Hg and PaO2: 200 mm Hg).
After 7 days, the patient was successfully weaned from ECMO. During the first 3 days of ECMO, she was curarized and sedated with midazolam and morphine. Thereafter, the sedation was progressively reduced allowing adequate diaphragmatic activity, which is required for synchronized ventilation with the use of neurally adjusted ventilator assist (NAVA). NAVA was initiated 17 days after disease onset. One day later, the WBC count rose again to 40,000/μL. Thus, the patient was treated with 1 last exchange leukofiltration through the central venous catheter (Figure, panel C, Supplemental Digital Content 1, http://links.lww.com/INF/C178). She was discharged to the referring hospital, where she was weaned off the ventilator within 15 days. She was discharged home a few weeks afterwards with no apparent pulmonary or neurologic impairment.
Hyperleukocytosis (>100,000/μL) is typically seen in pertussis infants with the most severe disease, and is associated with high mortality rate.3 Pertussis toxin produces lymphocyte proliferation and is thought to be responsible for the leukocytosis observed with pertussis infection.4 It results in a hyper viscosity syndrome that leads to obstruction of the pulmonary arterioles. The leukocytes aggregate within the pulmonary circulation and produce pulmonary arteriolar thrombosis, which explains the unusually refractory nature of pulmonary hypertension. Infants with severe respiratory failure, hyperleukocytosis and pulmonary hypertension are poorly responsive to inhaled nitric oxide and high frequency ventilation. ECMO can be considered in these cases, although its impact on the outcome of severe pertussis infection is debatable. The extracorporeal life support organization registry from 1992 to 2009 demonstrated a significantly higher mortality rate for pertussis infants requiring ECMO, than for other ECMO indications, because of the potential permanent refractory pulmonary hypertension.5–7 Many reports recommend ECMO for infants with multiple poor prognostic factors, such as an age less than 1 year, pulmonary hypertension and WBC count >100,000/μL. However, to improve survival, ECMO should be initiated early.7 Furthermore, VV ECMO appears to be superior to VA ECMO in these patients, probably by maintaining normal pulmonary flow.8 Infants treated with VA ECMO required circulatory support because of severe pulmonary hypertension. Almost all of them died because of pulmonary consolidation and leukocytes infiltration of lung parenchyma.
In our case, the high WBC count associated with pulmonary hypertension and right ventricular dysfunction led us to promptly initiate VA ECMO, which in turn, allowed us to associate it with aggressive leukodepletion using the ECMO circuit. In infants less than 90 days of age, a rapid leukodepletion associated with ECMO was proved to be more efficient than ECMO solely (survival rate of 90% vs. 55%).5 In our case, the WBC increased again after the ECMO removal and a second leukofiltration was performed. We do not have any valid explanation for this recurrent increase of the WBC. In cases of extreme leukocytosis and when used early, double exchange transfusion may prevent leukocyte aggregates in the pulmonary vasculature, thus avoiding pulmonary hypertension, hemodynamic collapse and ECMO.
Multiple therapies have also been described to treat acute respiratory failure in these infants. Surfactant therapies decrease mortality and are associated with shorter mechanical ventilation duration.9 NAVA improves patient-ventilator synchrony and reduces the work of breathing in patients on ECMO.10 The combination of ECMO, prone positioning and NAVA may provide protective ventilation with optimized gas exchange.
The combination of hyperleukocytosis and pulmonary hypertension defines severe pertussis, which is associated with high mortality rate. Even though the strategy we describe is not a standard recommendation in such cases, we believe that early ECMO initiation associated with leukodepletion techniques can improve outcome.
1. Mattoo S, Cherry JD. Molecular pathogenesis, epidemiology, and clinical manifestations of respiratory infections due to Bordetella pertussis
and other Bordetella subspecies. Clin Microbiol Rev. 2005;18:326–382
2. World Health Organization. . Pertussis: immunization surveillance, assessment and monitoring. Available at: http://www.who.int/gho/immunization/en/
. Accessed February 22, 2015
3. Pierce C, Klein N, Peters M. Is leukocytosis a predictor of mortality in severe pertussis infection? Intensive Care Med. 2000;26:1512–1514
4. Kerr JR, Matthews RC. Bordetella pertussis
infection: pathogenesis, diagnosis, management, and the role of protective immunity. Eur J Clin Microbiol Infect Dis. 2000;19:77–88
5. Rowlands HE, Goldman AP, Harrington K, et al. Impact of rapid leukodepletion on the outcome of severe clinical pertussis in young infants. Pediatrics. 2010;126:e816–e827
6. De Berry BB, Lynch JE, Chung DH, et al. Pertussis with severe pulmonary hypertension and leukocytosis treated with extracorporeal membrane oxygenation. Pediatr Surg Int. 2005;21:692–694
7. Halasa NB, Barr FE, Johnson JE, et al. Fatal pulmonary hypertension associated with pertussis in infants: does extracorporeal membrane oxygenation have a role? Pediatrics. 2003;112(6 Pt 1):1274–1278
8. Theilen U, Johnston ED, Robinson PA. Rapidly fatal invasive pertussis in young infants–how can we change the outcome? BMJ. 2008;337:a343
9. Duffett M, Choong K, Ng V, et al. Surfactant therapy for acute respiratory failure in children: a systematic review and meta-analysis. Crit Care. 2007;11:R66
10. Mauri T, Bellano G, Confalonieri A, et al. neutrally adjusted ventilator assist (nava) improves patient-ventilator interaction in severe patients undergoing ECMO
. Intensive Care Med. 2013;39:282–91