Fatal Pulmonary Hypertension Associated With Pertussis in Infants: Does Extracorporeal Membrane Oxygenation Have a Role?
* Divisions of Pediatric Infectious Disease
Pediatric Critical Care, Department of Pediatrics, Vanderbilt University Medical Center, Nashville, Tennessee
Department of Pathology, Vanderbilt University Medical Center, Nashville, Tennessee
 |
ABSTRACT |
|---|
 TOP ABSTRACT CASE PRESENTATIONSMETHODSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Objective. The deaths of 4 infants who had confirmed pertussis infection at our hospital during the past year motivated us to review their cases as well as the use of extracorporeal membrane oxygenation (ECMO) in respiratory failure caused by pertussis.
Methods. Retrospective chart reviews of the 4 infants who had pertussis and died at Vanderbilt University Medical Center (VUMC) from May 2001 to May 2002 were conducted. The Extracorporeal Life Support Organization (ELSO) database is an international voluntary registry established in 1986 to compile passive reports of adult and pediatric patients placed on ECMO. We searched this database for pertussis cases and analyzed the clinical outcomes by age, ventilatory management, and measurements of cardiopulmonary status.
Results. All 4 infants who died from pertussis infection at VUMC were younger than 3 months and had severe pulmonary hypertension, and ECMO therapy was considered for respiratory failure. Review of the international ELSO database, focusing only on children from 1986 to July 2002, revealed a total of 23 970 patients placed on ECMO. Since the first pertussis case treated with ECMO in 1990, a total of 61 children with pertussis have been treated with ECMO, representing an increase from 0.09% to 1% of the total cases. Mean age of the pertussis patients placed on ECMO was 88 days (1 day–2.7 years). The overall mortality was 70.5% (43 of 61) but was significantly higher for infants who were younger than 6 weeks (84%) compared with infants who were older than 6 weeks (61%). When evaluating pre-ECMO management, survivors had received significantly higher mean positive end expiratory pressures than nonsurvivors (11.1 ± 4.5 vs 7.3 ± 3.1 cmH2O) and had significantly higher serum pH than nonsurvivors (7.31 ± 0.14 vs 7.14 ± 0.19). There was no effect of duration of ECMO, positive inspiratory pressure, mean arterial pressure, ECMO mode (venoarterial vs venovenous mode), or sex on mortality.
Conclusions. Although the ECMO registry is not an active surveillance system, it suggests that the numbers of children who require ECMO for pertussis have significantly increased. The high fatality rates of pertussis patients who have placed on ECMO are alarming and should stimulate reevaluation of whether this high-risk intervention is beneficial to patients with pertussis. This review, coupled with the 4 infants who died of pertussis at VUMC, suggests that improved measures are needed to prevent pertussis in infants.
Key Words: pertussis • ECMO • pulmonary hypertension • infants
Abbreviations: VUMC, Vanderbilt University Medical Center • ECMO, extracorporeal membrane oxygenation • ELSO, Extracorporeal Life Support Organization • ECHO, echocardiogram • PCR, polymerase chain reaction • NO, nitric oxide • DFA, direct fluorescent antibody • PEEP, positive end expiratory pressure • ECLS, extracorporeal life support
Pertussis (whooping cough) is a bacterial respiratory infection caused by Bordetella pertussis, a Gram-negative bacillus. Its major manifestation is a protracted cough illness that lasts many weeks, often referred to as the "100-day cough."1 The disease is most severe in infants and young children. In fact, pertussis was a major cause of death among infants and children in the United States during the prevaccine era, with approximately 9000 pertussis-related deaths occurring in 1923.2 After the introduction of the whole-cell pertussis vaccine in the 1940s, the number of pertussis cases markedly declined. However, reported pertussis cases in infants have tripled in the past few years.3 Infants younger than 6 months, too young to have completed their primary immunization series, have the highest rates of morbidity and mortality.3 Many experience paroxysmal cough as well as apnea, pneumonia, respiratory collapse, and even death. Furthermore, the role of adolescents and adults in transmission of pertussis to infants has been demonstrated.4
We report 4 fatal cases of pertussis in infants who were too young to have been fully immunized and were seen at Vanderbilt University Medical Center (VUMC) during the past year. All 4 infants developed pneumonia, refractory pulmonary hypertension, and respiratory failure and died. Because they failed conventional methods of oxygenation, extracorporeal membrane oxygenation (ECMO) support was planned. Patterns of ECMO use and patient complications and outcomes obtained from the international ECMO registry, the Extracorporeal Life Support Organization (ELSO), were also analyzed.
|
CASE PRESENTATIONS |
|---|
|
TOPABSTRACTCASE PRESENTATIONSMETHODSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Case 1
A 16-day-old term white female infant was admitted to VUMC in May 2002 with a 4-day history of sneezing, poor feeding, and cough and the acute onset of paroxysmal cough and increasing respiratory distress. She had been evaluated by her pediatrician 2 days earlier and prescribed oral erythromycin. The mother denied sick exposures, child care attendance, and other siblings. On admission, she was afebrile with a respiratory rate of 110 breaths per minute and pulse rate of 158 beats per minute. There were moderate chest retractions, and rales were heard bilaterally. A chest radiograph revealed a right upper lobe pneumonia with atelectasis and/or infiltrate in the left lower lobe (Table 1). Parenteral ampicillin, gentamicin, and erythromycin were administered. During the next 48 hours, the infant progressed to respiratory failure requiring first mechanical ventilation and then ECMO support. An echocardiogram (ECHO) revealed right ventricular enlargement, bidirectional flow at the foramen ovale, good biventricular wall motion, and tricuspid insufficiency consistent with pulmonary hypertension. On hospital day 13, the ECMO circuit clotted and tissue plasminogen activator was administered. On hospital day 14, the patient’s pupils were noted to be unequal and head ultrasound revealed a massive intracranial hemorrhage. She died that same day.
|
An autopsy was performed and revealed diffuse hemorrhagic consolidation of the left lung and large patchy areas of hemorrhagic consolidation of the right lung, with the remaining lung parenchyma showing consolidation without hemorrhage (Fig 1). Arteries of all sizes were acutely thrombosed, with extensive tissue necrosis. Acute bronchiolitis was the predominant inflammatory lesion, with only rare foci of inflammatory alveolar consolidation (pneumonitis). Tissue Gram stain did not reveal any organisms, but B pertussis polymerase chain reaction (PCR) on autopsy lung tissue was positive despite 14 days of erythromycin therapy before death.
|
Case 2
A 53-day-old former 31-week estimated gestational age white female infant was admitted in August 2001 with a 7-day history of nasal congestion and cough. The infant’s 5-year-old sibling had a 4-week history of cough and posttussive emesis, and several other family members, including her twin sister, also had nasal congestion and cough. One day before admission, she was evaluated at an outside emergency department, where tachypnea and retractions were documented and a normal chest radiograph was obtained. During the next 24 hours, her respiratory distress worsened and she returned to the emergency department, was intubated for respiratory failure, and was transported by helicopter to VUMC. A repeat chest radiograph revealed a right upper lobe infiltrate (Table 1). During the next 48 hours, she developed severe pulmonary hypertension documented on ECHO with right ventricle dilation and tricuspid valve regurgitation. Hypoxemia developed that was unresponsive to high-frequency oscillatory ventilation and inhaled nitric oxide (NO). Arrangements were made for initiation of ECMO on hospital day 3, but the patient died just before cannulation.
Case 3
A 61-day-old former 32-week estimated gestational age black female infant was admitted in January 2001 to VUMC with a 2-day history of cough and tachypnea. Chest radiograph on admission revealed a right upper lobe infiltrate (Table 1). Parenteral ampicillin, cefotaxime, and erythromycin were begun on admission. Respiratory failure requiring intubation occurred 24 hours after admission. An ECHO revealed right ventricle dysfunction and suprasystemic pulmonary artery pressures, consistent with pulmonary hypertension. Increasing hypoxia led to a therapeutic trial with inhaled NO and then ECMO support on hospital day 3. After 10 days of ECMO support, the right ventricular function and pulmonary artery pressures improved and the patient was decannulated and inhaled NO was restarted. A repeat ECHO showed worsening right ventricular function and pulmonary hypertension. Profound hypoxemia and systemic hypotension recurred, and life support was withdrawn on hospital day 20.
Case 4
A 23-day-old term white female infant was admitted to a referring pediatric intensive care unit in May 2001 with a 3-day history of cough and nasal congestion. Five other family members had cough symptoms. Bacterial cultures of blood, cerebrospinal fluid, and urine were negative. A nasal swab for pertussis direct fluorescent antibody (DFA) was positive. Parenteral ampicillin, gentamicin, and erythromycin were initiated on admission. The patient required intubation and mechanical ventilation on hospital day 7 and was transferred to VUMC for worsening hypoxemia (Table 1). ECHO revealed right ventricle dysfunction and suprasystemic pulmonary arterial pressures, consistent with pulmonary hypertension. Severe hypoxemia developed and was unresponsive to inhaled NO; ECMO was initiated shortly after admission. On ECMO day 7, she had neurologic deterioration and a head ultrasound revealed a large intraventricular hemorrhage. Life support measures were discontinued.
|
METHODS |
|---|
|
TOPABSTRACTCASE PRESENTATIONSMETHODSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
ELSO Database Review
The ELSO is an international voluntary registry housed in Ann Arbor, Michigan. It has been in existence since 1986, and it receives passive reports of both adult and pediatric patients who are placed on ECMO. To assess the role of ECMO in the treatment of patients with pertussis, we queried the database for all pertussis cases and analyzed the data by age, ventilatory management, measurements of cardiopulmonary status, and mortality. From January 1986 to July 2002, a total of 23 970 children had been placed on ECMO. The first pertussis case on ECMO was reported in 1990, with a total of 61 children since that time. Means between groups were compared using the t test, and rates were compared by
2 using STATA (STATA Press). Assessing the temporal trend of pertussis cases treated with ECMO over time, a significant increase from 0.09% to 1% of the total cases was noted (P < .001; Table 2). Demographic characteristics of the pertussis patients on ECMO revealed that the mean age was 88 days (range: 1–964 days; median: 47 days). The overall mortality was 70.5% (43 of 61). Mortality rates were significantly higher for infants younger than 6 weeks (84%) compared with infants older than 6 weeks (61%; P = .05). Before initiation of ECMO, survivors had received significantly higher pre-ECMO mean positive end expiratory pressures (PEEP) than nonsurvivors (11.1 ± 4.5 vs 7.3 ± 3.1 cmH2O; P = .0058; Table 3). Pre-extracorporeal life support (ECLS) arterial blood gas data demonstrated survivors also had a significantly higher serum pH than nonsurvivors (7.31 ± 0.14 vs 7.14 ± 0.19; P = .0018; Table 4). There were no significant differences in mortality between the 2 groups with respect to duration of ECMO, mean arterial pressure, ECMO mode (venoarterial versus venovenous), or sex (Table 5).
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTCASE PRESENTATIONSMETHODSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Reported cases of infant pertussis have steadily increased in the past few years with a recent tripling in rates.3 Earlier reports of infants with pertussis have highlighted the increased mortality and have described pulmonary hypertension, documented by either ECHO or direct measurements of pulmonary pressures.1,5–12 All 4 of our patients had documented pulmonary hypertension.
Our patients were unresponsive to conventional management of pulmonary hypertension with NO and high-frequency ventilation, and ECMO was attempted. However, very low survival rates in our patients and others in the literature prompted us to review the ECLS database for patients with the diagnosis of pertussis. An earlier comprehensive review including the study years of 1987–1997 revealed 17 deaths in the 22 pertussis patients (78% mortality). Reanalysis of the same database from 1986 through July 2002 identified 43 deaths in 61 pertussis patients (70.5% mortality; P = .85 comparing % deaths from the 2 studies). Patients younger than 6 weeks had the highest mortality. When comparing mortality rates in patients with pertussis with those on ECMO for other reasons, the observed mortality with pertussis was higher. Overall ECMO mortality rates for all respiratory causes varied from 14% to 35% for respiratory causes (P < .001 for pertussis deaths compared with respiratory deaths), 45% for cardiac disease (P < .001 for pertussis deaths compared with cardiac deaths), and 38% to 55% for other causes. In addition, over time, the number of children who had pertussis and were placed on ECMO increased significantly. The increase in ECMO use with pertussis may reflect an increase in disease severity, an increase in disease rates, improved reporting of ECMO cases (despite consistent numbers of centers participating in ELSO), more sensitive techniques for diagnosing pertussis (eg, PCR), greater exposure of infants to adolescents and adults with waning immunity, or a combination of these.
Risk factors for mortality in the ECLS database included ventilator strategy and acidosis before the initiation of ECMO. Survivors had received higher PEEP than nonsurvivors, probably reflecting improved ventilation in infants who ultimately survived. Pre-ECMO pH was significantly lower in infants who died but was not explained by differences in partial pressure of arterial carbon dioxide or bicarbonate levels. Serum lactate levels have been shown to be a better predictor of poor tissue perfusion and oxygen delivery than serum bicarbonate levels, but these values were not available in the ELSO registry.13
The high mortality rate for infants with pertussis and intractable pulmonary hypertension raises the question of whether ECMO should be provided to this patient population, and, if so, what approach should be used. The findings that survivors had higher PEEP settings and higher pH values before ECMO than nonsurvivors may indicate that open-lung ventilation strategies to reduce barotrauma and early initiation of ECMO before cardiovascular failure ensues may improve outcomes. Although ECMO should still be offered as a therapeutic modality, physicians should discuss with parents that the majority of these infants do not survive.
Randomized clinical trials to determine efficacy of potential treatments for intractable pulmonary hypertension in the setting of pertussis would be difficult to perform. Such a study would require a number of centers to participate, and when children are failing conventional management, some physicians may be reluctant to withhold therapy that might have potential benefit. One approach might include a randomized trial of high-frequency oscillatory ventilation plus inhaled NO compared with early ECMO intervention in infants with documented suprasystemic pulmonary artery pressures.
In addition, when infants with unexplained pulmonary hypertension and respiratory failure present, the diagnosis of pertussis should be investigated. The standard approaches to the diagnosis of pertussis include culture, DFA, serology, and PCR. At this point, culture remains the gold standard for diagnosis, but the fastidious nature of the B pertussis organism makes this method less sensitive. In addition, factors such as previous use of antibiotics, patient age, immunization status, duration of symptoms before specimen collection, and specimen transport conditions also affect culture results.14,15 DFA can provide more rapid results than cultures but is also a test with poor sensitivity and varying specificity.14 Serologic tests can be very specific and sensitive but usually require both an acute and a convalescent phase and are not as reliable in infants as in older children and adults.15 Several commercial and state laboratories have developed PCR assays for the diagnosis of pertussis and have been found to be more sensitive than culture.14,16 However, PCR is not uniformly available at all centers, and calcium-alginate swabs used to obtain nasopharyngeal aspirates have been shown to inhibit PCR studies.17
The exact mechanism for the induction of intractable pulmonary hypertension by pertussis infection is not known. Hypoxia alone does contribute to pulmonary hypertension, but ECMO mortality rates with pertussis are much higher than those seen with persistent pulmonary hypertension of the newborn, a similar pathophysiologic situation (21% mortality rate). Whether other factors such as toxin production by B pertussis inducing direct endothelial damage or pulmonary venous leukocyte thrombi5 are unknown and warrant additional investigation. Autopsy findings in case 1 demonstrated massive arterial thromboses that could contribute to persistent pulmonary hypertension through mechanical obstruction. However, it is not known whether the thrombi were the result of a primary or a secondary event. Paradoxically, in vitro studies associate pertussis toxin with the attenuation of pulmonary hypertension.18,19 A better understanding of the cause of pulmonary hypertension in pertussis in greatly needed.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTCASE PRESENTATIONSMETHODSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Although the increase in infant deaths from pertussis and the associated refractory pulmonary hypertension have been previously reported, this article highlights the extremely high mortality rates in patients who have pertussis and are placed on ECMO. Although the ELSO registry is not an active surveillance system, an estimated 85% of ECMO cases in the United States are reported (Dr Peter Rycus, personal communication), and it does indicate that ECMO use in patients with pertussis has increased 10-fold, in line with national statistics showing a tripling of the number of cases of infant pertussis.3 The distressingly high fatality rates on ECMO, especially in infants younger than 6 weeks, should be taken into consideration while weighing the risks and benefits of this highly invasive therapy. If the pertussis rates in infants continue to increase, then there may be sufficient numbers of patients to conduct a multicenter study to evaluate various therapeutic modalities. However, the most practical approach would be to increase public awareness about infant pertussis and develop pertussis immunization programs for adolescents, adults, and pregnant women in the third trimester or to accelerate the infant vaccination schedule.
|
ACKNOWLEDGMENTS |
|---|
This work was supported in part by Public Health Service award K-12 RR017697 through the Vanderbilt Mentored Clinical Research Scholar Program (N.B.H.).
We thank Dr Peter Rycus, Extracorporeal Life Support Organization (Ann Arbor, MI), for providing the data from the registry.
|
FOOTNOTES |
|---|
Received for publication Feb 12, 2003; Accepted May 8, 2003.
Address correspondence to Kathryn M. Edwards, MD, Department of Pediatrics, Division of Infectious Diseases, CCC-5323 Medical Center North, Vanderbilt University Medical Center, 1161 21st Ave South, Nashville, TN 37232-2573. E-mail: Kathryn.Edwards{at}vanderbilt.edu
|
REFERENCES |
|---|
|
TOPABSTRACTCASE PRESENTATIONSMETHODSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
- Christie CDC, Baltimore RS. Pertussis in neonates.
Am J Dis Child.1989; 143
:1199
–1202
[Abstract/Free Full Text] - Pertussis-United States, January 1992-June 1995. MMWR Morb Mortal Wkly Rep.1995; 44 :525 –529[Medline]
- Pertussis-United States, 1997–2000. MMWR Morb Mortal Wkly Rep.2002; 51 :73 –76[Medline]
- Senzilet LD, Halperin SA, Spika JS, et al. Pertussis is a frequent cause of prolong cough illness in adults and adolescents. Clin Infect Dis.2001; 32 :1691 –1697[CrossRef][Web of Science][Medline]
- Williams GD, Numa A, Sokol J, Tobias V, Duffy BJ. ECLS in pertussis: does it have a role? Intensive Care Med.1998; 24 :1089 –1092[CrossRef][Web of Science][Medline]
- Wortis N, Strebel P, Wharton M, Bardenenheir B, Hardy I. Pertussis deaths: report of 23 cases in the United States, 1992 and 1993.
Pediatrics.1996; 97
:607
–612
[Abstract/Free Full Text] - Sreenan CD, Osiovich H. Neonatal pertussis requiring extracorporeal membrane oxygenation. Pediatr Surg Int.2001; 17 :201 –203[CrossRef][Web of Science][Medline]
- Chan L. Cough and respiratory distress in a two-month-old infant. Pediatr Infect Dis J.1997; 16 :631 –632[CrossRef][Web of Science][Medline]
- Gillis J, Grattan-Smith T, Kilham H. Artificial ventilation in severe pertussis.
Arch Dis Child.1988; 63
:364
–367
[Abstract/Free Full Text] - Goulin G, Kaya K, Bradley J. Severe pulmonary hypertension associated with shock and death in infants infected with Bordetella pertussis. Crit Care Med.1993; 21 :1791[Web of Science][Medline]
- Pierce C, Klein N, Peters M. Is leukocytosis a predictor of mortality in severe pertussis infection? Intensive Care Med.2000; 26 :1512 –1514[CrossRef][Web of Science][Medline]
- Plott KF, Pascual F, Bisgard KC, Vitek C, Murphy T, Curtis C. Pertussis deaths-United States, 2000. MMWR Morb Mortal Wkly Rep.2002; 51 :616 –618[Medline]
- Charpie J, Dekeon M, Goldberg C, Mosca R, Bove E, Kulik T. Serial blood lactate measurements predict early outcome after neonatal repair or palliation for complex congenital heart disease.
J Thorac Cardiovasc Surg.2000; 120
:73
–80
[Abstract/Free Full Text] - Loeffelholz MJ, Thompson CJ, Long KS, Gilchrist MJ. Comparison of PCR, culture, and direct fluorescent-antibody testing for detection of Bordetella pertussis.
J Clin Microbiol.1999; 37
:2872
–2876
[Abstract/Free Full Text] - von Konig CH, Halperin S, Riffelmann M, Guiso N. Pertussis of adults and infants. Lancet Infect Dis.2002; 2 :744 –750[CrossRef][Web of Science][Medline]
- Lingappa JR, Lawrence W, West-Keefe S, Gautom R, Cookson BT. Diagnosis of community-acquired pertussis infection: comparison of both culture and fluorescent-antibody assays with PCR detection using electrophoresis or dot blot hybridization.
J Clin Microbiol.2002; 40
:2908
–2912
[Abstract/Free Full Text] - Cloud JL, Hymas W, Carroll KC. Impact of nasopharyngeal swab types on detection of Bordetella pertussis by PCR and culture.
J Clin Microbiol.2002; 40
:3838
–3840
[Abstract/Free Full Text] - Olson NC, Kruse-Elliott KT, Whorton AR, Dodam JR. Pertussis toxin attenuates platelet-activating factor-induced pulmonary hemodynamic alterations in pigs. Am J Physiol.1993; 264 :L213 –L221
- Stenmark KR, Gerasimovskaya E, Nemenoff R, Das M. Hypoxic activation of adventitial fibroblasts—role in vascular remodeling.
Chest.2002; 122
:326S
–334S
[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2003 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  U Theilen, E D Johnston, and P A Robinson Rapidly fatal invasive pertussis in young infants--how can we change the outcome? BMJ, November 27, 2008; 337(nov27_2): a343 - a343. [Full Text]
|
|
|
|
 |
|
 J. Surridge, E. R Segedin, and C. C Grant Pertussis requiring intensive care Arch. Dis. Child., November 1, 2007; 92(11): 970 - 975. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
