Pumpless extracorporeal carbon dioxide removal for life-threatening asthma

Case Report

Pumpless extracorporeal carbon dioxide removal for life-threatening asthma Stuart C. Elliot, BHSc; Kumar Paramasivam, MD, MRCP; John Oram, MB ChB, FRCA; Andrew R. Bodenham, FRCA; Simon J. Howell, MD, FRCA; Abhiram Mallick, MD, FRCA

Objective: To report the use of pumpless extracorporeal carbon dioxide removal in two cases of acute severe asthma. Design: Case reports. Setting: Adult general intensive care unit, Leeds General Infirmary, Leeds, UK. Patients: A 74-yr-old male and 52-yr-old female with lifethreatening asthma developed progressive hypercapnia and severe acidosis that proved nonresponsive to all other therapies. Intervention: Initiation of extracorporeal arteriovenous carbon dioxide removal using the Novalung device (Novalung GmbH, Lotzenäcker 3, D-72379 Hechingen, Germany).

S

evere life-threatening asthma can manifest in dangerous levels of hypercapnia and acidosis, the permissible extremes for which are unknown. Patients with severe asthma requiring mechanical ventilation have a significant mortality and morbidity (1). Mechanical ventilation itself can cause increased air trapping and hyperinflation, predisposing the lungs to barotraumas (2). Equally, some patients fail to respond to drug treatment, including the less conventional therapies such as volatile inhalation anesthetics and intravenous ketamine (3). Extracorporeal membrane oxygenation has shown some promise in acute respiratory distress syndrome in adults (4), but its role in severe asthma has been limited to case reports (5). Its usage requires high intensity and much specialized care (3). Conversely, extracorporeal carbon dioxide removal using the Novalung is comparatively simple. It is a pumpless low-resistance gas exchange membrane reliant on arteriovenous

From Leeds Teaching Hospitals NHS Trust, Anaesthetic Department, Leeds General Infirmary, Leeds, UK. Copyright © 2007 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/01.CCM.0000257462.04514.15

Crit Care Med 2007 Vol. 35, No. 3

Main Results: The addition of extracorporeal carbon dioxide removal to mechanical ventilation corrected hypercapnia and acidosis, allowing reduction of other supportive measures. In both cases, adequate gas exchange was maintained until their underlying condition improved sufficiently for device removal. The two patients were subsequently weaned from mechanical ventilation and made a full recovery. Conclusions: Extracorporeal carbon dioxide removal proved to be a valuable adjunct to mechanical ventilation and other medical treatment. (Crit Care Med 2007; 35:945–948) KEY WORDS: extracorporeal carbon dioxide removal; Novalung; acute severe asthma; mechanical ventilation; hypercapnia

shunt and a sweep flow of oxygen for membrane diffusion of oxygen and carbon dioxide (Fig. 1). The use of this device has been reported recently for acute respiratory distress syndrome (6), but to our knowledge, this is the first use in life-threatening asthma. The features of each technique are summarized in Table 1.

CASE REPORTS Case 1. A 74-yr-old male ex-smoker and known asthmatic, not receiving regular inhaled medications, was admitted through the emergency room following collapse and respiratory arrest. On arrival, he required tracheal intubation and assisted ventilation. A chest radiograph revealed hyperinflated lung fields. Initial arterial blood gas showed pH 7.08, PaO2/ FIO2 158.2 mm Hg, and PaCO2 106.5 mm Hg. He failed to respond to nebulized salbutamol and ipratropium bromide and therapeutic doses of intravenous aminophylline, magnesium sulfate, and steroids. At transfer to the intensive care unit, mechanical ventilation was attempted in accordance with general principles of ventilation for severe acute asthma, namely sedation and analgesia, neuromuscular blocking drugs, low respiratory rate, prolonged expiratory time, and low extrinsic positive

end-expiratory pressure. This, however, proved extremely difficult because of air trapping, and despite attempting numerous maneuvers, the patient remained hypercapnic and acidotic. Measured expiratory times were ⬎45 secs and tidal volumes ⬍200 mL. Soon after intensive care unit (ICU) admission, the patient became hypotensive and was administered an epinephrine infusion, and later norepinephrine, to maintain adequate mean arterial pressure. Ketamine and isoflurane were administered but produced no improvement. After 9 hrs of ventilation, the PaCO2 had reached 147.52 mm Hg, with pH 6.87. This was associated with worsening hypotension, despite aggressive hemodynamic support and volume loading. It was, therefore, decided to reduce PaCO2 with the Novalung. Under ultrasound guidance, the left femoral artery and vein were cannulated with size 15- and 17-Fr catheters, respectively. Blood was then gently released into the Novalung circuit, with an initial flow of 1.5 L·min⫺1. During a period of 4 hrs, oxygen sweep across the membrane was increased to a maximum rate of 15 L·min⫺1. This produced a corresponding fall in PaCO2 to 62.85 mm Hg and pH 7.14 (Fig. 2). Pressure-controlled ventilation was maintained at 10 breaths per min, with an inspiratory pressure of 33 cm 945

Figure 1. The Novalung connected from the femoral artery to the femoral vein. Table 1. Differences between Novalung and extra corporeal membrane oxygenation (ECMO)

Pump Mode Action Membrane surface area Blood flow rate Fill volume O2 sweep flow

Novalung

ECMO

Pumpless Arterial-venous CO2 removal 1.3 m2 0.5–4.5 L.min⫺1 0.24 L 1–15 L䡠min⫺1

Roller pump Venovenous/venoarterial Oxygenator/CO2 removal 1.9 m2 or greater 1–7 L䡠min⫺1 Up to 1.4 L Normally 1–10 L䡠min⫺1

Figure 2. Changes in pH and PaCO2 with Novalung.

H2O and an inspiratory:expiratory ratio of 1:8.8. Over time, tidal volumes improved up to 600 mL. From admission, PaO2 was never problematic. Novalung treatment was continued for a total of 121 hrs, but the membrane gas exchanger required changing after 37 hrs. At this time, blood flow through the Novalung had fallen to 0.6 L·min ⫺1 , suggesting coagulation within the device, and the PaCO2 had increased to 67.5 mm Hg. A white marbling effect was apparent within the membrane. This was attributed to the lipid component of propofol. As a consequence, the sedation regimen was changed to midazolam and a new membrane gas exchanger was attached to the existing catheters. On Novalung day 3, the patient’s platelet count was 29 ⫻ 109/L. Systemic 946

anticoagulation with low-dose heparin at 140 IU/hr was stopped in view of a possible heparin-induced thrombocytopenia, but the Novalung was continued despite the risk of heparin washout from the heparinbonded device. After the PaCO2 had remained stable for 93 hrs between 45 and 60 mm Hg and signs of bronchospasm had resolved, oxygen flow to the Novalung was reduced to zero. The extracorporeal circuit was kept in place for a further 28 hrs and then disconnected. Both cannulas were individually removed and manual compression applied until cessation of bleeding. Unfortunately, the patient became intermittently hypertensive while coughing, and this caused bleeding at the femoral arterial catheter site. This did not settle with manual compression,

and the patient was taken to the operating room for exploration of the femoral artery and closure by the vascular surgical team. The patient remained intubated for 2 additional days as mechanical ventilatory support was slowly weaned. Thereafter, the patient was successfully extubated and transferred to the ward 2 days later. He spent an additional 19 days in the hospital before discharge to home. All serial cultures of respiratory secretions returned negative throughout the period of hospitalization. Case 2. A 52-yr-old female, who was a known brittle asthmatic, was admitted to the emergency room with increasing shortness of breath. She had been a heavy smoker for 35 yrs and had been admitted to the hospital four times in the previous 2 yrs with acute exacerbations of asthma. None of these had warranted intensive care, but she had required noninvasive ventilation on her last hospitalization. At this admission, examination revealed widespread wheeze and pronounced accessory muscle breathing. Her chest radiograph was unremarkable. She was initially treated with nebulized bronchodilators, steroids, and noninvasive ventilation but soon became tired and drowsy. Arterial blood gases revealed increasing hypercapnia, PaCO2 ⬎60 mm Hg, with worsening respiratory acidosis, pH ⬍7.2, prompting tracheal intubation and mechanical ventilation. She was then transferred to the ICU. Here, despite numerous ventilatory changes, maximal bronchodilator therapy, and antibiotics, the patient’s condition continued to necessitate sedation and paralysis, with PaCO2 values never dropping below 52.5 mm Hg. On ICU day 4, the PaCO2 rose to 75 mm Hg and the patient’s chest radiograph revealed patchy bilateral shadowing. The patient’s airway pressures remained high, and her tidal volumes were unacceptably low. At this point, volatile anesthetic inhalation was started, but this had little effect. PaCO2 continued to rise to a peak of 130.5 mm Hg, with a corresponding pH 7.2. The Novalung was then inserted for extracorporeal carbon dioxide removal. The left femoral artery and right femoral vein were cannulated using 15- and 17-Fr catheters, respectively. Oxygen sweep flow across the Novalung membrane was initially titrated slowly upwards for a period of approximately 4 hrs, to a maximum of 10 L/min. This was found sufficient to maintain PaCO2 at approximately 45 mm Hg. Blood flow across the filter at this time was between 1.4 and 1.6 L/min⫺1. Despite normalization of PaCO2, hypoxia became problematic for the first Crit Care Med 2007 Vol. 35, No. 3

time since admission. A FIO2 of 0.8 was required to maintain adequate oxygenation, and the patient was placed in the prone position. There were increasing signs of pulmonary infection, with profuse dirty green sputum, pyrexia, elevated C-reactive protein, and white cell count. This problem resolved within 24 hrs, and the FIO2 was reduced to 0.4. Four days after commencement of extracorporeal gas exchange, the patient’s overall condition improved sufficiently to enable weaning of the membrane oxygen sweep flow to zero while maintaining satisfactory arterial blood gases. The cannulas were explanted the following day by the vascular surgical team, and the patient subsequently began a slow ventilatory wean. Tracheal aspirates grew Pseudomonas aeruginosa. She was extubated on intensive care day 11 and discharged to the adult respiratory care unit on day 12. The patient was discharged to home after a hospital stay of 32 days. The local ethics committee does not require the review of clinical case reports, but consent for publication was obtained from both patients.

DISCUSSION The use of the Novalung in conjunction with mechanical ventilation in both of these cases of life-threatening asthma was able to correct severe hypercapnia and acidosis until the underlying disease improved. Despite advances in the understanding of ventilating asthmatic patients and newer drug therapies, there remains a significant mortality and morbidity in severe cases (7). Those who require tracheal intubation and ventilation are, by definition, at the severe end of the spectrum. Such patients are difficult to manage, and the institution of ventilation, although potentially life-saving, may present formidable problems in terms of achieving adequate gas exchange (5). Mechanical ventilation can result in dynamic hyperinflation, despite measures such as low respiratory rate, prolonged expiratory time, and ensuring extrinsic positive endexpiratory pressure is less than intrinsic positive end-expiratory pressure (2). There is a risk of secondary barotrauma resulting in complications such as pneumothoraces and air leaks (8). Typically, patients with acute severe asthma have single organ failure. If they can be stabilized for a period of time, bronchospasm should settle, and a full recovery can be expected unless other Crit Care Med 2007 Vol. 35, No. 3

complications ensue. Both of these patients were known asthmatics but may have had undiagnosed chronic obstructive airway disease resulting from their smoking. Whatever the case, the use of extracorporeal carbon dioxide removal enabled normalization of unacceptable hypercapnia and acidosis and allowed us to restrict conventional ventilation and reduce pharmacologic supportive measures. The use of extracorporeal carbon dioxide removal as a rescue therapy is attractive in the clinical context of severe asthma because of the following: 1) bronchospasm in severe asthma will usually settle over time, meaning that any extracorporeal support should only be required for a relatively short time period; 2) the physical characteristics of gas transfer and blood flow over the Novalung mean that it is much more effective at CO2 clearance than oxygenation. Most severe asthmatics, as seen in these two cases, do not have major problems with oxygenation but do suffer from progressive failure of ventilation manifested by rising PaCO2; 3) severe asthmatic patients normally have sufficient cardiac output and blood pressure to maintain arteriovenous shunt across the membrane. Most should not have peripheral vascular disease with attendant risks of large-bore arterial cannulation. Patients with intractable bronchospasm, who have failed all usual interventions, have an arterial PaCO2 ⬎110 mm Hg, and have severe acidosis, are possible candidates for extracorporeal carbon dioxide removal. Ideally, this should be studied in a formal trial setting, although there would be obvious difficulties given the small numbers involved in each centre. It is interesting to speculate whether the Novalung might be better formatted in a venovenous pumped system. A pump would allow reliable blood flow, even in a patient in shock and avoids the use of large-bore arterial cannula and the related risks. Theoretically, increasing blood flow through the membrane would also enable large venous blood volumes of 4 – 6 L·min⫺1 return to the systemic circulation saturated with oxygen impacting

on PaO2. However, any pump system would necessitate systemic anticoagulation, would prove more labor intensive, and may be proinflammatory (6). The feedback from the intensive care staff managing the Novalung was that they liked it because of its simplicity and ease of use. To limit potential insertion and removal problems, we suggest insertion of the cannulas under ultrasound guidance and surgical removal of the arterial cannula. Because the Novalung membrane is fully heparin-bonded, additional anticoagulation, although desirable, is not essential (9). When we stopped anticoagulation because of a suspected heparininduced thrombocytopenia, which later was confirmed to be negative, the device worked adequately. We were, however, initially unaware of the fact that high concentrations of circulating triglycerides were known to deposit in the Novalung membrane, reducing circulating blood flow and impairing membrane function (9). Certainly this was not an issue in the second case because when initiating Novalung, sedation was changed from propofol to midazolam. There may be a benefit to instituting extracorporeal carbon dioxide removal before mechanical ventilation if the situation permits. Timely use of such intervention in the severe asthmatic who is obviously deteriorating could avoid the need for mechanical ventilation altogether. Oxygenation could be supported with continuous positive airway pressure or noninvasive ventilation. It would then be a matter for investigation whether the overall risks of arterial and venous cannulation, the use of a membrane, and associated anticoagulation would have less risk-benefit than traditional tracheal intubation and positive-pressure ventilation. We believe this area of study is well worth pursuing in the future.

REFERENCES 1. McFadden ER Jr: Acute severe asthma. Am J Respir Crit Care Med 2003; 168:740 –759 2. Oddo M, Feihl F, Schaller M, et al: Management of mechanical ventilation in acute severe asthma: Practical aspects. Intensive Care Med 2006; 32:501–510 3. Phipps P, Garrad CS: The pulmonary physician in critical care. 12: Acute severe asthma in the intensive care unit. Thorax 2003; 58:81– 88 4. Lewandowski K, Rossaint R, Pappert D, et al: High survival rate in 122 ARDS patients managed according to a clinical algorithm includ-

947

ing extracorporeal membrane oxygenation. Intensive Care Med 1997; 23:819 – 835 5. Leiba A, Bar-Yosef S, Bar-Dayan Y: Early administration of extracorporeal life support for near fatal asthma. Isr Med Assoc J 2003; 5:600 – 602 6. Bein T, Weber F, Philipp A, et al: A new pump-

948

less extracorporeal lung assist in critical hypoxemia/hypercapnea. Crit Care Med 2006; 34: 1372–1377 7. Krishnan V, Diette GB, Rand CS, et al: Mortality in patients hospitalised for asthma exacerbations in the United States. Am J Respir Crit Care Med 2006; In Press

8. Williams TJ, Tuxen DV, Scheinkestel CD, et al: Risk factors for morbidity in mechanically ventilated patients with acute severe asthma. Am Rev Respir Dis 1992; 146:607– 615 9. Novalung GmbH, Lotzenäcker 3, Hechingen, Germany. http://www.novalung.com. Accessed on August 1, 2006

Crit Care Med 2007 Vol. 35, No. 3