Abstract
Coronavirus disease 2019 pneumonia with respiratory failure refractory to maximum medical therapy has been successfully managed with venovenous extracorporeal membrane oxygenation. This report describes a process of using directed hypercapnia in 5 patients to wean them from prolonged extracorporeal support secondary to refractory hypercarbic respiratory failure.
Abbreviations and Acronyms: ABG, arterial blood gas; CRRT, continuous renal replacement therapy; ECMO, extracorporeal membrane oxygenation; Fio2, fraction of inspired oxygen; IJ, internal jugular; PC-AC, pressure control–assist control; PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressure; PMH, past medical history; RR, respiratory rate; TEE, transesophageal echocardiographic; VV, venovenous
A recent analysis of the Extracorporeal Life Support Organization registry data suggested an in-hospital 90-day mortality of 38% in patients with coronavirus 2019 (COVID-19) and acute respiratory distress syndrome who undergo extracorporeal membrane oxygenation (ECMO).1 The median duration of ECMO support in this analysis was 13.9 days, but the patients who survived to hospital day 40 had an in-hospital 90-day mortality of only 14.1%. We previously reported that patients requiring ECMO for COVID-19 pneumonia have similar outcomes as patients with influenza but significantly longer ECMO courses.2 In our experience, patients with prolonged courses of ECMO often have improved oxygenation but persistent hypercarbia. Although ventilation improves over time, patients may require several additional weeks of ECMO support before they can be weaned from ECMO. At our center (Baylor Scott & White The Heart Hospital, Plano, TX), we have used directed hypercapnia to induce subacute, metabolically compensated respiratory acidosis as a strategy to overcome long-term refractory hypercarbic respiratory failure in patients undergoing prolonged ECMO courses. In this report, we describe the outcomes of this strategy in 5 patients with COVID-19 pneumonia and respiratory failure.
Case Reports
This case series describes 5 patients with COVID-19–related acute respiratory distress syndrome who were transferred to a single ECMO referral center for initiation of venovenous (VV) ECMO after maximal conventional ventilatory strategies had failed. All patients met the Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome (EOLIA) criteria3 for emergency ECMO cannulation and subsequently required prolonged extracorporeal support secondary to refractory hypercarbic respiratory failure. We used directed hypercapnia to induce subacute, metabolically compensated respiratory acidosis as a strategy to wean and decannulate these patients from ECMO. In all cases, our directed hypercapnia protocol to induce metabolically compensated respiratory acidosis involved titrating the ECMO sweep gas for a mildly acidotic pH (7.30-7.35) while allowing the serum bicarbonate level to rise intentionally.
Patient 1
A 51-year-old man with a past medical history (PMH) of obesity who had a diagnosis of COVID-19 requiring mechanical ventilatory support was transferred to our facility and underwent emergency VV-ECMO cannulation through the right internal jugular (IJ) vein with a 32-F Crescent cannula (Medtronic) inserted under transesophageal echocardiographic (TEE) guidance.
By ECMO day 92, he no longer required oxygenation support from the ECMO circuit but experienced persistent hypercapnia. Arterial blood gas (ABG) measurements at that time, at a sweep gas of 2.0 L/min, were as follows: pH, 7.37; Paco 2, 64 mm Hg; bicarbonate, 37 mEq/L; and Pao 2, 109 mm Hg. Our directed hypercapnia protocol was initiated. On ECMO day 111, the circuit was capped. The patient tolerated a 3-day capping trial and was decannulated on ECMO day 114. ABG values were as follows: pH, 7.40; Paco 2, 75 mm Hg; bicarbonate, 46 mEq/L; and Pao 2, 79 mm Hg. Ventilator settings were as follows: pressure control–assist control (PC-AC); respiratory rate (RR), 24 breaths/min; positive end-expiratory pressure (PEEP), 8 cm H2O; peak inspiratory pressure (PIP), 29 cm H2O; and fraction of inspired oxygen (Fio 2), 50%. The patient was discharged 9 days after ECMO decannulation to a long-term acute care facility tolerating pressure support trials.
Patient 2
A 54-year-old man with a PMH of obesity, hypertension, and type 2 diabetes and a diagnosis of COVID-19 was transferred to our facility and underwent emergency VV-ECMO cannulation at the bedside through the right IJ vein with a 32-F Crescent cannula inserted under TEE guidance.
On ECMO day 75, his ABG values, at a sweep gas of 2.5 L/min, were as follows: pH, 7.39; Paco 2, 51 mm Hg; bicarbonate, 31 mEq/L; and Pao 2, 77 mm Hg. Our directed hypercapnia protocol was initiated. The patient underwent capping trials and was decannulated ECMO day 93. His ABG values before decannulation were as follows: pH, 7.42; Paco 2, 61 mm Hg; bicarbonate, 40 mEq/L; and Pao 2, 73 mm Hg. The ventilator settings before decannulation were as follows: PC-AC; RR, 16 breaths/min; PEEP, 10 cm H2O; PIP, 20 cm H2O; and Fio 2 50%. The patient was discharged 9 days after ECMO decannulation to a long-term acute care facility tolerating pressure support trials.
Patient 3
A 47-year-old man with a PMH of hypertension and obesity who had a diagnosis of COVID-19 was transferred to our facility and underwent emergency VV-ECMO cannulation at the bedside through the right IJ vein with a 30-F Crescent cannula inserted under TEE guidance.
On ECMO day 38, his ABG values, at a sweep gas of 7.5 L/min, were as follows: pH, 7.45; Paco 2, 35 mm Hg; bicarbonate; 24 mEq/L; and Pao 2, 65 mm Hg. Our directed hypercapnia protocol was initiated. On ECMO day 55, the circuit was capped, and the patient was decannulated on ECMO day 56. His ABG values before decannulation were as follows: pH, 7.4; Paco 2, 58 mm Hg; bicarbonate, 36 mEq/L; and Pao 2, 70 mm Hg. The ventilator settings before decannulation were PC-AC; RR, 16 breaths/min; PEEP, 8 cm H2O; PIP, 36 cm H2O; and Fio 2, 50%. After decannulation, patient had transiently worse hypercarbia with respiratory acidosis attributed to ventilatory desynchrony that resolved with increased sedation. The patient was transferred back to the referring hospital 11 days after decannulation on minimal ventilatory settings for continued weaning from ECMO.
Patient 4
A 58-year-old man with no PMH who had a diagnosis of COVID-19 was transferred to our facility and underwent emergency VV-ECMO cannulation through the right IJ vein with a 30-F Crescent cannula inserted under TEE guidance.
On ECMO day 69, his ABG values, at a sweep gas of 5.0 L/min, were as follows: pH, 7.42; Paco 2, 39 mm Hg;, bicarbonate, 25 mEq/L; and Pao 2, 64 mm Hg. Our directed hypercapnia protocol was initiated. On ECMO day 82, the circuit was capped, and on ECMO day 83, the patient was decannulated. His ABG values before decannulation were as follows: pH, 7.40; Paco 2, 64 mm Hg; bicarbonate, 40 mEq/L; and Pao 2, 80 mm Hg. The ventilator settings before decannulation were as follows: PC-AC; RR, 24 breaths/min; PEEP, 10 cm H2O; PIP, 32 cm H2O; and Fio 2, 50%. The patient was able to sit at the edge of bed with therapy and returned to his transferring facility 6 days after decannulation.
Patient 5
A 48-year-old male never smoker with a PMH of hypertension and obesity who had a diagnosis of COVID-19 was transferred to our facility and underwent emergency VV-ECMO cannulation through the right IJ vein with a 30-F Crescent cannula inserted under TEE guidance.
Unlike the other 4 patients, this patient experienced renal failure requiring continuous renal replacement therapy (CRRT). On ECMO day 150, his ABG values, at a sweep gas of 2.0 L/min, were as follows: pH, 7.35; Paco 2, 50 mm Hg; bicarbonate, 27 mEq/L; and Pao 2, 67 mm Hg. Our directed hypercapnia protocol was initiated. ECMO day 163 the circuit was capped, and patient was decannulated on ECMO day 167. His ABG values before decannulation were as follows: pH, 7.32; Paco 2, 75 mm Hg; bicarbonate, 39 mEq/L; and Pao 2, 77 mm Hg. The ventilator settings before decannulation were as follows: PC-AC; RR, 24 breaths/min; PEEP, 10 cm H2O; PIP, 22 cm H2O; and Fio 2 50%.
After decannulation, the patient experienced renal recovery, and CRRT was appropriately discontinued 3 days after ECMO decannulation. Initially, the patient was able to maintain hemodynamic stability and ventilation. Unfortunately, worsening hypercarbia, acidosis, and sepsis refractory to medical management suddenly developed in this patient. CRRT was restarted, but we were unable to compensate for his acidosis. Given the patient’s prolonged course, we elected not to recannulate. The family chose to stop aggressive therapies and instead opted to pursue comfort care. The patient died 7 days after decannulation.
Comment
This case series demonstrates successful weaning from VV-ECMO in 5 patients with long-term refractory hypercarbic respiratory failure. Although oxygenation requirements improved dramatically during initial ECMO support, these patients had poor CO2 clearance requiring persistent sweep gas. By weaning from sweep gas with a lower target pH on ABGs, we created directed respiratory hypercapnia that induced metabolic compensation. After several days of allowing the patients’ serum bicarbonate levels to rise, the patients were weaned from the sweep gas completely, and a steady-state elevated level of Paco 2 was achieved with a normal pH.
The 4 patients in this series with adequate renal function tolerated directed hypercapnia as a weaning strategy and were successfully decannulated and discharged from the ECMO center. All 4 patients are alive as of the submission of this paper. Patient 5, who had renal failure requiring CRRT, died of a combination of acidosis, sepsis, and respiratory failure after decannulation. The impaired ability of this patient’s native kidneys to regulate serum bicarbonate levels and intrinsically compensate for respiratory acidosis likely played a role in his failure to thrive after decannulation and discontinuation of CRRT.
This case series demonstrates successful use of directed hypercapnia to induce subacute metabolically compensated respiratory acidosis as a strategy to overcome long-term refractory hypercarbic respiratory failure in patients with COVID-19 pneumonia who require prolonged VV-ECMO with adequate renal function (Figure ). We believe this strategy may allow for earlier decannulation from ECMO that may ultimately reduce morbidity and mortality related to extracorporeal support and liberate valuable resources.
Figure.
Paco2 and serum bicarbonate before and after directed hypercapnia. The circle represents the Paco2 and bicarbonate levels before directed hypercapnia, and the arrow represents the level before decannulation. The pH at the time of decannulation was within normal limits for all patients.
References
- 1.Barbaro R.P., MacLaren G., Boonstra P.S., et al. Extracorporeal membrane oxygenation support in COVID-19: an international cohort study of the extracorporeal life support organization registry. Lancet. 2020;396:1071–1078. doi: 10.1016/S0140-6736(20)32008-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Shih E., Squiers J.J., DiMaio J.M., et al. Outcomes of extracorporeal membrane oxygenation in patients with severe acute respiratory distress syndrome caused by COVID-19 versus influenza. Ann Thorac Surg. 2022;113:1445–1451. doi: 10.1016/j.athoracsur.2021.05.060. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Combes A., Hajage D., Capellier G., et al. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. N Engl J Med. 2018;378:1965–1975. doi: 10.1056/NEJMoa1800385. [DOI] [PubMed] [Google Scholar]