ECMO and endogenous carboxyhemoglobin formation

Sir,

We would like to report a succinct observation regarding an extremely high (poisonous) carboxyhemoglobin (COhb) level in a critically ill patient on extracorporeal membrane oxygenation (ECMO). A 41-year-old, 115 kg, 188 cm male presented to our facility with shortness of breath and hypoxemia necessitating placement of an endotracheal tube for mechanical ventilation. A diagnosis of H1N1 influenza pneumonia was made via polymerase chain reaction. Seventy-two hours later, his worsening hypoxemia (PaO₂ of 41 mmHg) required veno-venous ECMO with inhaled nitric oxide (iNO) as an adjunct treatment. Additionally, the patient became septic and went into renal failure.

On day 21 of ECMO, his COhb level spiked to 11.6%; levels of 0.8–2.0% are usually found in critically ill non-ECMO patients.[1] Carbon monoxide (CO), in effect, acts as a toxin that binds hemoglobin readily, more so than oxygen. Thus, there is less oxygen transported and ultimately less released to the tissues when COhb levels increase. In ECMO, hemolysis is frequently a problem, especially secondary to disruption of the hemoglobin molecule by mechanical trauma from the ECMO cannulas.[2] Massive hemolysis was confirmed by lactate dehydrogenase levels (1612–5475 U/L), and free hemoglobin levels (170–2220 g/dL). When hemoglobin is degraded, free iron, biliverdine, and CO are produced by heme oxygenase enzymes. Heme oxygenase enzymes come in three variations: inducible HO-1, the constitutive form HO-2, and the oxygen-sensing form HO-3 (not catalytically active). HO-1 can be induced by heme, and/or any oxidative stress, including hypoxia, hyperoxia, cytokines (sepsis), heavy metals, and acute lung injury.[1,3] Heme oxygenase can also be induced by iNO;[3] this leads to the production of methemeglobin with heme release (substrate). Our patient had treatment with iNO at 20 ppm for 21 days, and this may have contributed to the increased levels. However, the iNO was discontinued and this resulted in only a minimal decrease in subsequent COhb levels. Other sources of elevated COhb include clots in the ECMO circuit which may perpetuate hemolysis (there were none in this case), and blood transfusions (which were in a considerable number in this patient).[4,5] It should be noted that there were no difficulties with the pressures across the Quadrox^; oxygenator membrane (inlet or outlet), and the oxygen blender FiO₂ was 100% throughout the course of ECMO. While there were multiple possible contributory considerations regarding the elevated COhb levels, the most likely was assumed to be mechanical trauma secondary to the cannulas, as mentioned above. An attempt was made to correct this problem with cannula positions adjustments and decreased cannula flows. However, these maneuvers produced no abatement of the massive hemolysis, nor did they correct the elevated COhb levels. Ultimately, the patient succumbed to H1N1 pneumonia.

This scenario of hemolysis and elevated COhb levels is concerning for ECMO patients because of the already limited supply of oxygen to the tissues. We encourage our colleagues to be vigilant in regard to an elevated COhb level during ECMO, its differential diagnosis, and the appropriate interventions.