Prolonged extracorporeal membrane oxygenation in pediatric leukemia with severe acute respiratory distress syndrome and persistent fungemia

Abstract

To the Editor, We read with great interest the articles by Steppan et al.1 and Peta et al.2 on extracorporeal membrane oxygenation (ECMO) in children with cancer. Steppan et al. reported that 18 children with cancer supported on ECMO had similar outcomes as nononcologic ECMO controls and that 61% survived to hospital discharge.1 Given the improved survival of childrenwith cancer, data suggest that ECMOcan be considered in this population.3–13 However, prolonged use (>28 days) of ECMO in children with refractory cardiac/respiratory failure has been associated with low survival and significant long-term neurologic and respiratory sequelae.14,15 We report a 7-year-old boy with acute lymphoblastic lymphoma (ALL) who was admitted to the intensive care unit (ICU) with septic shock and candidemia. He recovered after being on ECMO for a prolonged period (55 days) due to severe acute respiratory distress syndrome (ARDS) during his 147 days in the ICU. The patient developed neutropenic fever after induction chemotherapy. Empirical antibiotic therapy was initiated. Due to clinical deterioration, antibiotic therapy was modified and liposomal amphotericin B added. He was transferred to the ICU with septic shock and was intubated. Candida tropicalis was detected in blood culture. Caspofungin was added. The port-a-catheter was removed. Antimicrobials were modified according to clinical status and microbiological results. Despite aggressive supportive care, he continued to deterioriate. Granulocyte transfusions were given 11 times. Chest tubes were placed multiple times due to pneumothorax. Due to refractory hypoxia, secondary to severe ARDS, he was put on venovenous ECMO on the 29th day in ICU. Candida parapsilosis grew on blood culture during ECMO. Micafungin was added due to its inhibitory effect on Candida biofilms. Excluding the cannulae, the ECMO circuit was changed without any effect on fungemia. There was no possibility to stop ECMO. Antifungals were continued, speculating that they would at least decrease/control the growth of the fungi. He had cytomegalovirus reactivation and hemophagocytic syndrome, responding to gancyclovir and intravenous immunoglobulin. Acinetobacter baumanii was identified in his tracheal aspirate. After 55 days of ECMO, he was decannulated and the fungemia ceased. He had a respiratory deterioration on the 95th day of his ICU stay, was diagnosed with pulmonary hypertension and treated with iloprost and sildenafil. He had a tracheostomy. He was weaned off the mechanical ventilator to oxygen support via nasal cannula after 145 days in ICU, as his lung parenchyma restored. Bone marrow was in remission. Chemotherapy was continued. He is alive in remission at 4 years from the end of treatment. He attends school with a mobile oxygen device and has not needed continuous oxygen support for the past year. This case represents a pediatric cancer with the longest duration of ECMO in the literature. Sustained fungemia/bacteremia during an ECMOrun should not be regarded as a reason towithdrawextracorporeal support, although efforts are clearly warranted towean off ECMO at the earliest opportunity. Children with leukemia or other cancers should be candidates for ECMO in severe respiratory/cardiac failure.