892: Extracorporeal Membrane Oxygenation (Ecmo) Support for High-Risk Phosphine Intoxication

Introduction: Phosphide poisoning is highly fatal with no antidotes available to date. It causes profound, yet reversible, cardiogenic and circulatory collapse that may be refractory to conventional therapy. Veno-arterial extracorporeal membrane oxygenation (V-A ECMO) may temporarily bridge these patients to recovery with good but limited evidence. Methods: This is a tertiary-care, single-center, observational study examining critically ill phosphide intoxicated patients receiving ECMO support in Cairo University from May 2022 to August 2023. Results: We included 21 patients during the study period; 20 with aluminum phosphide and one with zinc phosphide intoxication. They had a mean age of 27.4 ± 10.52 years and no major comorbidities. Despite supportive therapy, they had a mean arterial pressure of 47.8 ± 11.7 mmHg and 57% suffered cardiac arrest before ECMO cannulation and extracorporeal cardiopulmonary resuscitation (ECPR) was deployed in 33% of the patients. Marked myocardial dysfunction dominated before ECMO with a mean left ventricular ejection fraction of 22.4% ± 14.2, which reached 49.37% ± 7.69 by decannulation and normalized on discharge. After ECMO runs lasting 93.6 ± 35.5 hours, 15 patients (71%) survived to hospital discharge. Survivors had significantly higher mean arterial pressure (p=0.01) and left ventricular contractility (p=0.003) at time of cannulation and mortality significantly correlated with ECPR, cardiac arrest duration before ECMO, and lactate resolution after support. Anemia (76%), thrombocytopenia (95%), and cannula site bleeding (33.3%) were the most reported ECMO-related complications. Septic shock, multiorgan failure, and severe neurologic dysfunction were the causes of death. Conclusions: In light of severely deranged cardiac and hemodynamic profiles of phosphine intoxicated patients, ECMO can provide a sufficient temporary support till recovery. However, ECPR and markedly impaired hemodynamics and ejection fraction at support onset may predict mortality.