Introduction: Recently, neuroprotective effects of magnesium (Mg) have been demonstrated in animals and preterm neonates. Therapeutic use for Mg in babies to prevent or ameliorate hypoxic-ischemic encephalopathy (HIE) is now under intense investigation. The purpose of this pilot study was to measure Mg levels in ECMO circuits, and in the blood of term or near term infants requiring ECMO, who are also at risk for HIE.

Methods: Retrospective review (n=27) and prospective acquisition(n=6) of total serum Mg, calcium (Ca), and ionized-Ca (iCa) concentrations were undertaken in ECMO treated infants. Circuit primes were analyzed in 9.

Results: Mg and iCa were below normal ranges in the circuit prime(mean Mg=0.8±0.1 SD mEq/L, iCa=1.2±0.3 mg/dL). Retrospective analysis revealed a small decrease in infant Mg with ECMO initiation(Mg=1.3±0.3 mEq/L <5 hour pre versus 1.1±0.2 mEq/L on ECMO, P=0.08). In prospective evaluations, a significant drop in infant iCa occurred at 5-10 minutes after the initiation of ECMO (iCa=3.7±0.7 mg/dL <2 hour pre versus 2.3±1.1 mg/dL after 5-10 minutes on bypass, P=0.03). Recovery to baseline occurred by 1 hour after initiation of bypass(3.8±0.9 mg/dL, P=0.99 versus pre). Infant Mg was below normal prior to and at 5-10 min on ECMO, and trended downwards after the initiation of ECMO(1.1±0.1 to 1.0±0.1 mEq/L, P=0.06).

Conclusions: Total Mg and iCa were below normal in most infants and in all ECMO circuits prior to the initiation of ECMO. As with Ca, Mg binds to citrate in the circuit prime and is not routinely added. Although only a mild reduction in total Mg with the initiation of bypass has so far been seen, we speculate that there may be a dramatic reduction in ionized Mg with priming and initiation of ECMO, much like iCa. Prevention of hypomagnesemia may be important in ameliorating HIE in this population of infants who are already hypomagnesemic and at risk for brain injury.