The Impact of Regional Citrate Anticoagulation on Magnesium Replacement During CRRT

A patient admitted to the ICU with shock and acute kidney injury required continuous renal replacement therapy (CRRT). CRRT was initiated using regional citrate anticoagulation (RCA) with an initial magnesium (Mg) level of 1.7 mg/dL. Over 12 days the patient received 68 g of Mg sulfate. After 58 g the patient's Mg level was 1.4 mg/dL. On day 13, CRRT was changed to a heparin circuit from concerns of citrate toxicity. Over the next 7 days the patient required no Mg replacement with a mean Mg level of 2.22. This was significantly higher than the final 7 days on RCA (1.99; P = .00069). This case illustrates the challenges in maintaining Mg stores during CRRT. RCA is now the preferred method of circuit anticoagulation, with prolonged filter life and fewer bleeding complication compared to heparin circuits. Citrate inhibits coagulation within the circuit by chelating ionized calcium (Ca2+). Free Ca2+ and Ca-citrate complexes diffuse across the hemofilter with a percentual calcium loss as high as 70%, requiring continuous post-filter infusions of calcium to prevent systemic hypocalcemia. Magnesium loss during CRRT is also significant and may approach 15% to 20% of the total body pool within a week. Citrate chelates Mg with percentual losses comparable to calcium. Twenty-two CRRT patients on RCA had median losses >6 g/day. Doubling the Mg content in the dialyzate of 45 CRRT patients significantly improved Mg balance, but with the potential risk of increased citrate toxicity. A major obstacle to replacing Magnesium loss with the same precision as calcium is few hospitals can measure ionized Mg++ levels and must rely on total magnesium levels to guide replacement, despite a literature showing poor correlation with total body stores. Post-circuit continuous replacement of magnesium, as with calcium, in the absence of ionized magnesium levels would likely be very inexact and arduous. Being aware of the losses that can occur with CRRT, especially with RCA, and adjusting magnesium replacement empirically on rounds may be the only pragmatic action plan for this clinical issue.