Sustained low-efficiency dialysis for metformin-associated lactic acidosis in patients with acute kidney injury.

Abstract

The choice of the specific modality and treatment duration of renal replacement therapy (RRT) to adopt in metformin-associated lactic acidosis (MALA) is still debated. We aimed to verify if sustained low-efficiency dialysis (SLED) is a rational choice in patients with MALA and acute kidney injury (AKI). We collected serial serum metformin measurements, clinical parameters, and outcome data in ten consecutive patients (mean age 77years [range 58-88], 5 males) admitted to our renal intensive care unit for suspected MALA associated with AKI and hemodynamic instability. Patients underwent a 16-h SLED session performed with either conventional dialysis machines or machines for continuous RRT (CRRT). A 2-compartment open-infusion pharmacokinetic model with first-order elimination was fitted to each subject's serum concentration-time data to model post-SLED rebound and predict the need for further treatments. Two patients died within 24h after SLED start. Three patients needed one further dialysis session. Surviving patients (n = 8) were dialysis-free at discharge. Metformin levels were in the toxic range at baseline (median [range] 32.5mg/l [13.6-75.6]) and decreased rapidly by the end of SLED (8.1mg/l [4.5-15.8], p < 0.001 vs. baseline), without differences according to the dialysis machine used (p = 0.84). We observed a slight 4-h post-SLED rebound (9.7mg/l [3.5-22.0]), which could be predicted by our pharmacokinetic model. Accordingly, we predicted that the majority of patients would need one additional dialysis session performed the following day to restore safe metformin levels. A 16-h SLED session, performed with either conventional dialysis machines or CRRT machines, allows effective metformin removal in patients with MALA and AKI. However, due to possible post-SLED rebound in serum metformin levels, one additional dialysis treatment is required the following day in the majority of patients.