#2589 Predictive value of serum lactate in ICU dialysis-dependent AKI patients

Abstract Background and Aims Hyperlactatemia is frequently observed in critically ill patients. It is often associated with unfavorable outcomes and serves as a strong predictor of mortality. Renal replacement therapy aims to mitigate lactate accumulation in this particular subgroup of patients, but the effectiveness of this therapy in reducing morbidity and mortality in acute kidney injury (AKI) associated lactic acidosis remains inconclusive. The aim of this retrospective study was to evaluate the predictive role of serum lactate in dialysis-dependent AKI patients stratified on three different underlying etiologies (septic shock, concomitant liver failure and metformin-associated lactic acidosis). Method A retrospective, single-center cohort analysis was conducted to examine patients with AKI who underwent hemodialysis in our ICU between 2022 and 2023 and had a serum lactate > 4 mmol/L. Patients were stratified according to underlying etiology: isolated septic shock with AKI, shock with concomitant AKI and fulminant liver failure (serum ammonia > 100 µmol/L), and AKI with metformin-associated lactic acidosis (MALA). We compared patients according to age and 3-day, 30-day, and 90-day mortality. Finally, we performed receiver-operating-curve (ROC) analysis to potential predictive power of serum lactate for 30-day mortality. Results The median age of our cohort of 63 patients was 67 (range 52 - 74) years, with a median lactate level of 12.9 (7.6 – 18.0) mmol/L. The 3-day, 30-day and 90-day mortality rate was 57%, 78%, and 83%, respectively. We observed that patients with concomitant liver failure were significantly younger (p = 0.004). Interestingly, while all groups have comparable 90-day mortality, patients with septic shock exhibit extremely high 3-day mortality (74%), compared to 35% and 25% for acute liver failure and metformin-associated shock. ROC curve analysis showed that serum lactate was highly predictive of 30-day mortality in group of AKI with concomitant fulminant liver dysfunction. On the contrary, in metformin-associated hyperlactatemia the level of serum lactate showed no such prediction. Conclusion Our retrospective analysis underscores the crucial importance of differentiating between various groups of patients (septic shock, concomitant liver failure and metformin-associated lactic acidosis) in patients with hyperlactatemia. In metformin-induced kidney failure, the level of lactate does not predict mortality, while in the cases of AKI with concurrent acute liver dysfunction our results are highly suggestive that dialysis may not be effective for severe lactacidosis. The findings contribute valuable insights for designing future prospective studies in this complex patient population.