Acute kidney injury (AKI) occurs in up to 50% of patients undergoing cardiac surgery and is associated with an increased long-term risk of chronic kidney disease and mortality. Its pathophysiology is multifactorial, involving renal hypoperfusion and hypoxia, ischemia-reperfusion injury, oxidative stress, and inflammation. Until recently, no pharmacological intervention had been shown to effectively reduce its incidence after cardiac surgery. Sodium-glucose cotransporter 2 (SGLT2) inhibitors have demonstrated renoprotective effects, reducing the progression of chronic kidney disease and the incidence of AKI by decreasing proximal tubular sodium and glucose reabsorption, glomerular pressure, and renal workload while preserving kidney perfusion. These agents may also attenuate renal hypoxia, ischemia-reperfusion injury, and inflammation. Preliminary studies with empagliflozin had suggested a substantial reduction in AKI following cardiac surgery. The MERCURI-2 trial was a multicenter, randomized, double-blind, placebo-controlled superiority trial conducted at 7 centers in the Netherlands between June 2023 and January 2025. Adult patients scheduled for any type of elective cardiac surgery were eligible. Exclusion criteria included type 1 diabetes, type 2 diabetes requiring intensive insulin therapy, baseline estimated glomerular filtration rate (eGFR) below 20 mL/min/1.73 m², a history of diabetic ketoacidosis, systolic blood pressure below 100 mm Hg, previous treatment with SGLT2 inhibitors, known allergy to these drugs, pregnancy, breastfeeding, and certain circumstances related to childbearing potential. Patients were randomly assigned in a 1:1 ratio to receive dapagliflozin 10 mg daily for 4 doses, starting the day before surgery, or placebo. Randomization was computer-generated, double-blinded, and stratified according to sex, type 2 diabetes status, and participating center.

The primary outcome was the incidence of AKI during the first 7 postoperative days, defined according to Kidney Disease: Improving Global Outcomes (KDIGO) criteria. Secondary outcomes included the incidence of stage 3 AKI, maximum postoperative change in eGFR, new-onset atrial fibrillation, intensive care unit and hospital length of stay, major adverse cardiac events (MACE30) and major adverse kidney events (MAKE30), disability assessed using the World Health Organization Disability Assessment Schedule 2.0 (WHO-DAS 2.0), days at home during the first 30 postoperative days (DAH30), and quality of life measured using the EQ-5D-3L questionnaire. Genital mycotic infections, ketoacidosis, and hypoglycemia were also evaluated as safety outcomes.

A total of 784 patients were randomized, with 392 allocated to each group. Baseline characteristics were comparable between groups. Approximately 60% underwent a single procedure, either isolated coronary artery bypass grafting (CABG) or another non-CABG procedure, while 40% underwent combined procedures. All 4 study doses were completed by approximately 88% of patients receiving dapagliflozin and 87% of those receiving placebo. Dapagliflozin significantly reduced the incidence of AKI compared with placebo (RR = 0.54, p < .001), corresponding to an absolute risk reduction of 24 percentage points. Among patients who completed all 4 doses (per-protocol analysis), the effect remained similar (RR = 0.52). Regarding secondary outcomes, stage 3 AKI was uncommon, with no significant difference between groups (dapagliflozin 0.8% vs placebo 0.3%, p = nonsignificant). No statistically significant differences were observed in the remaining secondary outcomes. After adjustment for sex, dapagliflozin remained significantly associated with a lower incidence of AKI (OR = 0.36, p < .001), with no evidence of treatment-effect heterogeneity between men and women.

The reduction in AKI with dapagliflozin was primarily observed in mild and moderate stages. Furthermore, when AKI was defined exclusively according to serum creatinine criteria, its incidence was very similar between groups. In contrast, when urine output criteria alone were applied, the difference was considerably greater (21% with dapagliflozin vs 48% with placebo). This suggests that much of the observed reduction in AKI was attributable to a lower incidence of postoperative oliguria.

Adverse events typically associated with SGLT2 inhibitors were uncommon. In the dapagliflozin group, there was one case each of genital mycotic infection, ketoacidosis, and hypoglycemia. None of these events occurred in the placebo group.

In their discussion, the investigators identified the reduction in AKI following elective cardiac surgery as the principal finding of perioperative dapagliflozin administration, although no significant improvements were demonstrated in the other secondary outcomes. These findings are consistent with previous studies of empagliflozin, which also reported a reduction in postoperative AKI. The incidence of AKI in the placebo group, based on the combined urine output and serum creatinine criteria, was relatively high at approximately 52%, exceeding rates commonly reported in the literature. When serum creatinine criteria alone were considered, AKI occurred in approximately 15% of patients in each group.

The main strengths of the study include its relatively large sample size and its multicenter, randomized, double-blind, placebo-controlled design. Nevertheless, several limitations acknowledged by the authors deserve particular attention. The primary outcome was assessed only during the first 7 postoperative days, and the trial was insufficiently powered to adequately evaluate stage 3 AKI or composite kidney outcomes. Its findings are directly applicable only to elective cardiac surgery, and the interval between baseline serum creatinine measurement and surgery was not consistently documented. Moreover, excluding patients already receiving SGLT2 inhibitors restricted the representation of individuals with heart failure, type 2 diabetes, and chronic kidney disease. Consequently, the study population was probably healthier than that encountered in routine clinical practice, limiting the applicability of these findings to more complex patients.

In conclusion, among patients undergoing elective cardiac surgery, probably representing a relatively low-risk population, 4 doses of dapagliflozin initiated the day before surgery significantly reduced the incidence of AKI during the first 7 postoperative days. However, this reduction was not accompanied by demonstrable improvements in other clinical outcomes, and a substantial proportion of the observed benefit was related to fewer episodes of oliguria.

COMMENTARY:

This study examines the effects of perioperative dapagliflozin administration (4 doses beginning the day before surgery) in patients who were SGLT2 inhibitor–naive. Two aspects of the findings are particularly striking. First, the remarkably low incidence of the complications that currently justify withholding these agents before elective surgery, particularly ketoacidosis. Prospective nonrandomized studies have identified a strong association between preoperative treatment with SGLT2 inhibitors and euglycemic ketoacidosis. Discontinuing treatment 3 days before surgery appears to prevent this complication without eliminating the renoprotective effect of empagliflozin, as observed in the POST-CABGDM trial. In MERCURI-2, only one patient assigned to dapagliflozin developed ketoacidosis. This raises the question of whether such a restrictive approach is always necessary, particularly since no large randomized trial has investigated the consequences for renal function of abruptly interrupting chronic SGLT2 inhibitor therapy before cardiac surgery. This represents an important avenue for future research. These patients were probably carefully monitored to prevent conditions that could precipitate ketoacidosis, although hypovolemia, systemic hypoperfusion, low cardiac output, and similar complications are common after cardiac surgery and cannot always be completely avoided.

The second striking finding is the substantial reduction in postoperative AKI among patients receiving dapagliflozin, although this difference was driven mainly by urine output criteria. No significant differences were observed when AKI was defined solely by serum creatinine, nor in the incidence of stage 3 AKI, the most severe category of renal injury. The investigators argue that the apparent renal protection is unlikely to be explained simply by a diuretic effect of SGLT2 inhibitors, since previous studies suggest that any increase in urine output is modest and largely transient. Unfortunately, despite the well-established association between AKI and adverse postoperative outcomes, including increased morbidity and mortality, prolonged hospitalization, and impaired quality of life, none of these clinical outcomes improved significantly with dapagliflozin compared with placebo, despite the relatively large study population. This raises doubts about the magnitude of the actual net clinical benefit. Similar findings were reported in the POST-CABGDM trial, in which empagliflozin significantly reduced AKI according to Acute Kidney Injury Network (AKIN) criteria despite being discontinued 3 days before surgery. However, that trial randomized only 145 patients, providing insufficient statistical power to detect differences in mortality in a population undergoing isolated coronary surgery.

Another important limitation, also acknowledged by the investigators, concerns the substantial selection bias introduced by excluding patients already receiving SGLT2 inhibitors. In contemporary practice, an increasing proportion of patients undergoing cardiac surgery are prescribed these drugs for type 2 diabetes, heart failure, chronic kidney disease, or other indications. This makes it even more important to determine whether preoperative discontinuation is truly necessary and, if so, in which patients. It also raises the possibility that the renal benefits of these agents could be enhanced by continuing treatment until the immediate preoperative period in selected patients with a lower risk of ketoacidosis but a high risk of postoperative renal dysfunction.

In more complex patients, who frequently present with comorbidities underrepresented in this trial and face a greater risk of postoperative morbidity and mortality, the renoprotective effects of SGLT2 inhibitors might hypothetically be even greater and could potentially translate into improved clinical outcomes. If this hypothesis were confirmed, continuing SGLT2 inhibitor therapy in selected patients, together with close postoperative surveillance for potential adverse effects to ensure early recognition and treatment, could become a reasonable alternative to routinely discontinuing these drugs in every patient.

REFERENCE:

Ginestal-Calvo J, Flordelís-Lasierra JL, Pérez-Vela JL, et al. Postoperative ketoacidosis in cardiac surgery: incidence, analytical profile, and association with sodium-glucose cotransporter 2 inhibitors-a prospective observational study. J Cardiothorac Vasc Anesth. 2026;40(2):539-545. doi:10.1053/j.jvca.2025.11.014.

Pitta FG, Lima EG, Tavares CAM, et al. Empagliflozin in patients with type 2 diabetes undergoing on-pump CABG: the POST-CABGDM randomized clinical trial. Diabetes Care. 2025;48(6):988-995. doi:10.2337/dc24-2807.