Dapagliflozin for Critically Ill Patients With Acute Organ Dysfunction: The DEFENDER Randomized Clinical Trial

Abstract

Importance: Sodium-glucose cotransporter 2 (SGLT-2) inhibitors improve outcomes in patients with type 2 diabetes, heart failure, and chronic kidney disease, but their effect on outcomes of critically ill patients with organ failure is unknown.

Objective: To determine whether the addition of dapagliflozin, an SGLT-2 inhibitor, to standard intensive care unit (ICU) care improves outcomes in a critically ill population with acute organ dysfunction.

Design, setting, and participants: Multicenter, randomized, open-label, clinical trial conducted at 22 ICUs in Brazil. Participants with unplanned ICU admission and presenting with at least 1 organ dysfunction (respiratory, cardiovascular, or kidney) were enrolled between November 22, 2022, and August 30, 2023, with follow-up through September 27, 2023.

Intervention: Participants were randomized to 10 mg of dapagliflozin (intervention, n = 248) plus standard care or to standard care alone (control, n = 259) for up to 14 days or until ICU discharge, whichever occurred first.

Main outcomes and measures: The primary outcome was a hierarchical composite of hospital mortality, initiation of kidney replacement therapy, and ICU length of stay through 28 days, analyzed using the win ratio method. Secondary outcomes included the individual components of the hierarchical outcome, duration of organ support-free days, ICU, and hospital stay, assessed using bayesian regression models.

Results: Among 507 randomized participants (mean age, 63.9 [SD, 15] years; 46.9%, women), 39.6% had an ICU admission due to suspected infection. The median time from ICU admission to randomization was 1 day (IQR, 0-1). The win ratio for dapagliflozin for the primary outcome was 1.01 (95% CI, 0.90 to 1.13; P = .89). Among all secondary outcomes, the highest probability of benefit found was 0.90 for dapagliflozin regarding use of kidney replacement therapy among 27 patients (10.9%) in the dapagliflozin group vs 39 (15.1%) in the control group.

Conclusion and relevance: The addition of dapagliflozin to standard care for critically ill patients and acute organ dysfunction did not improve clinical outcomes; however, confidence intervals were wide and could not exclude relevant benefits or harms for dapagliflozin.

Trial registration: ClinicalTrials.gov Identifier: NCT05558098.

Figure 1.. Flow of Participants Through the DEFENDER Randomized Clinical Trial

^aInformation was obtained via screening logs, reporting the single criterion that did not meet eligibility. ICU indicates intensive care unit; SGLT-2, sodium-glucose cotransporter 2.

Figure 2.. Win Ratio Analysis for the Primary Outcome

Distribution of wins, ties, and losses for the dapagliflozin group among the 64 232 paired comparisons, stratified by each level of the hierarchical primary composite outcome. Every possible pair of participants between groups was compared in a hierarchical fashion with a win, a loss, or a tie determined by the outcome evaluated at each level of the hierarchy. Early ties were determined when both participants in the pair died during hospitalization. Percentages are calculated for each level of the hierarchy. The win ratio equals the total wins for the dapagliflozin group divided by the total losses for the dapagliflozin group: 27 143/26 929 = 1.01 (95% CI, 0.90-1.13; P = .89). ICU indicates intensive care unit.

Figure 3.. Primary Outcome in the Prespecified Subgroup Analyses

^aDetermined by physician’s assessment. ^bBaseline serum creatinine levels. To convert creatinine from mg/dL to μmol/L, multiply by 88.4. ^cAccording to the Simplified Acute Physiology Score 3 subgroup. Shown is the win ratio for the composite hierarchical primary outcome of hospital mortality, initiation of kidney replacement therapy, and intensive care unit (ICU) length of stay stratified for prespecified subgroups. A win ratio greater than 1.0 indicates a favorable effect for the dapagliflozin group. The width of point estimates are scaled according to the number of participants in each subgroup. The 95% CIs were not adjusted for multiple comparisons and should not be used to infer treatment effects.