Identification of Trajectory-Based Acute Kidney Injury Phenotypes Among Cardiac Surgery Patients

Area of Science:

• Nephrology
• Cardiology
• Data Science

Background:

• Acute kidney injury (AKI) is a frequent and severe complication following cardiac surgery.
• Heterogeneity in AKI patient groups may explain the limited success of current therapeutic strategies.
• Understanding AKI patient subgroups is crucial for developing targeted renoprotective interventions.

Purpose of the Study:

• To identify distinct patient phenotypes based on postoperative serum creatinine trajectories after cardiac surgery.
• To group patients with similar AKI patterns for improved clinical trial design and analysis.
• To investigate the association between identified AKI phenotypes and long-term patient outcomes.

Main Methods:

• Retrospective, single-center cohort study of patients undergoing coronary artery bypass graft (CABG) procedures.
• Latent class mixed modeling applied to serum creatinine values to define AKI trajectory phenotypes.
• Development and validation cohorts used to ensure consistency and stability of identified AKI classes.

Main Results:

• Twelve distinct AKI trajectory phenotypes were identified in both development and validation cohorts.
• Phenotypes varied in the rate, timing, and degree of serum creatinine changes and recovery.
• Four identified high-risk AKI phenotypes were associated with a significantly greater long-term risk of mortality.

Conclusions:

• Latent class mixed modeling successfully identified 12 reproducible AKI phenotypes in CABG patients.
• Four AKI phenotypes represent high-risk groups with poor long-term prognosis.
• These findings provide a novel framework for patient stratification in renoprotection research and reanalysis of past trials.