Subphenotypes of Tumor Lysis Syndrome with AKI

Background: The diverse clinical outcomes observed in tumor lysis syndrome (TLS) patients with acute kidney injury (AKI) suggest underlying heterogeneity within TLS. We aimed to elucidate the subphenotypes of TLS with latent class analysis (LCA) to better understand their clinical characteristics and outcomes. Methods: A cohort of 1,377 patients presenting with TLS-related AKI between 2005 and 2019 was enrolled. Analysis of total forty clinical variables, including baseline characteristics, malignancy type, metastatic site, AKI stage, and laboratory measurements, was conducted using LCA to identify distinct subphenotypes. All-cause death, cancer-specific death, non-recovery of renal function, and progression to dialysis requiring AKI, were compared among the identified subphenotypes. Results: LCA revealed four distinct subphenotypes of TLS patients. Subphenotype 1 exhibited the lowest all-cause mortality (hazard ratio [HR] 0.27, 95% confidence interval [CI] 0.22-0.34), while subphenotype 4 had the highest mortality rate. Similar trend was observed in cancer-specific mortality with subphenotype 1 displaying the lowest risk (HR 0.28, 95% CI 0.22-0.35). Additionally, subphenotype 1 exhibited significantly lower risks of non-recovery of renal function (OR 0.30, 95% CI 0.19-0.47) and progression to dialysis-requiring AKI (OR 0.36, 95% CI 0.17-0.77) compared to the other subphenotypes. Conclusion: Our study identifies four novel subphenotypes, with subphenotype 1 associated with the most favorable outcomes, while subphenotype 4 exhibits the worst prognosis. This subphenotyping approach holds promise in guiding future research.Heat map of variable by subphenotypesSubgroup analyses of clinical outcomes of subphenotypes.