Sepsis-associated acute kidney injury (SA-AKI) is common in patients with sepsis and is linked to poor outcomes. Conventional markers such as serum creatinine and urine output have limited sensitivity for early assessment. This retrospective study examined whether admission levels of Klotho, fibroblast growth factor 23 (FGF23), and the Klotho/FGF23 ratio were associated with SA-AKI and short-term clinical and renal outcomes, and compared the discriminative performance of these biomarkers.
The investigation was a retrospective cohort study of 320 patients admitted to the intensive care unit (ICU) who met Sepsis-3 criteria for sepsis. SA-AKI was identified retrospectively in 112 patients according to Kidney Disease: Improving Global Outcomes (KDIGO) criteria.
Stored serum samples collected within 24 hours of ICU admission were analyzed. Klotho and FGF23 concentrations were measured by enzyme-linked immunosorbent assay (ELISA). Discrimination of biomarkers for SA-AKI was evaluated using receiver operating characteristic (ROC) analysis, and adjusted associations were assessed with multivariable logistic and linear regression models. The abstract does not list the specific covariates included in adjusted models or the absolute biomarker concentrations.
At ICU admission, patients who subsequently met criteria for SA-AKI had statistically lower serum Klotho, higher FGF23, and a reduced Klotho/FGF23 ratio compared with patients who did not develop SA-AKI (all p < 0.001). These differences indicate an inverse relationship between Klotho and FGF23 in the context of sepsis-related kidney injury in this cohort.
Among the three tested biomarkers, the Klotho/FGF23 ratio demonstrated the highest observed discrimination for SA-AKI. The area under the ROC curve (AUC) for the ratio was 0.84 with a 95% confidence interval of 0.79 to 0.88. The abstract reports this AUC value but does not provide AUCs for Klotho or FGF23 individually, nor does it report optimal cutoffs, sensitivity, or specificity values.
After multivariable adjustment, the admission Klotho/FGF23 ratio remained independently associated with SA-AKI. The adjusted odds ratio reported for the ratio was 2.25 (95% CI 1.46–3.53). The ratio also decreased progressively with increasing KDIGO stage, suggesting a relationship between lower ratio values and greater SA-AKI severity. The abstract does not specify which confounders were adjusted for in the multivariable models.
Within the subgroup of patients who developed SA-AKI, a low admission Klotho/FGF23 ratio was associated with several adverse short-term outcomes after adjustment. These included lower likelihood of renal recovery at 3 months, longer ICU and hospital lengths of stay, higher in-hospital mortality, and greater progression to chronic kidney disease at 3 months. The abstract reports these associations but does not supply absolute event rates, adjusted effect sizes for each outcome, or the full list of covariates used in the outcome models.
The PubMed abstract reports study design and key findings but omits several details typically necessary for full appraisal: the specific covariates used in adjusted analyses, the absolute biomarker concentrations and distributions, sensitivity and specificity or chosen cut points for discrimination, and detailed patient characteristics. As this summary is based solely on the abstract, additional methodological details and full results would need to be obtained from the full text for comprehensive evaluation.
In this retrospective ICU cohort of 320 sepsis patients (112 with SA-AKI), the admission Klotho/FGF23 ratio was associated with the presence of SA-AKI, higher KDIGO stage, and worse short-term renal and clinical outcomes. The ratio showed higher observed discrimination for SA-AKI (AUC 0.84; 95% CI 0.79–0.88) than either biomarker alone in the data presented in the abstract, and it remained independently associated with SA-AKI after multivariable adjustment (adjusted OR 2.25; 95% CI 1.46–3.53). The abstract indicates potential clinical relevance of the ratio for early risk stratification in sepsis but leaves several methodological details unreported in the abstract text.