Urinary proteomics in sepsis-associated AKI

To the Editor,

We read with interest the recent article by Stanaway et al. titled “Urinary proteomics identifies distinct immunological profiles of sepsis-associated AKI sub-phenotypes” [1]. The study represents a significant advancement in understanding acute kidney injury (AKI) in sepsis through urinary proteomics, offering promising insights to improve early recognition, predict responses to therapy, and implement or develop targeted treatment strategies. However, some aspects of the study warrant further discussion.

The observation that bacterial infections were more common in AKI-SP2, whereas COVID-19 predominated in AKI-SP1, supports the phenotype of endothelial dysfunction and inflammation. However, this raises the question of whether AKI-SP2 represents a distinct AKI sub-phenotype or reflects endothelial dysfunction typically associated with bacterial sepsis. Additionally, the observed overlap of proteins associated with AKI-SP2 and those linked to the risk of renal replacement therapy (RRT) raises the question of whether AKI-SP2 represents a distinct sub-phenotype or reflects a continuum of severe AKI. Clarification of this overlap could enhance our biological understanding of these processes. Moreover, the finding that urinary proteomic profiles of AKI-SP1 were largely similar to those of non-AKI patients suggests the need to refine diagnostic thresholds or explore alternative biomarker panels to improve classification accuracy. Addressing this issue may require the use of more specific biomarkers directly associated with the pathophysiological mechanisms of AKI to improve patient phenotyping. Given the study’s emphasis on urinary proteomics, defining sub-phenotypes directly from urinary proteomic data seems feasible, potentially yielding kidney-specific classifications that more accurately reflect local injury processes, which could enable tailored therapeutic strategies.

The timing of sample collection from patients with sepsis remains a significant challenge as the onset and progression of critical symptoms can vary widely between individuals. This variability is influenced by patient-related factors, type of pathogen involved, and specific organs affected. Given the dynamic progression of sepsis, aligning sample collection more precisely with the timing of sepsis onset could reduce variability and improve data consistency; however, this remains an extremely challenging, if not impossible, task. Alternatively, patients could be aligned based on the onset of AKI as defined by the KDIGO guidelines [2] or using predictive or functional AKI biomarkers such as Cystatin C rather than ICU admission, which may provide a more clinically relevant timeline for analysis, reduce inter-patient variability, and provide a clearer picture of AKI-related changes in urinary proteomic profiles as AKI evolves. Longitudinal analysis with repeated sampling could enable the identification of temporal proteomic changes associated with the onset or recovery of AKI. However, it is important to realize that the pathways differentially expressed after AKI onset may not represent therapeutic targets to prevent AKI development in sepsis effectively. Finally, while plasma biomarkers were used to define AKI sub-phenotypes, no direct comparison was made between urinary proteomic data and systemic circulation proteomic profiles. Integrating these datasets could provide a more comprehensive understanding of local kidney-specific processes versus systemic inflammatory responses in sepsis-associated AKI.

Future studies should incorporate renal outcomes, including the duration and severity of AKI and its progression to acute kidney disease (AKD) and chronic kidney disease (CKD). This would further strengthen the claim that this biomarker panel can identify meaningful AKI phenotypes beyond acute prognostication. While the study primarily focused on in-hospital mortality and the need for renal replacement therapy (RRT), these outcomes often correlate with sepsis severity and serve as prognostic indicators rather than actionable clinical insights for long-term care. The authors highlight potential treatable targets, yet these largely align with the well-established inflammatory response and endothelial activation pathways in sepsis. From a clinical perspective, a more pressing question is how these AKI phenotypes can directly inform patient management. Moreover, identifying patients at risk of long-term renal dysfunction could guide post-ICU follow-up strategies and interventions, ultimately improving outcomes for sepsis survivors. This approach would be far more impactful than yet another predictive biomarker panel primarily aimed at predicting mortality, especially when more accessible biomarkers already exist for that purpose.

In conclusion, the study by Stanaway et al., provides valuable insights into the proteomic characterization of sepsis-associated AKI, underscoring the potential of urinary biomarkers in advancing precision medicine for sepsis. Interventional studies in preclinical models or human patients are needed to assess whether the identified pathways are potential therapeutic targets for alleviating AKI in sepsis. Addressing these points will be essential to further enhance the potential of urinary proteomics to tailor sepsis care.