Distribution of Acute and Chronic Kidney Disease Across Clinical Phenotypes for Sepsis

Luca Molinari; Gaspar Del Rio-Pertuz; Priyanka Priyanka; Ali Smith; Joseph C Maggiore; Jason Kennedy; Hernando Gomez; Christopher W Seymour; John A Kellum; ProCESS and ProGReSS-AKI Investigators

doi:10.1016/j.chest.2024.03.006

Distribution of Acute and Chronic Kidney Disease Across Clinical Phenotypes for Sepsis

Chest. 2024 Mar 8:S0012-3692(24)00296-4. doi: 10.1016/j.chest.2024.03.006. Online ahead of print.

Authors

Luca Molinari¹, Gaspar Del Rio-Pertuz², Priyanka Priyanka³, Ali Smith⁴, Joseph C Maggiore⁵, Jason Kennedy⁴, Hernando Gomez⁶, Christopher W Seymour⁴, John A Kellum⁷; ProCESS and ProGReSS-AKI Investigators

Affiliations

¹ Department of Critical Care Medicine, University of Pittsburgh, Pittsburgh, PA; Center for Critical Care Nephrology, University of Pittsburgh, Pittsburgh, PA; Department of Translational Medicine, Università degli Studi del Piemonte Orientale, Novara, Italy.
² Center for Critical Care Nephrology, University of Pittsburgh, Pittsburgh, PA; Division of Cardiology, Department of Medicine, University of Minnesota, Minneapolis, MN.
³ Center for Critical Care Nephrology, University of Pittsburgh, Pittsburgh, PA.
⁴ Department of Critical Care Medicine, University of Pittsburgh, Pittsburgh, PA.
⁵ Department of Developmental Biology, University of Pittsburgh, Pittsburgh, PA.
⁶ Department of Critical Care Medicine, University of Pittsburgh, Pittsburgh, PA; Center for Critical Care Nephrology, University of Pittsburgh, Pittsburgh, PA.
⁷ Department of Critical Care Medicine, University of Pittsburgh, Pittsburgh, PA; Center for Critical Care Nephrology, University of Pittsburgh, Pittsburgh, PA. Electronic address: kellum@pitt.edu.

PMID: 38462074
DOI: 10.1016/j.chest.2024.03.006

Abstract

Background: Sepsis is the most common cause of acute kidney injury (AKI) in critically ill patients. Four phenotypes (α, β, γ, δ) for sepsis, which have different outcomes and responses to treatment, were described using routine clinical data in the electronic health record.

Research question: Do the frequencies of AKI, acute kidney disease (AKD), chronic kidney disease (CKD), and AKI on CKD differ by sepsis phenotype?

Study design and methods: This was a secondary analysis of a randomized clinical trial of early resuscitation, including patients with septic shock at 31 sites. After excluding patients with end-stage kidney disease and missing data, we determined frequencies of the following clinical outcomes: AKI (defined within 24 h as Kidney Disease: Improving Global Outcomes stages 2 or 3 or stage 1 with tissue inhibitor of metalloproteinases-2 × insulin-like growth factor binding protein 7 value of > 2.0), CKD, and AKD (persistence of AKI at any stage on day 7 after enrollment) across four phenotypes. We performed multivariable logistic regression to assess the risk-adjusted association between development of AKI and AKD and phenotype.

Results: Among 1,090 eligible patients, 543 patients (50%) had AKI. Across phenotypes, the frequencies of AKI varied, being highest in the δ and β phenotypes (78% and 71%, respectively) and the lowest in the α phenotype (26%; P < .001). AKD occurred most often in the δ phenotype (41%) and least often in the α phenotype (8%; P < .001). The highest frequencies of CKD and of AKI on CKD were found in the β phenotype (53% and 38% respectively; P < .001 for both). In the multivariable logistic regression models (α phenotype as reference), δ phenotype showed the strongest association with AKI (OR, 12.33; 95% CI, 7.81-19.47; P < .001) and AKD (OR, 9.18; 95% CI, 5.44-15.51; P < .001).

Interpretation: The rates of AKI and AKD differed across clinical sepsis phenotypes and are more common among patients with phenotypes β and δ. Phenotype β showed a higher level of underlying CKD that predisposed patients to new AKI. The α and γ phenotypes showed lower frequencies of AKI and less progression to AKD.

Keywords: acute kidney disease; acute kidney injury; chronic kidney disease; phenotypes; sepsis.