Acute kidney injury associated to COVID-19 leads to a strong unbalance of circulant immune mediators

Abstract

Background: Severe cases of coronavirus disease 2019 (COVID-19) have increased risk for acute kidney injury (AKI). The exacerbation of the immune response seems to contribute to AKI development, but the immunopathological process is not completely understood.

Objectives: To analyze levels of circulant immune mediators in COVID-19 patients evolving with or without AKI. We have also investigated possible associations of these mediators with viral load and clinical outcomes.

Methods: This is a longitudinal study performed with hospitalized patients with moderate to severe COVID-19. Serum levels of 27 immune mediators were measured by a multiplex immunoassay. Data were analyzed at two timepoints during the follow-up: within the first 13 days of the disease onset (early sample) and from the 14th day to death or hospital discharge (follow-up sample).

Results: We studied 82 COVID-19 patients (59.5 ± 17.5 years, 54.9% male). Of these, 34 (41.5%) developed AKI. These patients presented higher SARS-CoV-2 viral load (P = 0.03), higher frequency of diabetes (P = 0.01) and death (P = 0.0004). Overall, AKI patients presented significantly higher and sustained levels (P < 0.05) of CCL-2, CCL-3, CCL-4, CXCL-8, CXCL-10, IFN-γ, IL-2, IL-6, TNF-α, IL-1Ra, IL-10 and VEGF. Importantly, higher levels of CCL-2, CXCL-10, IL-2, TNF-α, IL-10, FGFb, and VEGF were observed in AKI patients independently of death. ROC curves demonstrated that early alterations in CCL-2, CXCL-8, CXCL-10, IFN-γ, IL-6, IL-1Ra and IL-10 show a good predictive value regarding AKI development. Lastly, immune mediators were significantly associated with each other and with SARS-CoV-2 viral load in AKI patients.

Conclusions: COVID-19 associated AKI is accompanied by substantial alterations in circulant levels of immune mediators, which could significantly contribute to the establishment of kidney injury.

Keywords: Acute kidney injury; COVID-19; Cytokines; Immune mediators.

Fig. 1

Flowchart demonstrating the inclusion of patients with suspect SARS-CoV-2 infection during April to August 2020.

Fig. 2

Analysis of immune mediators of COVID-19 hospitalized patients evolving with and without AKI. Circulating levels of (A) chemokines, (B) pro-inflammatory cytokines, (c) anti-inflammatory cytokines and (D) growth factors were assessed by a multiplex assay. Data is shown as median ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 indicate comparison between groups (No AKI vs. AKI) in each time point (t test/Mann-Whiney test). P-values on horizontal bars indicate the comparison of time points (early vs. follow-up samples) for each group - blue lines for “No AKI” and red lines for “AKI” - (paired t test/Wilcoxon).

Fig. 3

Exploring the profile of immune mediators in COVID-19 patients according to AKI development. Heatmaps show the landscape of immune mediators between “No AKI” and “AKI” (A) and “AKI + survival” and “AKI + death” (C). Best-fit decision trees identified the mediators that efficiently segregated “No AKI” from “AKI” (B) and “AKI + survival” from “AKI + death” (D). Levels were placed in the root of the tree according to the cytokine/chemokine value (pg/mL) that best divided groups. The total of classified registers (correct and incorrect) for each class are given in parentheses for each terminal node with the Full training (FULL) and Leave-one-out cross-validation (LOOCV) accuracies. If incorrectly classified registers exist, they will appear after slash “/”.

Fig. 4

Analysis of immune mediators of COVID-19 hospitalized patients who died according to the development of AKI. Data is shown as violin plots. P-values indicates comparison between groups (No AKI vs. AKI; t test/Mann-Whitney).

Fig. 5

Correlation matrix for comparing associations between immune mediators at early timepoint in COVID-19 patients according to AKI development. Spearman’s coefficients are represented in different colors, as demonstrated by the vertical bar on the right.