Thorac Cardiovasc Surg 2022; 70(S 01): S1-S61
DOI: 10.1055/s-0042-1742795
Oral and Short Presentations
Sunday, February 20
Perioperative Cardiac Surgical Therapy: Optimized Concepts

Prediction of Cardiac Surgery–Associated Acute Kidney Injury by Perioperative Quantification of Serum Proenkephalin Levels

P. Grieshaber
1   Department of Cardiac Surgery, University Hospital Heidelberg, Heidelberg, Deutschland
,
B. Arneth
2   University Hospital Giessen, Giessen, Deutschland
,
N. Kloke
2   University Hospital Giessen, Giessen, Deutschland
,
C. Gediz
2   University Hospital Giessen, Giessen, Deutschland
,
J. Schulte
3   Sphingotec GmbH, Hennigsdorf, Deutschland
,
O. Hartmann
3   Sphingotec GmbH, Hennigsdorf, Deutschland
,
H. Renz
2   University Hospital Giessen, Giessen, Deutschland
,
A. Böning
4   Rudolf-Buchheim-Str. 7, Gießen, Deutschland
› Author Affiliations
› Further Information
   All articles of this category

Background: Acute kidney injury (AKI) is a relevant and frequent complication of cardiac surgery. Its prediction, early diagnosis (usually based on serum creatinine measurements) and therapy are challenging. Therefore, biomarkers which predict AKI early and accurately are needed. Proenkephalin (PENK) has recently been suggested as a new biomarker for AKI. This study aimed to evaluate the predictive value of perioperative serum PENK levels for AKI.

Method: All patients, excluding those on chronic hemodialysis, undergoing surgery with cardiopulmonary bypass at our institution between June 2018 and May 2019 were invited to participate. Blood samples at eight perioperative time points (preoperative day, start of surgery, immediately postoperatively, 4 h, 8 h, 12 h, 24 h, 48 h postoperatively) were collected and PENK levels analyzed using a standardized immunoassay. AKI onset was observed until postoperative day 6 and classified using the KDIGO criteria. PENK levels were compared between patients who developed different AKI stages and performance characteristics for PENK cut-off 80 pmol/L were evaluated using receiver operator curves (ROC).

Results: AKI occurred in 48% of 258 patients (n = 124; thereof 35% AKI stage 1, 3% AKI stage 2, 10% AKI stage 3). The 30-day mortality in patients with AKI stage 1 was 3.6%, 0% for AKI stage 2 and 44% for patients with AKI stage 3 (p < 0.01). Already preoperatively, PENK levels were significantly discriminating between patients who would develop different stages of AKI postoperatively (no AKI vs. all-stage AKI [mean ± standard deviation: 63 ± 21 versus 88 ± 58 pmol/L, p < 0.001; no AKI or AKI stage 1 versus AKI stage 2 or 3: 71 ± 41 versus 104 ± 57 pmol/L, p < 0.001; no AKI, AKI stage 1 or 2 versus AKI stage 3: 71 ± 41 versus 106 ± 60 pmol/L, p < 0.001). These discriminations were consistent over all time points of PENK analysis. ROC analyses confirmed fair prediction for all AKI stages preoperatively (AUC [AKI stage 1, 2 or 3] 0.622, p < 0.001; AUC [AKI stage 2 or 3] 0.704, p < 0.001; AUC [AKI stage 3] 0.712, p < 0.001); AUC [AKI requiring dialysis] 0.75, p < 0.001), with sensitivities between 42% (AKI stage 1 or 3) and 64% (AKI stage 3). The strongest prediction was achieved 48 hours postoperatively (AUC [AKI stage 1, 2, or 3] 0.752, p < 0.001, sensitivity of 48%, specificity of 93%).

Conclusion: The analysis of serum PENK levels allows for prediction of AKI already preoperatively with fair accuracy. Therefore, PENK seems to be a promising tool for early identification of patients at increased risk for AKI. Further studies have to evaluate if PENK-based early nephroprotective therapeutic regimes in those risk patients can reduce the incidence of AKI.



Publication History

Article published online:
03 February 2022

© 2022. Thieme. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany