Freitag, 02.10.2020

14:00 - 15:30

Poster DGfN 2020

Akutes Nierenversagen 1 (P001 - P008; LA01 - LA04)

AKI epidemiology and outcomes – a single-centre, retrospective cohort study

K. Asmus, O. Ritter, S. Patschan, D. Patschan; Brandenburg

Objective: Acute kidney injury substantially worsens the prognosis of hospitalized patients. The Brandenburg Medical School was founded in 2014, a nephrology section was opened in summer 2017. Aim of the study was to analyze AKI epidemiology and outcomes in one of two university hospitals belonging to the Medical School.
Method: The investigation was designed as single-centre, retrospective cohort study at the Brandenburg hospital of the Brandenburg Medical School. All in-hospital patients treated between january and the end of december 2015 were included. AKI was defined as specified in the 2012 published KDIGO criteria (criterion 1 and 2). Four parameters were evaluated in particular: AKI incidence, in-hospital mortality, frequency of renal replacement therapy, and renal recovery during the stay at the hospital.
Results: A total number of 5,300 patients were included in the analysis, AKI was diagnosed in 491 subjects (9.2%). The in-hospital mortality was 26%. Five point six (5.6)% received dialysis therapy, renal recovery occurred in 23.5% of all AKI subjects.
Conclusion: Both, the AKI incidence and the frequency of dialysis were lower than reported in the literature. However, the overall mortality was high and fewer subjects recovered from AKI. The findings possibly result from the lack of pre-hospitalization creatinine values, the lack of follow-up data, and a generally lower awareness for the need to perform renal replacement therapy in AKI.

Electronic AKI alert – a single-centre pilot study

A. Assem, O. Ritter, D. Patschan; Brandenburg

Objective: Acute kidney injury substantially worsens the prognosis of hospitalized patients. The Brandenburg Medical School was founded in 2014, a nephrology section was opened in summer 2017. An electronic AKI aIert system was implemented in summer 2018. Aim of the current study was to analyze the efficacy of the system under non-interventional conditions.
Method: The investigation was designed as single-centre, retrospective cohort study at the Brandenburg hospital of the Brandenburg Medical School. All in-hospital patients treated between mid october and mid december 2018 were included. The nephrologist in charge received an electronic message if the following criterium was fulfilled: increase in serum creatinine of higher than 26.5 Mikromol/L within 48 hours. During the observational period, the nephrologist exclusively participated in AKI care if any other physicians officially applied for support.
Results: The nephrologist in charge received a total number of 200 alerts. Out of these, 150 (75%) were judged as correct. Other physicians applied for support in only 10% of all alerts. The diagnosis AKI was mentioned in only 36% of all discharge letters, follow-up of the patients was recommended in only 20%.
Conclusion: The AKI associated healthcare management in Brandenburg neeeds to be improved. Measures must include: (I) optimization of the electronic alert algorithm, (II) sensitization of physicians in general, and (III) hiring of additional experts in the field

Non-compliance and acute dehydration are main reasons for acute kidney failure in patients with non-infantile primary hyperoxaluria type 1 (PH1)

C. Martin-Higueras, B. B. Beck, B. Hoppe; La Laguna/E, Köln, Bonn

Objective: Patients with primary hyperoxaluria type 1 have an increased risk of acute or chronic kidney failure (AKF/CKF). There are two risk groups: first the infantile oxalosis patients with an unremarkable family history and second, the adult patients with only an oligosymptomatic course (minor stone events), who get into end-stage renal disease (ESRD) due to oxalate-induced chronic inflammatory processes in the kidney. We were interested in the prevalence of acute kidney failure in the PH1 patients seen in our Hyperoxaluria Center over the last 10 years.
Method: We retrospectively analyzed the database of the German Hyperoxaluria Center for those patients being routinely seen in our outpatient clinic for AKF and its specific reason. AKF was defined as patients with a sudden onset of kidney failure and the necessity of dialysis installment, but a documented stable kidney function (no worse than stage 2-3 CKD) prior to onset of AKF.
Results: Currently, we routinely follow (3-4 times a year) 49 of 117 patients with genetically proven PH1. For the patients not routinely seen, we act as the reference center and primary lab. Out of the 117 PH1 patients, 6 had infantile oxalosis and thus early ESRD; 8 patients are currently on dialysis; and 6 patients have died, 2 patients each either post liver or liver/kidney transplantation (LTx, LKTx), under dialysis or as sudden death, the latter two due to systemic oxalosis. Transplantations were performed in 35 patients with PH1 (LKTx in 28 patients, LTx in 4 patients and 3 had an isolated kidney transplantation). Out of the 49 PH1 patients we routinely follow, AKF was diagnosed in 8 patients aged 11-56 years (one pediatric patient, 3 patients 18-19 years old, and 4 patients 29-56 years old), which has led to CKF and hence, maintenance hemodialyis in 7/8 and even death in one patient. Non-compliance regarding either medication or recommended fluid intake was the reason for AKF in 6 of the 8 patients, the 11 year-old girl, 2 adult patients and the 3 young adults being on the edge of leaving pediatric care. Acute massive diarrhea without adequate fluid substitution led to AKF in 2 middle aged male patients.
Conclusion: AKF is not uncommon in patients with PH1 and is frequently related to either non-compliance or to situations of severe fluid losses with inadequate fluid substitution. In the situations described, AKF is not based on already extremely altered renal function. Therefore, interruption of medication or lack of fluid intake, even for short periods of time, can lead to severe clinical consequences in these patients.

Effect of coronary angiography on Dickkopf-3 concentrations

F. Seibert, A. Heringhaus, N. Pagonas, B. Rohn, F. Bauer, H.-J. Trappe, U. Landmesser, N. Babel, T. H. Westhoff; Herne, Berlin

Objective: Dickkopf-3 (DKK3) has recently been discovered as a urinary biomarker for the prediction of acute kidney injury (AKI), and its concentrations are associated with the progression of chronic kidney disease (CKD). It remains elusive, however, whether it is able to detect mild forms of tubular injury, e. g. due to contrast media application. We therefore compared DKK3 concentrations before and after coronary angiography
Method: We performed a prospective study in 490 patients undergoing coronary angiography. AKI was defined according to the Acute Kidney Injury Network (AKIN) criteria. The study population included both subjects with and without CKD. DKK3 was measured by ELISA (ReFiNE; DiaRen UG, Homburg/Saar, Germany) in urine samples obtained <24h before coronary angiography and within 72h after coronary angiography.
Results: CI-AKI was observed in 30 (6.1%) patients, from whom 27 corresponded to stage I and 3 to stage II. DKK3 ratios were 42 fold higher after coronary angiography (2.1 pg/mg [IQR 0.9-187.9] vs. 89.5 pg/mg [IQR 0-338.3]), p=0.001). The rise of DKK3/creatinine in those subjects without AKI was 43 fold (2.0 pg/mg [IQR 0.9-174] at baseline vs. 86.1 pg/mg [IQR 0-322.6] after the procedure; p=0.001). Those with AKI (7.1 pg/mg [IQR 1.1-1245] at baseline vs. 206.1 pg/mg [IQR 36.6-567.4] after the procedure; p=0.57) presented a 29 fold increase of DKK3/creatinine ratios.
Conclusion: Coronary angiography is associated with a strong increase of urinary DKK3/creatinine. Whether this reflects transient or permanent tubular injury needs to be investigated in the future.

Glomerular filtration barrier dysfunction in an RNA virus-induced glomerulopathy resembles findings of common nephrotic syndromes

C. Nußhag, A. Stütz, S. Hägele, C. Speer, F. Kälble, C. Eckert, C. Morath, J. Reiser, M. Zeier, E. Krautkrämer; Heidelberg, Chicago/USA

Objective: Virally induced kidney dysfunction is highlighted by the alarming incidence of SARS-CoV-2 associated acute renal disease including nephrotic syndrome (NS). Plasma levels of soluble urokinase plasminogen activator receptor (suPAR) are elevated in COVID patients and provide prognostic insights. SuPAR is also involved in proteinuric kidney diseases such as focal segmental glomerulosclerosis in which podocytes effacement/injury is a common feature. Hantavirus-induced hemorrhagic fever with renal syndrome (HFRS) represents another RNA virus-induced disease with acute kidney injury and NS. The exact pathophysiology of proteinuria is, however, unclear. We hypothesized that hantavirus infection results in podocyte injury and a dysfunctional glomerular filtration barrier (GFB), similar to findings in common NS.
Method: Renal biopsy specimens were analyzed by light and electron microscopy. Urinary nephrin and serum suPAR were measured over time in 26 patients with HFRS and 18 healthy controls.
Results: Hantavirus patients showed significantly increased urinary nephrin, immunoglobulin G (IgG), a1-microglobulin (a1-MG) and serum suPAR concentrations compared to healthy controls. Furthermore, nephrin and IgG levels were significantly higher in patients with severe than with mild proteinuria. Differences in a1-MG levels, however, disappeared after normalization to urinary creatinine. Urinary nephrin levels as a marker for podocyte damage correlated strongly with biomarkers of non-selective glomerular proteinuria. Interestingly, suPAR correlated significantly with urinary nephrin, IgG and albumin levels, suggesting suPAR as a potential pathophysiological mediator in GFB dysfunction in response to RNA virus infection. The main finding in microscopy analyses was a focal foot process effacement. Proteinuria and kidney dysfunction recovered autonomously in all patients.
Conclusion: Hantavirus infection causes a podocyte injury leading to GFB dysfunction. A better understanding of transient virally induced proteinuria syndromes and their often self-limiting disease character may generate new therapeutic approaches for NS.

suPAR determines outcomes in septic acute kidney injury

C. Nußhag, C. Wei, R. Szudarek, F. Kälble, C. Speer, F. Uhle, J. Eugen-Olsen, T. Krarup, M. Weigand, M. Zeier, C. Morath, T. Brenner, J. Reiser; Heidelberg, Chicago/USA, Hvidovre/DK, Birkeroed/DK, Essen

Objective: Sepsis is the main contributor to the development of acute kidney injury (AKI) in critically ill patients. Plasma soluble urokinase plasminogen activator receptor (suPAR) is a criculating risk factor for AKI and a prognostic marker for the need of renal replacement therapy (RRT). We analyzed the pathophysiological role and kinetic properties of suPAR in septic AKI in critically ill patients and in a murine model of septic AKI.
Method: 200 critically ill patients were enrolled prospectively after meeting Sepsis-3 criteria. Serum suPAR levels were measured at 0, 12, 24, 48, 72, 96, 120 and 168-hour after enrollment and the need for RRT within 7 days was assessed as the primary outcome measure. Polybacterial sepsis was induced by cecal slurry injection in three mouse strains, respectively wild type (WT, N=9), uPAR-knockout (KO, N=13), and suPAR transgenic overexpression (OE, N=11).
Results: No or mild AKI occurred in 62 patients (31.0%), moderate or severe AKI without the need for RRT in 102 patients (51.0%), criteria for RRT were met in 36 patients (18.0%) and 7 patients (3.5%) died within the 7-day period. Compared to all other maximum AKI stages and AKI disease courses within 7 days, patients requiring RRT showed significantly higher suPAR levels at all time-points. Patients with suPAR levels ≥ 12.7 ng/mL (highest quartile) had an adjusted odds ratio of 5.22 (95% confidence interval [CI], 2.16-12.65) for the need for RRT; and 4.44 (95% CI, 1.98-9.97) for RRT or death within 7 days compared to patients with levels < 12.7 ng/mL. Compared to KO mice, WT and OE mice showed a significantly greater impairment of renal function and structure 24 hours after induction of sepsis. Kaplan-Meier analysis revealed a survival benefit of KO mice over OE mice within 24h (84.6% vs. 45.5%, p=0.041).
Conclusion: SuPAR distinguishes between divergent AKI stages/courses and the need for RRT at any time within 7 days after sepsis diagnosis. Our experimental data suggest that suPAR is a pathophysiological driver of septic AKI and may serve as a target for future interventional strategies.

Studienbedingungen versus klinische Routine – differierende Zahlen zur Inzidenz der akuten Nierenschädigung nach herzchirurgischen Eingriffen

M. Schanz, N. Göbel, T. Oberacker, S. Schricker, M. D. Alscher, U. Franke, M. Ketteler; Stuttgart

Hintergrund: Die akute Nierenschädigung (Acute Kidney Injury: AKI) ist mit einer hohen Morbidität und Mortalität verbunden. Insbesondere die Prävention ist aktuell Gegenstand intensiver Forschungsbemühungen. Inwiefern Daten aus klinischen Studien in die tägliche Praxis übertragbar sind, ist bisher unklar. Daher vergleicht diese Arbeit die unter Studienbedingungen erfasste und unter Alltagsbedingungen im gleichen Kollektiv dokumentierte AKI-Inzidenz.
Methode: Diese retrospektive Studie umfasst n=100 konsekutive Patienten ohne vorbestehende AKI, welche sich im Zeitraum von März bis April 2019 einem herzchirurgischen Eingriff mit Herz-Lungen-Maschine unterzogen. Es wurde die in der klinischen Routine dokumentierte (ICD-Codes/Briefe) mit der gemäß KDIGO-Definition von 2012 (Urinausscheidung, Serum-Kreatinin-Werte) verzeichneten Inzidenz einer postoperativen AKI innerhalb von 72 Stunden nach dem Eingriff erfasst und verglichen.
Ergebnisse: Von den 100 Patienten entwickelten 52% eine AKI innerhalb 72 Stunden nach Operation gemäß KDIGO-Definition. Die unter Alltagsbedingungen erfasste und dokumentierte Inzidenz (Routine) betrug jedoch nur 12%, was einer Rate von lediglich 23.1% der wahren Inzidenz in dieser Kohorte entspricht. Dieser Unterschied zeigte sich mit p<0.0001 statistisch hochsignifikant. Die Verteilung der AKI-Stadien zwischen KDIGO-Inzidenz und Routine-Inzidenz (% der KDIGO-Inzidenz) zeigte sich wie folgt: Stadium 1: KDIGO: n=31 versus Routine: n=1 (3.2%); Stadium 2: KDIGO: n=10 vs. Routine: n=3 (30%); Stadium 3: KDIGO: n=11  vs. Routine: n=7 (63.6%). N=1 wurde in der Routine ohne Stadium dokumentiert. Bei n=4 (33.3%) der routinemäßig dokumentierten AKI wurde ein inkorrektes Stadium angegeben. Die Gruppe der KDIGO-AKI-Inzidenz wies im Vergleich zu der in der Routine erfassten Inzidenz eine signifikant (p<0.0001) kürzere Intensiv-Aufenthaltsdauer (Median: 1 [IQR 1.0-3.0] vs. 7 Tage [3.8-10.8]) auf, wohingegen sich die Krankenhausverweilzeit nicht signifikant (p=0.24) unterschied.
Zusammenfassung: Es besteht eine große, signifikante Diskrepanz zwischen der unter Studienbedingungen anhand von Leitlinienkriterien erfassten und in der klinischen Routine wahrgenommenen Inzidenz einer AKI nach herzchirurgischem Eingriff. Dies stellt einen potentiell komplizierenden Faktor im postoperativen Management einer AKI dar, da ein beträchtlicher Anteil von AKI im Alltag nicht erkannt wird. Bisherige Daten zur Inzidenzsenkung einer AKI lassen sich daher nicht ohne Weiteres auf die klinische Praxis übertragen.

Nephroprotektion durch Aktivierung von HIF-Signaling: eine vergleichende Analyse pharmakologischer PHD-Inhibition und repetitiver Hypoxie

M. Späth, K. Bohl, K. J. R. Hoyer-Allo, I. Flamme, F. Köhler, F. Grundmann, C. D. Dèbes, A. Beyer, B. Schermer, T. Benzing, V. Burst, R.-U. Müller; Köln, Wuppertal

Hintergrund: Die akute Nierenschädigung (AKI) ist eine der häufigsten Komplikationen, die zu einer erhöhten Morbidität und Mortalität führt. Es fehlen jedoch präventive oder therapeutische Strategien in der klinischen Routine. Im Tiermodell kann die AKI durch Präkonditionierungsstrategien, z. B. durch wiederholte Exposition gegenüber Hypoxie (HP) wirksam verhindert werden. Eine direkte Translation in das klinische Umfeld ist nicht praktikabel, jedoch könnte die Hemmung von Prolylhydroxylasen (PHD) und die anschließende Aktivierung von „Hypoxie induzierbaren Faktoren“ (HIF) eine attraktive Strategie sein. Unser Ziel war es, die Protektion durch HP und PHD-Inhibition (PHDi) im Mausmodell des renalen Ischämie-Reperfusionsschadens (IRI) zu bestätigen und gemeinsame molekulare Muster zu charakterisieren, um neue therapeutische Ziele zu identifizieren.
Methode: Männliche C57Bl6-Wildtyp-Mäuse wurden zunächst entweder an drei folgenden Tagen Hypoxiephasen zunehmender Dauer ausgesetzt oder mit verschiedenen Dosierungen eines PHD-Inhibitors behandelt, bis eine ähnliche Induktion des HIF-Zielgens EPO - und damit die Vergleichbarkeit beider Ansätze - erreicht wurde. Nach dieser Dosisfindung wurden weitere Mäuse nach Präkonditionierung (HP/PHDi) sowie Kontrollen einer einseitigen Nephrektomie und einer 40-minütigen kontralateralen renalen Ischämie unterzogen. Hiernach erfolgte eine funktionelle (z. B. Serum-Kreatinin), histologische und molekularbiologische Analyse (z. B. RNAseq) der rechten und linken Nieren.
Ergebnisse: HP und PHDi verbesserten das Outcome 24 h nach der Reperfusion funktionell und histologisch signifikant. Die Transkriptomanalysen der ungeschädigten Nieren zeigten nur einen geringen Einfluss von HP und PHDi. Nach der Schädigung unterschieden sich die durch HP bzw. PHDi behandelten Nieren stark von den Kontrollen. Hierbei zeigten sich einerseits deutliche Unterschiede zwischen den Gruppen, andererseits jedoch einige interessante Gemeinsamkeiten. Diese erlauben es nun Gene und Signalwege zu identifizieren und zu testen, welche mechanistisch an der Nephroprotektion beteiligt sind.
Zusammenfassung: Die protektive Wirkung von HP und PHDi im renalen IRI konnte bestätigt werden. Zudem wurde die erste vergleichende molekulare Phänotypisierung von Nieren durchgeführt, die mit diesen Strategien behandelt wurden. Zukünftige Studien werden die Frage beantworten, ob die identifizierten Gene und Signalwege moduliert werden können, um eine AKI zu verhindern.

Discovery of a necroptosis-inducing receptor in acute kidney injury

W. Tonnus, J.-H. Bräsen, A. von Mäßenhausen, H. Walczak, C. Hugo, A. Linkermann; Dresden, Hannover, Köln

Objective: By utilizing RIPK3- and MLKL-deficient mice as well as the small molecule Nec-1, evidence was gained about an important pathophysiological role of necroptosis in renal ischemia/reperfusion injury (IRI). Necroptosis is a form of regulated necrosis, which can be triggered by stimulation of death receptors (such as TNFR1/2, TRAIL or Fas), receptors of innate immunity (TLR3/4) or via Z-DNA-binding protein ZBP1. However, it is unclear which of these actually contribute to necroptosis in IRI.
Method: Knock-out mice and pharmacological inhibitors of necroptosis-assocated receptors were utilized in renal IRI.
Results: First, we found that ZBP1-ko mice demonstrate higher levels of serum creatinine and urea with a concordantly increased tubular damage pattern after IRI. We then assessed TLR2- and TLR4-deficient mice in the same experimental setting but could not detect any significant effect compared to wildtype littermates.
In the last set, we investigated death receptors. Whereas pharmacological inhibition of TRAIL did not affect the experimental outcome, TNFR1/2-deficiency significantly deteriorated tubular damage as well as serum levels of creatinine and urea, respectively. Importantly, inhibition of TNFα by the chimeric TNF-receptor fragment Etanercept mimicked this effect. Treatment of mice undergoing IRI with APG101, an inhibitor of Fas (also known as CD95), significantly ameliorated tubular damage after IRI.
In support of this notion, we induced a necroptotic phenotype in cell culture by a combination of anti-Jo2 (murine Fas ligand) and zVAD-fmk (pan caspase inhibitor), which could be prevented by necrostatins such as Nec-1s and NSA.
Conclusion: In summary, we identified Fas as the functional trigger of necroptosis during IRI and provide evidence, that APG101, a drug currently in clinical trials for cancer therapy, might be repurposed for the treatment of AKI. Furthermore, we found other receptors in the necroptosis pathway to be protective, as their deletion deteriorated IRI. To this regard, further studies are required to mechanistically address this phenomenon.In summary, we identified Fas as the functional trigger of necroptosis during IRI and provide evidence, that APG101, a drug currently in clinical trials for cancer therapy, might be repurposed for the treatment of AKI. Furthermore, we found other receptors in the necroptosis pathway to be protective, as their deletion deteriorated IRI. To this regard, further studies are required to mechanistically address this phenomenon.

Gasdermin D-mediated pyroptosis does not functionally affect the outcome in murine models of AKI

W. Tonnus, M. Latk, A. von Mäßenhausen, F. Shao, C. Hugo, A. Linkermann; Dresden, Bejing/CN

Objective: For the pathophysiology of renal ischemia/reperfusion injury (IRI) a central role for regulated necrosis (RN) has been established within recent years. However, several forms of RN have been described. Whereas solid evidence suggests a role for necroptosis and ferroptosis in IRI, the functional contribution of pyroptosis, a highly inflammatory form of RN, is unclear.
Method: To address this, we utilized mice lacking the pore-forming effector protein of pyroptosis gasdermin D (GSDMD) and evaluated the outcome after IRI.
Results: In comparison to wildtypes, GSDMD-deficient mice demonstrated slightly increased tubular damage and levels of serum creatinine and urea, respectively. Of note, this was also true for the model of cisplatin-induced AKI.
However, gasdermin E (GSDME) is a closely related protein to GSDMD and has been described to induce pyroptosis under specific circumstances. To further rule out a contribution of pyroptosis to AKI, we utilized GSDME-deficient mice. Neither GSDME-ko mice nor the GSDMD/E-double knockout mice displayed any significant reduction in the analyzed parameters after IRI.
Conclusion: In summary, we could rule out a major functional role for pyroptosis in IRI. However, all knock-outs displayed slightly increased levels of kidney damage.

Phase-specific IRF8 and IRF4 regulation/expression of mononuclear phagocytes during ischemic acute kidney injury/disease

N. Li, S. Steiger, Z. Zheng, H.-J. Anders, J. Lichtnekert; München, Guangdong/CN

Objective: Ischemic acute tubular necrosis is a common cause of acute kidney disease (AKD). We hypothesized that the dynamically altered MPC subsets contribute to the progression of AKD, accompanied with varied expression of IRF4 and IRF8.
Method: AKD was induced by transient unilateral renal pedicle clamping in C57BL / 6N mice. After 1, 3, and 7 days kidneys, renal lymph nodes and spleens were collected. Flow cytometry was performed to identify different MPC subsets. GFR was measured and mRNA expression of inflammatory, anti-inflammatory transcription factors determined via RT-PCR. For in vitro experiments, bone marrow-derived macrophages and DCs, tubular epithelial cells (TECs), renal and splenic resident CD11b + immune cells isolated from naïve mice were stimulated with LPS or cultured under hypoxic condition for 3 and 10 hours. Bone marrow-derived monocytes were differentiated into macrophages or DCs, and stimulated with LPS for 3 and 10 hours.After stimulation, cells were harvested for mRNA analysis via RT-PCR.
Results: We identified four renal phenotypically distinct MPC subsets with various expression patterns of CD11b / CD11c during the different phases of post-ischemic AKD. During the early (day 1) and late injury phase (day 3), the number of infiltrating CD11bhi CD11clow R2 and CD11bhi CD11chi R3 macrophage-like subsets increased, along with a significant GFR decline compared with sham-operated mice.


During the repair phase (day 7), the number of CD11blow CD11clow R1 (resident macrophage-like MPCs) and CD11blow CD11chi R4 (infiltrating cDC-like MPCs) subsets significantly increased. Both resident macrophage-like (R1) and cDC-like (R4) MPCs significantly upregulated the expression of IRF8, whereas cDCs-like (R4) MPCs were also positive for IRF4 during the repair phase.This pattern of MPCs was consistent in spleen and renal lymph node. In vitro stimulation of renal and splenic CD11b + cells from naïve mice with LPS or under hypoxic condition induced a significant upregulation of IRF4 and IRF8 compared to untreated cells. This was also observed in bone marrow-derived macrophages and cDCs but not in TECs.
Conclusion: Our data indicate that infiltrating macrophage-like and cDC-like MPCs appear in high numbers during the early and late injury phase. Furthermore, both resident macrophage-like and cDC-like MPCs are predominately present during the late injury and recovery phase in AKD with altered IRF4 and IRF8 expression pattern. Further studies are needed to unravel a potential role of IRF4 and IRF8 during the progression of AKD and CKD by eg using fate mapping approaches.

Kidney biomarkers on ICU admission to predict initiation of acute renal replacement therapy after cardiac surgery

M. Ernst, M. Haase, B. Isermann, S. Westphal, P. Devarajan, C. Albert, H. Kuppe, S. Kropf, R. Bellomo, A. Haase-Fielitz, S. Elitok; Magdeburg, Potsdam, Leipzig, Cincinnati/USA, Berlin, Melbourne/AUS, Bernau

Objective: We assessed the predictive ability of urinary neutrophil gelatinase-associated lipocalin (NGAL)-to-hepcidin-25 ratio for severe adverse events after cardiac surgery.
Method: In 200 adult patients undergoing cardiac surgery at a University hospital, NGAL and hepcidin-25 concentrations were measured in urine using samples obtained on ICU admission. Primary endpoint was acute renal replacement therapy (aRRT), secondary endpoint was in-hospital mortality. We performed ROC-curve analysis, logistic regression analysis and cross-validated (leave-one-out) reclassification analysis to assess changes in model’s predictive ability. Cleveland risk score functioned as reference model for prediction of aRRT and EuroScore for prediction of in-hospital mortality. Finally, patients were allocated to four different combinations of AKI- and urinary NGAL/hepcidin-25 ratio-status: NGAL-to-hepcidin-25 ratio(-)/AKI(-), NGAL-to-hepcidin-25 ratio(+)/AKI(-), NGAL-to-hepcidin-25 ratio(-)/AKI(+), NGAL-to-hepcidin-25 ratio(+)/AKI(+) and endpoints were reported per group.
Results: The value of NGAL/hepcidin-25 ratio (AUC 0.82 [95% CI 0.68-0.97]) to predict aRRT (N=13 patients) was higher than when NGAL (0.74 [0.59-0.89]) and hepcidin (0.75 [0.65-0.85]) were used individually and persisted after adjusting for potential confounders (adjusted OR 1.8, 95% CI 1.2-2.8). Only NGAL/hepcidin-25 ratio showed improvement of cfNRI 0.71 (95% CI 0.17-1.26) and IDI 0.06 (95% CI 0.002-0.12). NGAL/hepcidin-25 ratio was also an independent predictor of in-hospital mortality (adjusted OR 2.0, [95% CI 1.3-3.0]). Finally, highest rates of aRRT and in-hospital mortality were found in NGAL-to-hepcidin-25 ratio(+)/AKI(+) patients and NGAL-to-hepcidin-25 ratio(+)/AKI(-) patients, whereas best outcomes in NGAL-to-hepcidin-25 ratio(-)/AKI(-) patients.

Conclusion: On ICU admission after cardiac surgery, the NGAL/hepcidin-25 ratio is a stronger predictor of aRRT than the single biomarkers alone. NGAL/hepcidin-25 ratio complements classical kidney risk criteria.
Figure 1. Risk assessment plot of NGAL-to-hepcidin-25 ratio for acute renal replacement therapy (aRRT) after cardiac surgery
X-axis: calculated risk for aRRT
Y-axis: sensitivity
If the model’s calculated risk improves after adding the biomarker to it, the dots (specificity) will move closer to 0.0 (non-events) and the squares (sensitivity) move closer to 1.0 (events).



Fragen und Antworten der Posterbewerter und Autoren

Nur Posterbewerter und Autoren dieser Postergruppe können Fragen/Antworten erstellen.
Sie müssen sich anmelden, um Kommentare hinzuzufügen.