糖心vlog

Importance听 No interventions have yet been identified to reduce the risk of acute kidney injury in the setting of cardiac surgery.

Objective听 To determine whether remote ischemic preconditioning reduces the rate and severity of acute kidney injury in patients undergoing cardiac surgery.

Design, Setting, and Participants听 In this multicenter trial, we enrolled 240 patients at high risk for acute kidney injury, as identified by a Cleveland Clinic Foundation score of 6 or higher, between August 2013 and June 2014 at 4 hospitals in Germany. We randomized them to receive remote ischemic preconditioning or sham remote ischemic preconditioning (control). All patients completed follow-up 30 days after surgery and were analyzed according to the intention-to-treat principle.

Interventions听 Patients received either remote ischemic preconditioning (3 cycles of 5-minute ischemia and 5-minute reperfusion in one upper arm after induction of anesthesia) or sham remote ischemic preconditioning (control), both via blood pressure cuff inflation.

Main Outcomes and Measures听 The primary end point was the rate of acute kidney injury defined by Kidney Disease: Improving Global Outcomes criteria within the first 72 hours after cardiac surgery. Secondary end points included use of renal replacement therapy, duration of intensive care unit stay, occurrence of myocardial infarction and stroke, in-hospital and 30-day mortality, and change in acute kidney injury biomarkers.

Results听 Acute kidney injury was significantly reduced with remote ischemic preconditioning (45 of 120 patients [37.5%]) compared with control (63 of 120 patients [52.5%]; absolute risk reduction, 15%; 95% CI, 2.56%-27.44%; P鈥�=鈥�.02). Fewer patients receiving remote ischemic preconditioning received renal replacement therapy (7 [5.8%] vs 19 [15.8%]; absolute risk reduction, 10%; 95% CI, 2.25%-17.75%; P鈥�=鈥�.01), and remote ischemic preconditioning reduced intensive care unit stay (3 days [interquartile range, 2-5]) vs 4 days (interquartile range, 2-7) (P鈥�=鈥�.04). There was no significant effect of remote ischemic preconditioning on myocardial infarction, stroke, or mortality. Remote ischemic preconditioning significantly attenuated the release of urinary insulinlike growth factor鈥揵inding protein 7 and tissue inhibitor of metalloproteinases 2 after surgery (remote ischemic preconditioning, 0.36 vs control, 0.97 ng/mL²/1000; difference, 0.61; 95% CI, 0.27-0.86; P鈥�<鈥�.001). No adverse events were reported with remote ischemic preconditioning.

Conclusions and Relevance听 Among high-risk patients undergoing cardiac surgery, remote ischemic preconditioning compared with no ischemic preconditioning significantly reduced the rate of acute kidney injury and use of renal replacement therapy. The observed reduction in the rate of acute kidney injury and the need for renal replacement warrants further investigation.

Trial Registration听 German Clinical Trials Register Identifier:

Acute kidney injury is a well-recognized complication after cardiac surgery and significantly affects morbidity and mortality.^1,2 Up to 30% of patients develop acute kidney injury after cardiac surgery, whereas severe acute kidney injury requiring dialysis is relatively rare.³ Approximately 1% of all patients undergoing cardiac surgery develop a severe dialysis-dependent acute kidney injury, and this severity of injury is associated with especially poor outcomes.¹ Although the mechanisms of acute kidney injury are not fully understood, injury to renal tubular epithelial cells is a universal aspect of the disease. Despite numerous clinical trials using several interventions,⁴ a reliable means to prevent acute kidney injury remains elusive.

Remote ischemic preconditioning elicited by brief episodes of ischemia and reperfusion in distant tissue may provide protection from subsequent injury.⁵ In cardiac surgery, adverse outcomes are mainly linked to perioperative myocardial injury.⁶

Remote ischemic preconditioning may attenuate renal injury by releasing various molecules such as damage-associated molecular patterns that are then filtered by the kidney and signal through Toll-like receptors in the proximal tubule epithelia.^5,7 This signaling may then induce natural defenses such as bioenergetic down-regulation and temporary cell-cycle arrest.⁸ These defenses, once engaged, can then protect the kidney during subsequent inflammatory or ischemic stress.

However, despite this rationale, 3 small single-center randomized trials investigating the effect of remote ischemic preconditioning on acute kidney injury after cardiac surgery have shown conflicting results.^9-11 Thus, a large randomized study with a robust and relevant clinical end point has been called for.¹² As an initial step to achieving this goal, we performed a multicenter, randomized, controlled, clinical trial to investigate whether remote ischemic preconditioning could reduce the occurrence and severity of acute kidney injury as defined by Kidney Disease: Improving Global Outcomes criteria¹³ and to analyze other relevant clinical outcomes of remote ischemic preconditioning in cardiac surgery patients at high risk for acute kidney injury after on-pump cardiac surgery. Our goal was to acquire phase 2 equivalent data to support a larger multicenter trial.

After obtaining approval from the institutional review boards at each site, we performed a multicenter, double-blind, randomized clinical trial (study protocol appears in Supplement 1). Consecutive patients were approached for enrollment during preadmission consultations and provided written informed consent. The study was conducted according to the principles of the Declaration of Helsinki. Eligible patients were adults at high risk for acute kidney injury who underwent cardiac surgery with the use of cardiopulmonary bypass at the universities of M眉nster, T眉bingen, Freiburg, or Bochum (all in Germany) between August 2013 and June 2014. A Cleveland Clinic Foundation score (eTable 1 in Supplement 2) of 6 or higher was used to define patients at high risk for acute kidney injury.¹⁴ The score is composed of different risk factors, including patient characteristics, comorbidities, and type of surgery.¹⁴ Exclusion criteria were acute myocardial infarction up to 7 days before surgery, age younger than 18 years, off-pump heart surgery, preexisting acute kidney injury, kidney transplantation, chronic kidney disease with a glomerular filtration rate less than 30 mL/min, pregnancy, peripheral vascular disease affecting the upper limbs, hepatorenal syndrome, and drug therapy with sulfonamide or nicorandil (preconditioning-blocking and preconditioning-mimetic medication, respectively).

Patients were randomized on a 1:1 basis, stratified by center. Randomization codes were computer generated and concealed from investigators. On the day of surgery, patients were assigned to undergo either remote ischemic preconditioning or sham remote ischemic preconditioning (control) (Figure 1), and the intervention was provided by an investigator not involved in the care of the patient. Patients, anesthesiologists, staff providing care of the patient, cardiac surgeons, and intensive care physicians were unaware of treatment assignment.

Anesthesia was induced according to the standard of care at each center and maintained with volatile anesthetics because propofol may interfere with remote ischemic preconditioning.¹⁵ According to a recently published review,¹⁶ we standardized the management of cardiopulmonary bypass as follows: mean arterial blood pressure of 60 to 70 mm Hg, the use of nonpulsatile cardiopulmonary bypass, 伪-stat acid-base management to regulate carbon dioxide tension, hematocrit values of 25% to 30%, blood glucose levels less than 200 mg/dL, and the use of arterial line filters.

After induction of anesthesia and before skin incision, we performed remote ischemic preconditioning consisting of 3 cycles of 5-minute inflation of a blood pressure cuff to 200 mm Hg (or at least to a pressure 50 mm Hg higher than the systolic arterial pressure) to one upper arm, followed by 5-minute reperfusion with the cuff deflated. In patients assigned to the control group, sham remote ischemic preconditioning intervention was induced by 3 cycles of upper limb pseudo ischemia (low pressure, 5-minute blood pressure cuff inflation to a pressure of 20 mm Hg and 5-minute cuff deflation). The surgical procedure and perioperative care were performed according to the standard at each center.

Our primary end point was the occurrence of acute kidney injury within the first 72 hours after surgery. We defined acute kidney injury according to the Kidney Disease: Improving Global Outcomes criteria (eTable 2 in Supplement 2).¹³ Secondary end points were severe acute kidney injury (stage 2-3) within 72 hours, 30-day all-cause mortality, need for renal replacement therapy during index hospitalization, duration of ventilator support, length of stay in the intensive care unit, length of hospital stay, in-hospital death, concentrations of various urinary biomarkers in the first 24 hours after surgery, and perioperative myocardial infarction and stroke during the index hospital stay.

We abstracted clinical variables from the medical record. Initiation of renal replacement therapy was at the discretion of the intensive care unit clinicians blinded to treatment assignment. Criteria for renal replacement therapy were not included in the protocol. Perioperative myocardial infarction and stroke were defined as described previously.¹⁷ Perioperative myocardial infarction was defined as cardiac troponin I concentration in serum more than 5 times the 99th percentile of the reference range when associated with new left bundle-branch block pathologic Q waves, or angiography-confirmed new or native coronary occlusion. Postoperative myocardial infarction was defined as an increase in troponin complex I concentration from baseline to at least twice the upper limit of normal, together with evidence of myocardial ischemia, such as electrocardiographic changes or angina symptoms. Cerebrovascular accidents or stroke during or after hospital admission was assessed if at least 1 of the following criteria was fulfilled: a neurologic event resulting in new, temporary, or permanent focal or global neurologic deficit, any embolic event after the immediate perioperative period (when anesthesia-induced unconsciousness was completely reversed), or a stroke or permanent neurologic event lasting longer than 24 hours or less than 24 hours if a cerebral lesion was observed on imaging. Repeated revascularization was defined as any percutaneous coronary intervention or repeated coronary artery bypass graft surgery after the primary coronary artery bypass graft surgery.

Blood samples were drawn before surgery and at prespecified points after surgery for measurement of serum creatinine concentrations (4 hours after cardiac surgery and on every morning for at least 3 days after cardiac surgery). We estimated glomerular filtration rate with the Modification of Diet in Renal Disease formula. Urine samples for biomarkers were collected before remote ischemic preconditioning or sham remote ischemic preconditioning, after inducing each one and at 4, 12, and 24 hours after surgery. Insulinlike growth factor鈥揵inding protein 7 (IGFBP7) and tissue inhibitor of metalloproteinases 2 (TIMP-2), both inducers of G1 cell cycle arrest, are implicated in acute kidney injury and serve as biomarkers to predict it. The product of urine TIMP-2 and IGFBP7 concentrations, (TIMP-2)鈥壝椻��(IGFBP7), was measured with the NephroCheck Test (Astute Medical). Urine neutrophil gelatinase鈥揳ssociated lipocalin (NGAL), was measured with a commercially available assay (Dianova) according to the manufacturer's protocol. Urine high-mobility group box (HMGB) 1 was measured with a commercially available assay () according to the manufacturer's protocol.

We calculated a necessary sample size based on the primary end point, using nQuery Advisor version 7. The primary efficacy analysis was intended to show superiority of remote ischemic preconditioning in high-risk cardiac surgery patients, applying a 2-sided 蠂² test on significance level 伪=.05. According to an observational study we performed in a similar patient population,¹⁸ the expected acute kidney injury rate in the control group treated with sham remote ischemic preconditioning was 50%. The expected absolute risk reduction for acute kidney injury was 18% according to a published single-center study investigating the effect of remote ischemic preconditioning on acute kidney injury after cardiac surgery.¹⁰ As a result of these considerations and a power of 80%, the required sample size was calculated to be 117 evaluable patients per treatment group, ie, 234 in total. An additional 6 patients were recruited to account for loss to follow-up or nonevaluable data.

The primary efficacy analysis included all randomized patients (full analysis set) and was performed according to the intent-to-treat principle, ie, all patients were analyzed according to their randomization (see statistical analysis plan in Supplement 1). For the primary outcome and the secondary end points acute kidney injury severity, need for renal replacement therapy, and mortality, all patients had complete data. For the analysis of biomarkers over time, information of all patients who had evaluable data for the respective time was included. For the logistic regression analyses, only patients with complete data regarding the included covariates were included. No imputation of the data was performed. Descriptive statistics are summarized for categorical variables as frequency (%) and were compared between groups with 蠂² test (or Fisher exact test if the produced matrixes contained cells with expected counts <5). Continuous variables, expressed as mean (standard deviation), were compared between groups with an unpaired t test. Continuous variables, which were not distributed normally, were analyzed with nonparametric tests (Mann-Whitney U and Wilcoxon for unpaired and paired observations, respectively). We estimated the relative risk (RR) reduction and the absolute risk reduction, including 95% CIs, for the occurrence of acute kidney injury, comparing the 2 study cohorts. The 95% CIs for median differences were calculated by bootstrapping (10鈥�000 random samples taken equally distributed from both randomization groups).

To identify the association between various risk factors and acute kidney injury, we used multivariable logistic regression with acute kidney injury within 72 hours of surgery (yes or no) as the dependent variable. We included variables from the Cleveland Clinic Foundation score¹⁴ (age, sex, diabetes, chronic obstructive pulmonary disease, previous heart surgery, and preoperative creatinine level), along with HMGB-1 (TIMP-2)鈥壝椻��(IGFBP7) (difference between pre- and post鈥搑emote ischemic preconditioning) and remote ischemic preconditioning as dependent variables, using backward likelihood ratios for variable retention in the model. We used the Wald test and reported P value odds ratios with 95% CIs. To identify factors associated with (TIMP-2)鈥壝椻��(IGFBP7) immediately after remote ischemic preconditioning, we used a predefined cutoff of 0.5 ng/mL²/1000¹⁸ and used multivariable logistic regression with the same variables as described above (except [TIMP-2] 脳 [IGFBP7]) as independent variables. Model performance was assessed by the analysis of the area under the receiver operating characteristic curve. P value is given for the hypothesis test area under the curve鈥�=鈥�0.5. IBM SPSS version 21.0 software was used. Two-sided P values 鈮�.05 were considered indicative of statistical significance.

Of 790 patients screened for the trial, 240 were enrolled and randomized to receive either remote ischemic preconditioning (n鈥�=鈥�120) or sham remote ischemic preconditioning (control) (n鈥�=鈥�120) and included in the primary analysis (Figure 1). The baseline and intraoperative characteristics were similar between the groups (Table 1). The number of patients with a low ejection fraction was similar between the groups.

Significantly fewer patients in the remote ischemic preconditioning arm developed acute kidney injury within 72 hours after surgery compared with the control group (37.5% vs 52.5%; P鈥�=鈥�.02; RR, 71%; 95% CI, 54%-95%; absolute risk reduction, 15.0%; 95% CI, 2.56%-27.44%; RR reduction, 28.6%; 95% CI, 5%-47%) (Table 2). Correction of serum creatinine level for fluid balance slightly changed the occurrence of acute kidney injury but did not change the difference in acute kidney injury rate between the remote ischemic preconditioning group and control group (42.5% vs 53.3%; P鈥�=鈥�.03). We performed a stratified analysis of the primary end point to check for site effects, using the Cochran and Mantel-Haenszel t test. The 2-sided P value of the test was .02 (OR, 0.56; 95% CI, 0.33-0.93).

Remote ischemic preconditioning significantly reduced the number of moderate and severe acute kidney injury cases compared with that of the control group (12.5% vs 25.8%; P鈥�=鈥�.02; RR, 85%; 95% CI, 75%-97%) but did not reduce the rate of mild acute kidney injury (25% vs 26.7%; P鈥�=鈥�.77; RR, 98%; 95% CI, 84%-114%). Use of renal replacement therapy (5.8% vs 15.8%; P鈥�=鈥�.01; absolute risk reduction, 10%; 95% CI, 2.25%-17.75%) and length of intensive care unit stay (3 days [interquartile range, 2-5] vs 4 days [interquartile range, 2-7]; 95% CI, 0-2 days, median difference; P鈥�=鈥�.04) were significantly reduced with remote ischemic preconditioning (Table 2). However, we found no significant differences between groups in time receiving mechanical ventilation, myocardial infarction, and perioperative stroke (Table 2). Length of hospital stay after surgery was comparable. All-cause in-hospital mortality and 30-day mortality were not different between groups (Table 2).

Although baseline urinary (TIMP-2) 脳 (IGFBP7) and NGAL, tested immediately before the intervention, did not differ between groups, the control group had significantly higher urinary (TIMP-2)鈥壝椻��(IGFBP7) at 4 hours after (remote ischemic preconditioning, 0.36 vs control, 0.97 ng/mL²/1000; difference, 0.61 [95% CI, 0.27-0.86]; 笔鈥�<鈥�.001) and 12 hours after cardiopulmonary bypass (P鈥�<鈥�.001) and higher NGAL at 4 hours after cardiopulmonary bypass (P鈥�=鈥�.04) compared with the remote ischemic preconditioning group (Figure 2A and B).

By contrast, remote ischemic preconditioning increased urinary (TIMP-2)鈥壝椻��(IGFBP7) immediately after remote ischemic preconditioning before cardiopulmonary bypass compared with that of the control group (Figure 2A), whereas urinary NGAL was unchanged (Figure 2B). Patients with urinary (TIMP-2)鈥壝椻��(IGFBP7) level greater than or equal to 0.5 ng/mL²/1000 before the initiation of the cardiopulmonary bypass had a significantly reduced rate of acute kidney injury compared with patients with lower urinary (TIMP-2)鈥壝椻��(IGFBP7) concentration (RR, 67%; 95% CI, 53%-83%; P鈥�<鈥�.001) (eFigure 1A in Supplement 2). However, patients with urinary (TIMP-2)鈥壝椻��(IGFBP7) greater than or equal to 0.5 ng/mL²/1000 4 hours after cardiopulmonary bypass had a significantly increased rate of acute kidney injury compared with patients with lower urinary (TIMP-2)鈥壝椻��(IGFBP7) (RR, 299%; 95% CI, 188%-473%; P鈥�<鈥�.001) (eFigure 1B in Supplement 2).

High-mobility group box 1, a damage-associated molecular pattern, was measured at baseline and after the intervention before cardiopulmonary bypass. Urinary HMGB-1 was similar in both groups at baseline. However, it significantly increased immediately after remote ischemic preconditioning (Figure 2C). In multivariable logistic regression analysis, preoperative serum creatinine level and previous heart surgery were associated with increased risk for acute kidney injury, whereas post鈥搑emote ischemic preconditioning HMGB-1 (OR, 0.75; 95% CI, 0.61-0.91; P鈥�=鈥�.005) and (TIMP-2)鈥壝椻��(IGFBP7) (OR, 0.57; 95% CI, 0.35-0.94; P鈥�=鈥�.03) were associated with lower risk for acute kidney injury (Table 3). Furthermore, both HMGB-1 and remote ischemic preconditioning were significant predictors of post鈥搑emote ischemic preconditioning (TIMP-2)鈥壝椻��(IGFBP7) 鈮モ��0.5 ng/mL²/1000 (Table 3).

The results of this multicenter, randomized, double-blind, clinical trial confirm the findings of a previous single-center study that remote ischemic preconditioning reduces the rate of acute kidney injury after cardiac surgery in high-risk patients.¹⁰ In our study, the intervention achieved more than a 15% absolute reduction in the rate of perioperative acute kidney injury. Especially the occurrence of moderate and severe acute kidney injury was reduced by remote ischemic preconditioning. We furthermore showed a benefit from remote ischemic preconditioning with a reduced use of renal replacement therapy and a shorter length of intensive care unit stay. Finally, we found that remote ischemic preconditioning reduced the post鈥揷ardiopulmonary bypass expression of biomarkers of acute kidney injury, including neutrophil gelatinase-associated lipocalin and the recently approved biomarker panel (TIMP-2)鈥壝椻��(IGFBP7). Remote ischemic preconditioning increased pre鈥揷ardiopulmonary bypass release of the 鈥渁larm鈥� markers (TIMP-2)鈥壝椻��(IGFBP7) while having no effect on the damage marker neutrophil gelatinase-associated lipocalin. This scenario is consistent with the known roles of TIMP-2 and IGFBP7 in the induction of G1 cell-cycle arrest, an epithelial defense mechanism.¹⁹

Several small feasibility and controlled clinical trials provided evidence that remote ischemic preconditioning can reduce myocardial injury during coronary bypass surgery,²⁰ during surgical repair of congenital heart defects,²¹ and before percutaneous coronary interventions.²² Two studies have reported a protective effect of remote ischemic preconditioning on renal function.^10,23 In contrast to these studies, 3 other trials failed to demonstrate renal protection with remote ischemic preconditioning (see eFigure 2 in the Supplement).^9,11,24 Our study provides new insight into the heterogeneity of treatment effect observed across these trials. Although the mechanisms responsible for the benefit of remote ischemic preconditioning are not completely understood, one possible explanation is that damage-associated molecular patterns released from the ischemic tissue engage self-protective mechanisms in the kidney such as cell-cycle arrest (see eFigure 3 in the Supplement). We measured HMGB-1, a well-known damage-associated molecular pattern, in urine and found that remote ischemic preconditioning resulted in increased release of this molecule. Early increases in HMGB-1 and (TIMP-2) 脳 (IGFBP7) were strongly associated with lower risk for acute kidney injury. However, not all patients responded to remote ischemic preconditioning with increased HMGB-1 release or increases in urine (TIMP-2) 脳 (IGFBP7). Future studies of this intervention might benefit from monitoring these biomarkers.

Differences in outcomes across remote ischemic preconditioning trials might also have been due to differences in study protocols, confounding comorbidities, anesthetic regimens, and surgical technique. Because ours is the first study to our knowledge to measure biomarkers, it is not possible to know whether previous trials with negative results failed to induce changes in these intermediate end points. In our protocol, patients with diabetes treated with sulfonylurea medications were excluded because these drugs inhibit adenosine triphosphate鈥搒ensitive potassium-channel conductance and may impede the effects of remote ischemic preconditioning.²⁵ Although volatile anesthetics might have a preconditioning effect,^26,27 we excluded the use of propofol because it may mitigate the effects of remote ischemic preconditioning.¹⁵ In our trial, we included only patients with a high risk for acute kidney injury, as identified by a Cleveland Clinic Foundation score greater than or equal to 6.¹⁴ We focused on this particularly high-risk patient population because 2 consensus conferences concluded that they would be most likely to benefit from remote ischemic preconditioning.²⁸

The pathophysiology of acute kidney injury is complex and still incompletely understood. New evidence suggests that adaptive responses by tubular epithelial cells to injurious signals are responsible for renal dysfunction and that renal inflammation and microcirculatory dysfunction further amplify these mechanisms.^7,29 Remote ischemic preconditioning induces the release of various molecules that appear to mediate the protective effect of this intervention.⁵ Here, we demonstrate that these mediators might be inducing G1 cell-cycle arrest in the kidney, as indicated by increased urinary (TIMP-2) 脳 (IGFBP7) after remote ischemic preconditioning. Cell-cycle arrest has been implicated in acute kidney injury,^30,31 and urinary (TIMP-2) 脳 (IGFBP7) has been shown to be predictive of acute kidney injury in patients undergoing cardiac surgery,¹⁸ as well as in general intensive care unit populations.³² However, cell-cycle arrest is a self-defense mechanism. When exposed to stress, epithelial cells may enter a short period of G1 cell-cycle arrest¹⁹ until the danger has passed or injury has been repaired. High-mobility group box-1 is an endogenous damage-associated molecular pattern molecule that can serve as an early mediator in the context of sterile inflammation, with release occurring as a consequence of acute cellular stress, hypoxia, or necrosis.³³ Extracellular HMGB-1 can bind to several pattern recognition receptors, including Toll-like receptors, which can directly or indirectly induce cell-cycle arrest.²⁹ Our data are in line with those of a recent animal study demonstrating that preconditioning with recombinant HMGB-1 provides protection against acute kidney injury.³⁴ We hypothesized that HMGB-1 (and other damage-associated molecular patterns) is released after remote ischemic preconditioning and these molecules induce cell-cycle arrest in tubular epithelial cells (see eFigure 3). Increases in urine (TIMP-2) 脳 (IGFBP7) immediately after remote ischemic preconditioning should therefore be protective from subsequent kidney injury induced by cardiac surgery, whereas late increases in these markers (for example, after cardiopulmonary bypass) should herald acute kidney injury. Our results fit this scenario exactly.

In cardiac surgery, perioperative acute kidney injury is closely associated with postoperative morbidity and mortality in the short and long term.^2,35-37 Several studies demonstrated an association between acute kidney injury and increased morbidity, short-term and long-term mortality, and use of resources in various patient populations.^2,35-41 This relationship holds true even with small increases of serum creatinine level for cardiac surgery patients.^2,35,39,41 Remote ischemic preconditioning could thus represent a simple and promising strategy to provide protection to the kidney and improve postoperative outcomes. Such measures would be particularly desirable to deal with the increasingly challenging risk profiles of patients who are referred for cardiac surgery.

Our study is not without limitations. Although this was a multicenter trial, it was adequately powered only to analyze prospectively the rate of perioperative acute kidney injury and thus a phase 2 equivalent study. The secondary end points, for which the study was not powered but which were assessed in view of a significant effect on the primary end point, indicated reduced kidney damage (by acute kidney injury stage, as well as by urinary (TIMP-2) 脳 (IGFBP7) and [neutrophil gelatinase-associated lipocalin]) in patients undergoing remote ischemic preconditioning. The use of renal replacement therapy was reduced in the intervention group as well. Although the critical care physicians treating the patients were blinded to the study group allocation, initiation of renal replacement therapy was at their discretion. Among critically ill patients with acute kidney injury, the timing of renal replacement therapy initiation remains an area of considerable controversy.¹³ Another limitation of this study is that although we have found important associations with intermediary end points, we cannot prove mechanism. Future experimental and clinical studies are needed to better establish the relationship between damage-associated molecular patterns, cell-cycle arrest, and rate or severity of acute kidney injury. Likewise, future studies will need to address the optimal methods for remote ischemic preconditioning and whether benefits are consistent across patients with various risks for acute kidney injury, such as those with preexisting chronic kidney disease or with lower Cleveland Clinic Foundation score. We did not detect a reduction in mortality between the 2 groups; as expected, this secondary end point is uncommon and our study was too small. According to our 30-day mortality results, we would need more than 4000 patients (183 deaths) to detect a difference in the mortality with 80% power. It remains to be determined whether preventing cardiac surgery鈥揳ssociated acute kidney injury with remote ischemic preconditioning will reduce morbidity, mortality, and use of resources other than renal replacement therapy.

Among high-risk patients undergoing cardiac surgery, remote ischemic preconditioning compared with control significantly reduced the rate of acute kidney injury and use of renal replacement therapy. The observed reduction in the rate of acute kidney injury and the need for renal replacement warrant further investigation.

Corresponding Author: Alexander Zarbock, MD, Department of Anesthesiology, Critical Care Medicine and Pain Therapy, University Hospital M眉nster, Albert-Schweitzer-Campus 1, Geb盲ude A1, 48149 M眉nster, Germany (zarbock@uni-muenster.de).

Published Online: May 29, 2015. doi:10.1001/jama.2015.4189.

Author Contributions: Drs Zarbock and Meersch had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Zarbock, Schmidt, Wempe, Martens, Haeberle, G枚rlich, Meersch.

Acquisition, analysis, or interpretation of data: Zarbock, Schmidt, Van Aken, Wempe, Zahn, Wolf, Goebel, Schwer, Rosenberger, G枚rlich, Kellum, Meersch.

Drafting of the manuscript: Zarbock, G枚rlich, Meersch.

Critical revision of the manuscript for important intellectual content: Zarbock, Schmidt, Van Aken, Wempe, Martens, Zahn, Wolf, Goebel, Schwer, Rosenberger, Haeberle, G枚rlich, Kellum, Meersch.

Statistical analysis: Zarbock, G枚rlich, Kellum, Meersch.

Obtained funding: Zarbock.

Administrative, technical, or material support: Zarbock, Schmidt, Wempe, Zahn, Wolf, Goebel, Schwer, Rosenberger, Kellum, Meersch.

Study supervision: Zarbock, Schmidt, Van Aken, Martens, Zahn, Wolf.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Zarbock reports receiving grant support and lecture fees from Astute Medical, unrelated to the current study. Dr Kellum reports receiving grant support and consulting fees from Astute Medical and Alere, unrelated to the current study. Drs Zarbock and Kellum have filed a patent application on the use of the biomarkers together with remote ischemic preconditioning.

Funding/Support: The study was funded by the German Research Foundation (ZA428/6-1 to Dr Zarbock).

Role of the Funder/Sponsor: The study sponsor had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.