Remote ischemic preconditioning has a neutral effect on the incidence of kidney injury after coronary artery bypass graft surgery

http://www.kidney-international.org

clinical trial

& 2014 International Society of Nephrology

Remote ischemic preconditioning has a neutral effect on the incidence of kidney injury after coronary artery bypass graft surgery Sean M. Gallagher1,2,3, Dan A. Jones1,2,3, Akhil Kapur1,2,3, Andrew Wragg1,3, Steve M. Harwood2, Rohini Mathur2, R. Andrew Archbold1,3, Rakesh Uppal2,3,4 and Muhammad M. Yaqoob2,3,5 1

Department of Cardiology, Barts Health NHS Trust, London, UK; 2William Harvey Research Institute, Queen Mary University, London, UK; NIHR Cardiovascular Biomedical Research Unit, London Chest Hospital, London, UK; 4Department of Cardiothoracic Surgery, Barts Health NHS Trust, London, UK and 5Department of Nephrology, Barts Health NHS Trust, London, UK

3

Acute kidney injury (AKI) is a frequent complication of cardiac surgery and usually occurs in patients with preexisting chronic kidney disease (CKD). Remote ischemic preconditioning (RIPC) may mitigate the renal ischemia–reperfusion injury associated with cardiac surgery and may be a preventive strategy for postsurgical AKI. We undertook a randomized controlled trial of RIPC to prevent AKI in 86 patients with CKD (estimated glomerular filtration rate under 60 ml/min per 1.73 m2) undergoing coronary artery bypass graft (CABG) surgery. Forty-three patients each were randomized to receive standard care with or without RIPC consisting of three 5-minute cycles of forearm ischemia followed by reperfusion. The primary end point was the development of AKI defined as an increase in serum creatinine concentration over 0.3 mg/dl within 48 h of surgery. Secondary end points included a comparison between the study and control groups of several serum biomarkers of renal injury including cystatin-C, neutrophil gelatinase–associated lipocalin (NGAL), and interleukin-18 (IL-18), and urinary biomarkers including NGAL, IL-18, and kidney injury molecule-1 measured at 6, 12, and 24 h after CABG, and the 72-h serum troponin T concentration area under the curve as a marker of myocardial injury. Clinical and operative characteristics were similar between the preconditioned and control groups. AKI developed in 12 patients in both groups within 48 h of CABG. There were no significant differences between the two groups in the concentrations of any of the serum or urinary biomarkers of renal or cardiac injury after CABG. Thus, RIPC induced by forearm ischemia–reperfusion had no effect on the frequency of AKI after CABG in patients with CKD.

Correspondence: Muhammad M. Yaqoob, Department of Translational Medicine and Therapeutics, William Harvey Research Institute, John Vane Building, Charterhouse Square, London EC1M 6BQ, UK. E-mail: [email protected] Received 10 January 2014; revised 31 May 2014; accepted 5 June 2014 Kidney International

Kidney International advance online publication, 30 July 2014; doi:10.1038/ki.2014.259 KEYWORDS: acute kidney injury; chronic kidney disease; ischemia–reperfusion; cardiovascular disease

Acute kidney injury (AKI) is a common problem following cardiac surgery. Depending upon the definition used, AKI complicates up to 45% of cardiac surgical procedures, with 1–6% of patients requiring renal replacement therapy following cardiac surgery.1–6 The development of AKI is associated with increased rates of short- and long-term postoperative mortality, prolonged intensive care unit and hospital stays, and increased resource utilization.1,4,7–9 The pathophysiology of AKI following cardiac surgery is complex and incompletely understood. Renal ischemia– reperfusion injury, however, is thought to be one of the principal contributors to a multifactorial renal insult, which also arises from the systemic inflammatory response to cardiopulmonary bypass (CPB), perioperative hemodynamic instability, and the use of perioperative nephrotoxins such as nonsteroidal anti-inflammatory drugs, angiotensinconverting enzyme inhibitors, iodinated contrast media, and nephrotoxic antibiotics.10 Currently, there are no prophylactic strategies proven to reduce the incidence of AKI in cardiac surgical patients. The application of brief periods of ischemia to an organ or tissue followed by its reperfusion has been shown to protect distant organs from the effects of subsequent episodes of ischemia–reperfusion. This phenomenon of remote ischemic preconditioning (RIPC) has been reported to reduce the incidence of AKI in preliminary trials of patients undergoing major vascular surgery,11 coronary angiography,12 and cardiac surgery.13 The single most important predictor for the development of AKI after cardiac surgery is preexisting chronic kidney disease (CKD).2,9,14 The prevalence of CKD in patients undergoing cardiac surgery is B25%.15,16 Whether or not RIPC prevents AKI after cardiac surgery in this high-risk 1

clinical trial

SM Gallagher et al: Effect of ischemic preconditioning on kidney injury after CABG

group of patients is not known. We investigated, in a prospective and randomized controlled trial, whether a forearm RIPC protocol reduces the incidence of AKI in patients with CKD undergoing coronary artery bypass graft (CABG) surgery. RESULTS

Of the 101 patients who were approached for trial inclusion, 11 did not wish to participate, 2 declined consent for cardiac surgery, 1 underwent ‘off-pump’ surgery, and 1 patient could not undergo the trial protocol for logistical reasons. The remaining 86 patients (69 men and 17 women) were randomized equally between the RIPC group (n ¼ 43) and the control group (n ¼ 43) (Figure 1). The baseline clinical characteristics of the study groups are shown in Table 1. There were no significant differences between groups in age, ethnicity, prior history, or drug therapy. The median estimated GRF and the median serum [Cr] were the same in both groups at 51 ml/min per 1.73 m2 and 1.37 mg/dl, respectively. The procedural characteristics were also similar, with 40 (93%) patients undergoing isolated CABG in the RIPC group compared with 41 (95.3%) in the control group (P ¼ 0.645), and similar frequencies of elective surgery, and similar numbers of grafts in the two groups (Table 2). The primary end point of postoperative AKI within 48 h of surgery occurred in 24 (27.9%) patients. The rate of AKI was the same in the RIPC group and in the control group, with 12 (27.9%) of 43 patients being affected in each (P ¼ 1.0). Seven (16.3%) patients in the RIPC group and 11 (25.6%) patients in the control group developed Acute Kidney Injury Network (AKIN) stage 1 AKI, three (7.0%) patients in the RIPC group and one (2.3%) patient in the control group developed AKIN stage 2 AKI, and two (4.6%) patients in the RIPC group and 0 (0%) patients in the control

101 Eligible patients within study period 11 Refused consent 1 Underwent off-pump surgery 1 Not enrolled for logistical reasons 2 Refused surgery 86 Consented for study

Randomization

43 RIPC

43 Control

Figure 1 | Trial flowchart. Enrolled patients are stratified by scheduled surgery and the presence of diabetes mellitus and then randomized to study groups. RIPC, remote ischemic preconditioning. 2

Table 1 | Baseline clinical characteristics Variable Demographics Age (years) Gender (male:female) BMI (kg/m2) Ethnicity Caucasian Afro-Caribbean South Asian

RIPC (n ¼ 43)

Control (n ¼ 43)

P-value

68.7±11.0 33:10 28.2±4.7

72.8±8.4 36:7 27.4±4.7

0.107 0.417 0.432

28 (65.1) 3 (7.0) 11 (25.6)

29 (67.4) 2 (4.7) 10 (23.3)

0.897

Diabetes mellitus Noninsulin-treated Insulin-treated

14 (32.6) 13 (30.2)

18 (41.9) 10 (23.3)

0.630

Hypertension Hypercholesterolemia Previous MI Previous stroke or TIA Peripheral vascular disease COPD

34 37 25 9 15 5

37 30 20 6 14 4

0.394 0.069 0.280 0.394 0.820 0.725

Clinical characteristics

Preoperative renal function Median eGFR (ml/min) Median serum [Cr] (mg/dl) CKD stage 3a 3b 4

(79.1) (86.0) (58.1) (20.9) (34.9) (11.6)

(86.0) (69.5) (46.5) (14.0) (32.6) (9.3)

51 (42–54) 51 (43–55) 1.37 (1.26–1.62) 1.37 (1.28–1.62)

0.952 0.746

29 (67.4) 11 (25.6) 3 (7.0)

28 (65.1) 11 (25.6) 4 (9.3)

0.923

Cardiac status Preoperative NYHA class 43 18 (41.8) Mean LVEF (%) 52.5±13.4

11 (25.6) 51.4±12.4

0.110 0.684

24 (55.8) 14 (32.6) 5 (11.6)

0.667

2 12 29 9

(4.7) (27.9) (67.4) (20.9)

0.331

0.579

42 6 39 16 39 36 19 2

(97.7) (14.0) (90.7) (37.2) (90.7) (83.7) (44.2) (4.7)

1.000 0.118 0.116 1.000 1.000 0.427 0.509 1.000

10 (23.3) 17 (39.5) 8 (18.6)

0.471 0.362 0.792

LVEF 455% 35–55% o35%

21 (48.5) 18 (41.9) 4 (9.3)

No. of diseased coronary arteries 1 1 (2.3) 2 7 (16.3) 3 35 (81.4) Left main stem disease 7 (16.3) Preoperative drug history Aspirin Clopidogrel b-blocker Ca2 þ channel blocker Lipid-lowering therapy ACE antagonist Long-acting nitrate Potassium channel opener Antidiabetic drugs Insulin Metformin Sulphonylurea

41 13 31 16 40 32 15 2

(95.3) (30.2) (72.1) (37.2) (93.0) (74.4) (35.0) (4.7)

14 (32.6) 12 (28.0) 10 (23.3)

Abbreviations: ACE, angiotensin-converting enzyme; BMI, body mass index; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; [Cr], creatinine; eGFR estimated glomerular filtration rate; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NYHA, New York Heart Association; RIPC, remote ischemic preconditioning; TIA, transient ischemic attack.

Kidney International

clinical trial

SM Gallagher et al: Effect of ischemic preconditioning on kidney injury after CABG

Table 2 | Procedural characteristics Variable Surgical procedure CABG CABG þ AVR

RIPC (n ¼ 43) 40 (93.0) 3 (7.0)

Table 3 | Postoperative outcomes stratified by study group Control (n ¼ 43) 41 (95.3) 2 (4.7)

P-value 0.645

Variable

RIPC (n ¼ 43)

Control (n ¼ 43)

P-value

Primary end point Frequency of AKI

12 (27.9)

12 (27.9)

1.000

7 (16.3) 3 (7.0) 2 (4.6) 2 (4.6)

11 (25.6) 1 (2.3) 0 (0) 0 (0)

0.270

Procedural urgency Elective Urgent

29 (67.4) 14 (32.6)

28 (65.1) 15 (34.9)

0.820

Isoflurane anesthesia Intermittent cross-clamp fibrillation

37 (86.0) 6 (14.0)

38 (88.4) 9 (20.9)

0.747 0.394

Classification of AKI AKIN stage 1 AKIN stage 2 AKIN stage 3 Requirement for postoperative RRT

Number of grafts 1 2 3 43

1 4 26 12

2 (4.7) 6 (14.0) 26 (60.5) 9 (20.9)

0.762

Secondary end points 72-h serum [TnT] AUC (ng/l)

Cross-clamp time (mins) Perfusion time (mins)

66 (49–90) 94 (78–119)

58 (45–77) 94 (74–123)

0.092 0.613

Fluid input during surgery Crystalloid PRBC Fluid input (0–24 h) Urine output (0–24 h) PRBC input (0–24 h)

(2.3) (9.3) (60.5) (27.9)

1993±1129 1826±376 0.302±0.118 0.279±0.096 2566±946 2600±949 0.97±0.29

3160±1583 2920±1340 1.32±0.25

0.875 0.879 0.202 0.274 0.152

0.154

34,686 31,269 0.367 (23,838–57,768) (22,374–41,958)

Other outcomes Inotrope use 31 (72.1) 34 (79.1) 0.451 Extubation time (mins) 483 (530–885) 720 (395–930) 0.239 ICU length of stay (hours) 22.4 (14.4–31.8) 22.0 (17.3–47.8) 0.242 Hospital length of stay (days) 8.0 (6–12) 8.0 (6–12) 0.512 30-Day mortality 2 (4.6) 2 (4.6) 1.000 Abbreviations: AKI, acute kidney injury; [Cr], creatinine; ICU, intensive care unit; RIPC, remote ischemic preconditioning; RRT, renal replacement therapy; [TnT] AUC, troponin T concentration area under the curve.

Abbreviations: AVR, aortic valve replacement; CABG, coronary artery bypass graft surgery; PRBC, packed red blood cells; RIPC, remote ischemic preconditioning.

group developed AKIN stage 3 AKI (P ¼ 0.270). Two (4.6%) patients in the RIPC group required renal replacement therapy compared with 0 (0%) patients in the control group (P ¼ 0.1524) (Table 3). Baseline clinical and procedural characteristics of the study groups were evenly matched after randomization, but this does not exclude an important effect of a baseline characteristic upon the primary end point. However, with univariable logistic regression, only estimated glomerular filtration rate was associated with subsequent development of AKI, although this relationship did not persist after multivariable logistic regression (using a model containing clinical factors consistently found to predict the development of AKI after cardiac surgery; namely age, gender, diabetes mellitus, left ventricle function, preoperative estimated glomerular filtration rate, surgical urgency, cross-clamp time, CPB time, and concurrent valve surgery). RIPC had no association with the subsequent development of AKI (data not shown). There was no significant difference between groups in serum [NGAL], [IL-18], or serum cystatin-C [CyC] measured at baseline, 6 h, 12 h, and 24 h after surgery (Table 4). Furthermore, urinary [NGAL]/creatinine ratio, urinary [IL-18]/creatinine ratio, and urinary [KIM-1]/creatinine ratio were not different between the groups when measured at baseline, 6 h, 12 h, and 24 h after surgery (Table 5). Myocardial biomarker release defined by 72 h AUC [TnT] was not significantly different between the groups (RIPC, 34686 [23838–57768] vs. control, 31269 [22374–41958] ng/L; P ¼ 0.3668) (Figure 2). Kidney International

There was no significant difference between groups in the proportion of patients who were treated with inotropic drugs or in the duration of mechanical ventilation postoperatively. Length of intensive care unit stay and hospital stay was 22.4 vs. 22.0 h (P ¼ 0.242) and 8.0 vs. 8.0 days (P ¼ 0.512) in the RIPC and control groups, respectively. Two patients in each group died within 30 days of surgery (Table 3). DISCUSSION

This randomized controlled study investigated whether or not RIPC induced by intermittent forearm ischemia in the anesthetic room immediately before cardiac surgery attenuated subsequent renal injury in patients with established CKD. We observed no effect from RIPC on the frequency of AKI within 48 h of surgery, or on the severity of AKI, whether assessed by AKIN criteria or by changes in the concentrations of several biomarkers for renal injury. Furthermore, RIPC had no effect upon serum [TnT], a biomarker for myocardial injury, or on clinical end points including length of intensive care unit stay, length of hospital stay, and 30-day mortality. AKI is a common problem following cardiac surgery; in this study, it occurred in 27.9% of patients. The development of AKI is associated with an increase in subsequent mortality.1,2,4,8,9 Mortality is highest in patients who require renal replacement therapy, and in one study of 3795 patients who underwent CABG or valve surgery 30-day mortality in this group exceeded 60%2 Smaller, transient postoperative increases in serum [Cr] without apparent clinical sequelae, however, are also associated with an increase in early and late mortality.4,8 The relation between AKI and excess mortality 3

clinical trial

SM Gallagher et al: Effect of ischemic preconditioning on kidney injury after CABG

Table 4 | Concentrations of serum biomarkers of renal injury at 6, 12, and 24 h after surgery Control (n ¼ 43)

P-value

Serum NGAL concentrations (ng/ml) Baseline 127.5 [74.5–179.5] 6 h post CABG 256.2 [158.0–364.8] 12 h post CABG 237.5 [159.6–407.9] 24 h post CABG 297.6 [166.8–440.4]

127.6 [79.5–178.0] 207.7 [157.7–318.0] 220.7 [174.5–314.0 277.7 [191.7–351.2]

0.8830 0.5804 0.5116 0.8022

Serum IL-18 concentrations (ng/ml) Baseline 18.92 [12.94–23.21] 6 h post CABG 14.26 [11.16–17.27] 12 h post CABG 23.17 [18.65–27.32] 24 h post CABG 30.76 [24.54–42.67]

18.75 [13.89–23.82] 15.13 [11.5–18.12] 24.15 [18.26–29.26] 32.2 [26.74–42.62]

0.8022 0.7955 0.3737 0.5687

Serum cystatin-C concentrations (ng/ml) Baseline 1376 [1163–1910] 6 h post CABG 1173 [938.5–1531] 12 h post CABG 1330 [874–1769] 24 h post CABG 1486 [1103–2268]

1359 [1187–1746] 1125 [913–1523] 1326 [1070–1654] 1452 [1155–1936]

0.7493 0.7493 0.7297 0.8089

Control

Table 5 | Concentrations of urinary biomarkers of renal injury at 6, 12, and 24 h after surgery

Urinary NGAL/urinary Baseline 6 h post CABG 12 h post CABG 24 h post CABG Urinary IL-18/urinary Baseline 6 h post CABG 12 h post CABG 24 h post CABG

Control (n ¼ 43)

P-value

creatinine ratios (ng/mg) 607.4 [269.5–2754] 633.6 [331.7–3100] 4616 [2917–10764] 3550 [1431–12333] 5673 [3250–11795] 5057 [1689–9851] 2783 [812.5–11191] 2760 [867.3–8647]

0.9518 0.7297 0.3691 0.8969

creatinine ratio (ng/mg) 1280 [522.2–2244] 881.7 [330.1–2291] 403.6 [148.8–897.3] 344.9 [96.26–969.5] 716.9 [197.8–3003] 534.1 [126.8–1750] 335.9 [75.2–980.1] 346.0 [154.4–16520]

0.4576 0.9589 0.2803 0.4788

Urinary KIM-1/urinary Baseline 6 h post CABG 12 h post CABG 24 h post CABG

creatinine (pg/mg) 988.1 [471.1–1530] 586.9 [305.3–1119] 1389 [767.8–2531] 1470 [931.8–2488]

908.7 [471–1463] 744.5 [338.3–1284] 1585 [732.1–2956] 1207 [695.1–3032]

0.9656 0.2615 0.7889 0.6409

Abbreviations: CABG, coronary artery bypass graft surgery; KIM-1, kidney injury molecule-1; NGAL, neutrophil gelatinase–associated lipocalin; RIPC, remote ischemic preconditioning.

persists after adjusting for potential confounding factors such as patient comorbidities and surgical complexity.4,8 Clinical factors that have been consistently implicated in the development of AKI following cardiac surgery include advanced age, diabetes mellitus, congestive heart failure, need for emergency surgery, surgical complexity, and preoperative CKD.2,6,14,17 The frequency of AKI is related to baseline renal function, being highest in patients with severe CKD, but those with moderate renal impairment also have a significantly higher rate of AKI than those without preoperative renal impairment. Among 2438 patients who underwent first-time elective CABG with or without valve surgery, for 4

RIPC

1000

500

P =0.3668

Abbreviations: CABG, coronary artery bypass graft surgery; IL-18, interleukin-18; NGAL, neutrophil gelatinase–associated lipocalin; RIPC, remote ischemic preconditioning.

RIPC (n ¼ 43)

Troponin T ng/l

RIPC (n ¼ 43)

1500

0 0

12

24

36 48 Time post CPB/h

60

72

Figure 2 | Cardiac troponin T release over 72 h. Median and interquartile ranges are presented. CPB, cardiopulmonary bypass.

example, AKI occurred in 12.8% of patients with baseline estimated GFR 460 ml/min per 1.73 m2 compared with 24.3% of patients with an estimated GFR of 30–59.9 ml/min per 1.73 m2 and 46.6% of patients with an estimated GFR o30 ml/min per 1.73 m2 (Po0.001).18 Recent improvements in the understanding of the etiology and pathophysiology of AKI after cardiac surgery have not been paralleled by decreases in its incidence or in the mortality rate associated with the condition,19 nor are there currently any established prophylactic strategies or treatments for AKI beyond general supportive measures.20 Effective prophylaxis and the improved clinical outcomes and reduced resource utilization which these would be expected to realize would be welcomed. RIPC induced using transient limb ischemia has been reported to reduce perioperative release of biomarkers for myocardial injury in adults and children undergoing cardiac surgery.21,22 RIPC has also been reported to afford renal protection in patients undergoing vascular surgery,11 coronary angiography with or without coronary intervention,12 and cardiac surgery.13 In the latter study, AKI within 48 h of cardiac surgery using CPB occurred in 12 (20.3%) of 59 patients who were randomized to RIPC and 28 (47.4%) of 59 patients randomized to control.13 RIPC is an attractive strategy for prophylaxis because it is easy to administer, safe, and cheap. Owing to their high rate of postcardiac surgery AKI, patients with CKD appear particularly likely to derive benefit from RIPC. Furthermore, patients with CKD comprise about one quarter of all patients who undergo cardiac surgery.15 Whether or not RIPC is effective prophylaxis against AKI after cardiac surgery in patients with CKD has not been specifically investigated. There are several reasons why CKD may render patients ‘resistant’ to the organ protection generated using RIPC. However, our research group found that the protective effect Kidney International

SM Gallagher et al: Effect of ischemic preconditioning on kidney injury after CABG

of RIPC was not attenuated by uremia in a rodent model of myocardial infarction.23 This finding stimulated our hypothesis that RIPC may offer important renal and myocardial protection in patients with CKD undergoing cardiac surgery. Not only did we find no effect of RIPC on the rate of AKI within 48 h of surgery, we were also unable to demonstrate any reduction in the proportion of patients who developed more severe AKI, as defined by AKIN criteria. AKI defined using these criteria has been associated with both short- and long-term complications after cardiac surgery.8 The AKIN criteria, however, are largely based upon perioperative changes in serum [Cr], which has several limitations as a measure of AKI. Serum [Cr] is affected by age, gender, diet, muscle mass, and some drugs; important changes in glomerular filtration may be masked, as up to 40% of creatinine clearance is due to the renal secretion of creatinine; serum [Cr] only becomes abnormal when 450% of glomerular filtering capacity is lost; and it may take up to 24 h following AKI for increases in serum [Cr] to become evident. The inherent limitations of using a creatinine-based diagnosis for AKI have stimulated the development of a number of novel urinary and serum/plasma biomarkers of renal injury. We chose to assess several of these novel serum (NGAL, IL-18, and CyC) and urinary (NGAL, IL-18, and KIM-1) biomarkers of renal injury as secondary end points. Many of these biomarkers have now been examined as predictors of AKI following CPB by post hoc analysis of cardiac surgical cohorts stratified by the subsequent development (or not) of AKI.24–27 Early analyses involving children undergoing CPB found these new biomarkers to perform well as early predictors of subsequent AKI. For example, both urinary and plasma NGAL measured in children between 2 and 6 h following cardiac surgery demonstrated an area under the receiver operating curve (AUC-ROC) of 40.9 for the subsequent development of AKI.28 IL-18 also performed impressively in the same pediatric cohort.29 However, pediatric cardiac surgical cohorts are relatively homogeneous, with little comorbidity and a known timing of renal injury. In more heterogeneous adult populations, or where the time of renal injury is poorly defined, the predictive performance of these biomarkers of renal injury has been proven to be limited.30 Most studies evaluating these biomarkers have excluded patients with CKD. Importantly, for the current study, the predictive performance of biomarkers of renal injury may be reduced in patients with CKD. Urinary NGAL has shown a lower AUCROC of between 0.6 and 0.7 for the prediction of AKI when studied in adult cardiac surgical cohorts including patients with CKD.31,32 The performance of biomarkers of renal injury to predict de novo AKI in patients with preexisting CKD is complicated by many factors; for example, there is a negative correlation between estimated glomerular filtration rate and serum NGAL and CyC. Following renal injury, blood concentration of these biomarkers becomes further elevated. Previously described biomarker cutoff values for predicting Kidney International

clinical trial

AKI may be inappropriate in this patient cohort. Furthermore, the release kinetics of these biomarkers in patients with CKD is poorly defined, and as a result the appropriate time windows for biomarker measurement following renal injury is unknown. When choosing to assess these biomarkers, we hoped that they would offer a sensitive and specific surrogate of renal injury that may enable the detection of small differences in renal outcomes. We found no effect of RIPC upon the postoperative levels of these biomarkers for renal injury during the first 24 h after cardiac surgery. The putative protective effect of RIPC is not limited to the kidneys. In this study, we assessed the degree of perioperative myocardial injury by serial serum [TnT] measurements in the first 72 h following surgery. Serial serum troponin concentration measurement is a standard end point in cardiac surgical RIPC studies.22 Furthermore, early postoperative serum [TnT] elevation is associated with increased early and late mortality following cardiac surgery.33 We found no significant difference in cardiac troponin release between patients who were allocated to RIPC compared with control. There are a number of potential explanations for the lack of organ protection afforded by RIPC in this study. First, comorbidities including hypertension, hyperlipidemia, and heart failure may render organs resistant to the protective effects of RIPC.34,35 These comorbidities are particularly prevalent in patients with CKD, and in our cohort 480% of patients had hypertension, 78% had hypercholesterolemia, and almost half had left ventricular impairment. Furthermore, diabetes mellitus, present in 64% of our study cohort, and antidiabetic medication may interfere with the signal transduction pathways believed to mediate the effects of RIPC.36 Second, although myocardial infarction within 7 days of surgery was a study exclusion, it is possible that the study included patients with subclinical episodes of myocardial ischemia without tissue infarction, which may confer subsequent organ protection during cardiac surgery, by mechanisms similar to RIPC, making any beneficial effect afforded by RIPC applied immediately before surgery impossible to detect. Third, the use of particular anesthetic and surgical techniques may influence whether or not RIPC provides an independent organ protective effect. Volatile anesthetics, propofol, and intravenous nitrates, for example, have all been reported to affect the perioperative organ protection afforded by RIPC,37,38 and the use of particular anesthetic regimes have often been cited as a potential reason for negative surgical RIPC studies.39–41

Study limitations

This was a single-blinded study, with patients, anesthetists, surgeons, and critical care staff blinded to study group allocation. For logistical reasons, researchers were not blinded. As RIPC is a binary intervention and the main renal and cardiac outcomes were all objective, defined by serum or urine test results, it is unlikely that this methodological issue could introduce any significant bias into the study. 5

clinical trial

SM Gallagher et al: Effect of ischemic preconditioning on kidney injury after CABG

The preconditioning stimulus used in this study was intermittent forearm ischemia. Previous studies that reported renal protection afforded by RIPC in patients undergoing vascular surgery11 and cardiac surgery13 used intermittent leg ischemia as the preconditioning stimulus. Whether the mass of tissue subjected to the preconditioning stimulus affects subsequent organ protection is unknown. Potentially, subjecting a larger mass of tissue to a preconditioning stimulus would yield greater subsequent organ protection. As patients with CKD may be ‘resistant’ to RIPC, intermittent leg ischemia could prove to be a more appropriate preconditioning stimulus. However, RIPC induced with intermittent leg ischemia has not been proven to be universally successful in reducing AKI following cardiac surgery.42 Even in children, who lack comorbid conditions and should be uniquely ‘sensitive’ to preconditioning, intermittent leg ischemia failed to reduce the incidence of postoperative AKI (in children undergoing cardiac surgery for complex congenital heart disease).43 Importantly, the preconditioning stimulus in our study (in terms of site, number of cuff inflations, duration of cuff inflation and interval of reperfusion between cuff inflations) was identical to that used in studies reporting reduction in postoperative troponin release with RIPC following cardiac surgery.22,44,45 Furthermore, intermittent forearm ischemia proved to be a sufficient stimulus to induce renoprotection in patients with CKD undergoing coronary angiography with or without subsequent coronary intervention (albeit 4 cycles of 5-min ischemia with 5 min of reperfusion were used).12 Whether the use of leg ischemia or the use of a greater number of, or longer cycles of, forearm ischemia and reperfusion might have produced a different outcome to this study in a potentially ‘RIPC resistant’ cohort can only be speculated upon. When estimating the sample size for this study, we assumed that the primary end point (AKI) would occur with a frequency of 48% in the control group, and that RIPC would reduce the incidence of AKI to 24%. Recruiting 40 pairs of patients would allow 90% power to detect an absolute reduction of 24% in the incidence of AKI. The actual incidence of AKI was far less than expected (27.9% in the control group), and there was no apparent decrease in the incidence of AKI in patients who received RIPC. As the trial was underpowered to detect a smaller difference in the incidence of AKI, we cannot provide a definitive statistical result. However, if we were to repower this study assuming an incidence of AKI of 28% in the control arm and that RIPC would reduce the incidence of AKI by 20% (to an expected incidence of 22.4% in the RIPC arm), then 1258 patients would be required in each arm of the study to achieve a statistical power of 90%. If the 86 patients in the current trial were considered as the first 86 patients recruited to this ‘repowered’ study, and an interim analysis performed at this point, then the probability of obtaining a statistically significant final result if the study were to continue to recruit all 2516 patients, conditional on this interim analysis, is only 6

2.3%. The conditional probability of rejecting our initial hypothesis (i.e., that RIPC reduces the incidence of AKI in patients with CKD undergoing cardiac surgery by 20%) is 97.7%. MATERIALS AND METHODS Study population Patients were eligible for the study if they were aged between 18 and 85 years, had established CKD, and were scheduled to undergo nonemergent CABG with or without aortic valve replacement. CKD was defined as a GFR o60 ml/min per 1.73 m2 estimated using the Modification of Diet in Renal Disease equation46 on two separate occasions preoperatively. For elective patients, these estimates were made in a preadmission clinic B2 weeks before surgery and on the day before surgery. For patients undergoing urgent surgery, GFR was estimated at hospital admission and on the day before surgery. Study exclusion criteria included myocardial infarction within 7 days of surgery, off-pump surgery, ‘re-do’ surgery, end-stage renal failure or the presence of a renal transplant, coronary angiography within 7 days of surgery, and presurgical AKI, defined as an increase in serum creatinine concentration ([Cr]) 40.3 mg/dl between the two preoperative measurements.47 Study protocol This was a single-center, single-blinded, prospective, randomized, placebo-controlled trial. The trial was approved by the local research ethics committee (MREC No. 10/H0703/92) and was registered with the Research and Development Registry of Barts Health NHS Trust, London, UK. Written informed consent was obtained from each study subject. The trial was conducted between February 2011 and April 2012. Consecutive patients were stratified by surgical procedure (isolated CABG or CABG and aortic valve replacement) and diabetes mellitus status, before randomization within strata in a 1:1 ratio to receive either standard perioperative care (control group) or standard perioperative care and RIPC. Patients, anesthetists, surgeons, and critical care teams were blinded to study group allocation, although investigators were not blinded. The protocol for achieving RIPC involved inducing forearm ischemia–reperfusion. This was undertaken after the induction of anesthesia by performing three 5-min cycles of upper arm ischemia, induced by inflating a 9-cm blood pressure cuff placed around the upper arm to a pressure of 50 mm Hg greater than the patient’s systolic blood pressure. Each cycle of ischemia was separated by a 5-min period of reperfusion, during which the blood pressure cuff was deflated. RIPC was conducted in the anesthetic room while patient monitoring, intravascular catheters, and a bladder catheter were being placed so that time from anesthetic induction to CPB was not prolonged in the RIPC group. Patients in the control group had a 9-cm blood pressure cuff placed on the upper arm (but not inflated) for 30 min. Immediately after the RIPC protocol was completed, patients underwent cardiac surgery. All operations were performed via median sternotomy and used standard nonpulsatile CPB with a membrane oxygenator and cardiotomy suction. The patient’s core temperature was allowed to drift down to 34 1C. The left internal mammary artery was the usual graft for the left anterior descending artery. Each distal anastomosis was completed with the use of either intermittent cross-clamp fibrillation or cardioplegia. Once all of the grafts were constructed, CPB was discontinued and core temperature was restored to normal. Kidney International

SM Gallagher et al: Effect of ischemic preconditioning on kidney injury after CABG

All patients received intraoperative and postoperative care at the discretion of the cardiac anesthetic and surgical teams.

Data collection Detailed demographic and clinical information was recorded prospectively at trial enrollment. This included age, sex, ethnicity, body mass index, New York Heart Association heart failure class, cardiac history including prior myocardial infarction and prior percutaneous coronary intervention, medical history including diabetes mellitus, hypertension, hypercholesterolemia, peripheral vascular disease, previous stroke or transient ischemic attack (TIA), and chronic obstructive pulmonary disease, preoperative drug therapy, preoperative renal function (including serum [Cr] and estimated GFR), left ventricular function, and anatomical severity of coronary artery disease. We also recorded the date of recent myocardial infarction, where relevant, and of most recent cardiac catheterization. Operative data recorded included procedural urgency, type of myocardial protection, cross-clamp time, perfusion time, use of internal mammary arterial grafts, number of coronary bypass grafts, and types of volatile anesthetics used. Elective surgery was defined as CABG in patients who were discharged from the hospital following coronary angiography and readmitted in a planned way for surgery at a later date. Urgent surgery was defined by the requirement for the patient to remain in hospital for CABG following coronary angiography. The following postoperative data were recorded: hourly urine output, 3-hourly hemodynamics and infusion volumes during the first 24 h, length of intensive care unit stay, hospital length of stay, and 30-day mortality. Samples for serum [Cr] were drawn less than 24 h before surgery and then daily for 5 days postoperatively unless the patient was discharged within this period. Serum [Cr] was measured using a buffered kinetic Jaffe reaction on an automated clinical chemistry analyzer (Roche Diagnostics, Burgess Hill, UK) (Roche Modular Analytics, Core unit, and Control unit). The intraassay coefficient of variation for [Cr] in our laboratory was 3.75%. Serum samples for biomarker analysis were collected immediately before surgery and at 6, 12, and 24 h postoperatively for measurement of the following serum biomarkers of renal injury: (1) neutrophil gelatinase–associated lipocalin [NGAL], (2) CyC, and (3) interleukin-18 [IL-18]. Serum was separated from blood by centrifugation at 1300g for 15 min and stored at  80 1C until assay. Urine samples for biomarker analysis were collected at the same time points, immediately before surgery, and at 6, 12, and 24 h postoperatively for measurement of the following urinary biomarkers of renal injury: (1) [NGAL], (2) [IL-18], and (3) kidney injury molecule-1 [KIM-1]. Urine samples were also centrifuged and stored at  80 1C until assay. Serum and urine [NGAL] were measured using a commercially available enzyme-linked immunosorbent assay kit (Human Lipocalin-2/NGAL Immunoassay; R&D Systems, Abingdon, UK). The intraassay coefficient of variation for [NGAL] was 7.2% in our laboratory. Serum and urine [IL-18] were measured using a commercially available enzyme-linked immunosorbent assay kit (Human IL-18 BPa Immunoassay; R&D Systems, Europe, Abingdon, UK). The intraassay coefficient of variation for [IL-18] was 11.9% in our laboratory. Serum [CyC] was measured using a commercially available enzyme-linked immunosorbent assay kit (Human Cystatin-C Immunoassay; R&D Systems). The intraassay coefficient of variation for [CyC] was 5.1% in our laboratory. Kidney International

clinical trial

Urinary [KIM-1] was measured using a commercially available enzyme-linked immunosorbent assay kit (TIM-1/KIM-1/HAVCR Immunoassay; R&D Systems). The intraassay coefficient of variation for [KIM-1] was 4.8% in our laboratory. All urinary biomarkers are reported as a ratio of urinary biomarker/urinary creatinine (ng/mg), which was calculated by dividing [biomarker] (ng/ml) by the urinary [Cr] (mg/ml). Samples were taken immediately before surgery and at 6, 12, 24, 48, and 72 h following surgery for the measurement of serum troponin T concentration ([TnT]), a biomarker for myocardial cell injury. Troponin T concentrations were measured using the commercially available Elecsys electrochemiluminescence immunoassay troponin T high-sensitivity assay kit. The intraassay coefficient of variation for [TnT] in our laboratory was 4.87%.

Study end points The primary end point was AKI, defined by an increase in serum [Cr] 40.3 mg/dl from baseline, within 48 h of cardiac surgery.47 The main secondary end points were comparisons between study groups of the serum biomarkers of renal injury—1) [NGAL], 2) [IL-18], and 3) [CyC]—and urinary biomarkers of renal injury—1) [NGAL], 2) [IL-18], and 3) [KIM-1]—all measured 6, 12, and 24 h after surgery. Other secondary end points included myocardial injury biomarker release defined by 72-hour serum [TnT] area under the curve (72 h AUC [TnT]).

Statistical analysis The sample size was determined based upon expected incidence of AKI. Pilot data upon 68 patients undergoing cardiac surgery with CPB reported an incidence of postoperative AKI (defined by an increase in serum [Cr] 40.3 mg/dl from baseline within 48 h of cardiac surgery) of 48%. RIPC reduced the incidence of AKI following surgery to 24%.48 Assuming a similar incidence of AKI within our control group, the sample size was estimated for a paired case–control study to look for a study rate difference of 24% (absolute) with a statistical power of 90% and the significance level declared at the two-sided 5% level. Using one control per case, we needed to recruit 40 pairs of patients. All data were analyzed according to the intention-to-treat principle. As the sample size was estimated for a paired case– control study, the primary end point was analyzed using the twosample test of proportions. Normality of distribution of continuous data was assessed using the Shapiro–Wilks test. Normally distributed continuous data are presented as mean (±standard deviation) and compared between groups using Student’s unpaired t-test. Nonnormally distributed continuous data are presented as median (interquartile range) and were compared using the Mann–Whitney U test. Categorical data are summarized using absolute values (percentage) and were compared using the Pearson w2 test or Fisher’s exact test, where appropriate. We undertook a predetermined analysis of the frequency of AKI in the RIPC group and the control group stratified by AKIN stage.47 An increase in serum [Cr] 40.3 mg/dl but o100% from baseline was classified as AKIN stage 1; an increase in serum [Cr] 100–200% from baseline was classified as AKIN stage 2; and increases in serum [Cr] 4200% from baseline were classified as AKIN stage 3. All statistical tests were two-tailed, and significance was defined by a P-value o0.05. Statistical analysis was undertaken with SPSS 19 (SPSS). Figure preparation was undertaken using GraphPad 7

clinical trial

SM Gallagher et al: Effect of ischemic preconditioning on kidney injury after CABG

Prism 5 for MacOS (GraphPad Software, San Diego, CA; http:// www.graphpad.com).

14.

15.

CONCLUSION

In this randomized controlled trial, we observed no additional renal or myocardial protection of RIPC induced by forearm ischemia–reperfusion beyond current standard anesthetic and surgical management in patients with CKD who underwent CABG.

16.

17. 18.

DISCLOSURE

All the authors declared no competing interests. 19.

ACKNOWLEDGMENTS

The authors thank colleagues in the cardiothoracic surgical department at Barts Health NHS Trust, including Ian Weir, Wael Awad, Kit Wong, Alex Shipolini, Stephen Edmondson and Kulvinder Lall, and colleagues at William Harvey Research Institute including Katie Qureshi. This study was funded by a project grant from Barts and the London Charity.

20.

21.

22.

AUTHOR CONTRIBUTIONS SMG, AK, and MMY conceived and designed the study. SMG and DAJ undertook the study protocol and acquired the study data. SMG, SMH, and MMY analyzed and interpreted the data. SMG, DAJ, AK, AW, RU, and MMY drafted the manuscript. SMG and MMY carried out statistical analysis. SMG, DAJ, AW, SMH AK, RU, RAA, and MMY critically revised the manuscript for important intellectual content.

23.

24.

25.

REFERENCES 1.

2. 3.

4.

5. 6.

7.

8.

9.

10. 11.

12.

13.

8

Brown J, Kramer R, Coca S et al. Duration of acute kidney injury impacts long-term survival after cardiac surgery. Ann Thorac Surg 2010; 90: 1142–1148. Chertow G, Lazarus J, Christiansen C et al. Preoperative renal risk stratification. Circulation 1997; 95: 878–884. Chertow G, Burdick E, Honour M et al. Acute kidney injury, mortality, length of stay, and costs in hospitalized patients. J Am Soc Nephrol 2005; 16: 3365–3370. Hobson C, Yavas S, Segal M et al. Acute kidney injury is associated with increased long-term mortality after cardiothoracic surgery. Circulation 2009; 119: 2444–2453. Karkouti K, Wijeysundera D, Yau T et al. Acute kidney injury after cardiac surgery: focus on modifiable risk factors. Circulation 2009; 119: 495–502. Wijeysundera D, Karkouti K, Dupuis J et al. Derivation and validation of a simplified predictive index for renal replacement therapy after cardiac surgery. JAMA 2007; 297: 1801–1809. Chertow G, Levy E, Hammermeister K et al. Independent association between acute renal failure and mortality following cardiac surgery. Am J Med 1998; 104: 343–348. Lassnigg A, Schmidlin D, Mouhieddine M et al. Minimal changes of serum creatinine predict prognosis in patients after cardiothoracic surgery: a prospective cohort study. J Am Soc Nephrol 2004; 15: 1597–1605. Thakar C, Arrigain S, Worley S et al. A clinical score to predict acute renal failure after cardiac surgery. J Am Soc Nephrol 2005; 16: 162–168. Rosner M, Okusa M. Acute kidney injury associated with cardiac surgery. Clin J Am Soc Nephrol 2006; 1: 19–32. Ali ZA, Callaghan CJ, Lim E et al. Remote ischemic preconditioning reduces myocardial and renal injury after elective abdominal aortic aneurysm repair: a randomized controlled trial. Circulation 2007; 116: I-98–I-105. Er F, Nia A, Dopp H et al. Ischemic preconditioning for prevention of contrast medium-induced nephropathy: randomized pilot RenPro Trial (Renal Protection Trial). Circulation 2012; 126: 296–303. Zimmerman RF, Ezeanuna PU, Kane JC et al. Ischemic preconditioning at a remote site prevents acute kidney injury in patients following cardiac surgery. Kidney Int 2011; 80: 1–7.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36. 37.

Brown J, Cochran R, Leavitt B et al. Multivariable prediction of renal insufficiency developing after cardiac surgery. Circulation 2007; 116: I139–I143. Chawla L, Zhao Y, Lough F et al. Off-pump vs. on-pump coronary artery bypass grafting outcomes stratified by preoperative renal function. J Am Soc Nephrol 2012; 23: 1389–1397. Cooper W, O’Brien SM, Thourani V et al. Impact of renal dysfunction on outcomes of coronary artery bypass surgery: results from the Society of Thoracic Surgeons National Adult Cardiac Database. Circulation 2006; 113: 1063–1070. Aronson D, Rayfield E. How hyperglycemia promotes atherosclerosis: molecular mechanisms. Cardiovasc Diabetol 2002; 1: 1. Charytan D, Yang S, Mcgurk S et al. Long and short-term outcomes following coronary artery bypass grafting in patients with and without chronic kidney disease. Nephrol Dial Transplant 2010; 25: 3654–3663. Hudson C, Hudson J, Swaminathan M et al. Emerging concepts in acute kidney injury following cardiac surgery. Semin Cardiothorac Vasc Anesth 2008; 12: 320–330. Patel NN, Rogers CA, Angelini GD et al. Pharmacological therapies for the prevention of acute kidney injury following cardiac surgery: a systematic review. Heart Fail Rev 2011; 16: 553–567. Cheung M, Kharbanda R, Konstantinov I et al. Randomized controlled trial of the effects of remote ischemic preconditioning on children undergoing cardiac surgery: first clinical application in humans. J Am Coll Cardiol 2006; 47: 2277–2282. Hausenloy D, Mwamure P, Venugopal V et al. Effect of remote ischaemic preconditioning on myocardial injury in patients undergoing coronary artery bypass graft surgery: a randomised controlled trial. Lancet 2007; 370: 575–579. Byrne C, Mccafferty K, Kieswich J et al. Ischemic conditioning protects the uremic heart in a rodent model of myocardial infarction. Circulation 2012; 125: 1256–1265. Haase M, Bellomo R, Story D et al. Urinary interleukin-18 does not predict acute kidney injury after adult cardiac surgery: a prospective observational cohort study. Crit Care 2008; 12: R96. Haase M, Bellomo R, Devarajan P et al. Accuracy of neutrophil gelatinaseassociated lipocalin (NGAL) in diagnosis and prognosis in acute kidney injury: a systematic review and meta-analysis. YAJKD 2009; 54: 1012–1024. Liangos O, Tighiouart H, Perianayagam MC et al. Comparative analysis of urinary biomarkers for early detection of acute kidney injury following cardiopulmonary bypass. Biomarkers 2009; 14: 423–431. Zhang Z, Lu B, Sheng X et al. Cystatin C in prediction of acute kidney injury: a systemic review and meta-analysis. Am J Kidney Dis 2011; 58: 356–365. Mishra J, Dent C, Tarabishi R et al. Neutrophil gelatinase-associated lipocalin (NGAL) as a biomarker for acute renal injury after cardiac surgery. Lancet 2005; 365: 1231–1238. Parikh C, Mishra J, Thiessen-Philbrook H et al. Urinary IL-18 is an early predictive biomarker of acute kidney injury after cardiac surgery. Kidney Int 2006; 70: 199–203. Endre Z, Pickering J, Walker R et al. Improved performance of urinary biomarkers of acute kidney injury in the critically ill by stratification for injury duration and baseline renal function. Kidney Int 2011; 79: 1119–1130. Koyner J, Bennett M, Worcester E et al. Urinary cystatin C as an early biomarker of acute kidney injury following adult cardiothoracic surgery. Kidney Int 2008; 74: 1059–1069. Wagener G, Gubitosa G, Wang S et al. Urinary neutrophil gelatinaseassociated lipocalin and acute kidney injury after cardiac surgery. Am J Kidney Dis 2008; 52: 425–433. Domanski M, Mahaffey K, Hasselblad V et al. Association of myocardial enzyme elevation and survival following coronary artery bypass graft surgery. JAMA 2011; 305: 585–591. Ferdinandy P, Schulz R, Baxter GF. Interaction of cardiovascular risk factors with myocardial ischemia/reperfusion injury, preconditioning, and postconditioning. Pharmacol Rev 2007; 59: 418–458. Lavi S, Lavi R. Conditioning of the heart: from pharmacological interventions to local and remote protection: possible implications for clinical practice. Int J Cardiol 2011; 146: 311–318. Whittington H, Babu G, Mocanu M et al. The diabetic heart: too sweet for its own good? Cardiol Res Pract 2012; 2012: 845698. Kokita N, Hara A, Abiko Y et al. Propofol improves functional and metabolic recovery in ischemic reperfused isolated rat hearts. Anesth Analg 1998; 86: 252–258. Kidney International

SM Gallagher et al: Effect of ischemic preconditioning on kidney injury after CABG

38.

39.

40.

41.

42.

43.

Peters J. Remote ischaemic preconditioning of the heart: remote questions, remote importance, or remote preconditions? Basic Res Cardiol 2011; 106: 507–509. Kottenberg E, Thielmann M, Bergmann L et al. Protection by remote ischemic preconditioning during coronary artery bypass graft surgery with isoflurane but not propofol - a clinical trial. Acta Anaesthesiol Scand 2012; 56: 30–38. Lucchinetti E, Bestmann L, Feng J et al. Remote ischemic preconditioning applied during isoflurane inhalation provides no benefit to the myocardium of patients undergoing on-pump coronary artery bypass graft surgery: lack of synergy or evidence of antagonism in cardioprotection? Anesthesiology 2012; 116: 296–310. Rahman I, Mascaro J, Steeds R et al. Remote ischemic preconditioning in human coronary artery bypass surgery: from promise to disappointment? Circulation 2010; 122: S53–S59. Choi Y, Shim J, Kim J et al. Effect of remote ischemic preconditioning on renal dysfunction after complex valvular heart surgery: A randomized controlled trial. J Thorac Cardiovasc Surg 2011; 142: 148–154. Pedersen K, Ravn H, Povlsen J et al. Failure of remote ischemic preconditioning to reduce the risk of postoperative acute kidney injury in children

Kidney International

44.

45.

46.

47.

48.

clinical trial

undergoing operation for complex congenital heart disease: a randomized single-center study. J Thorac Cardiovasc Surg 2012; 143: 576–583. Thielmann M, Kottenberg E, Kleinbongard P et al. Cardioprotective and prognostic effects of remote ischaemic preconditioning in patients undergoing coronary artery bypass surgery: a single-centre randomised, double-blind, controlled trial. Lancet 2013; 382: 597–604. Venugopal V, Hausenloy D, Ludman A et al. Remote ischaemic preconditioning reduces myocardial injury in patients undergoing cardiac surgery with cold-blood cardioplegia: a randomised controlled trial. Heart 2009; 95: 1567–1571. Levey A, Bosch J, Lewis J et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999; 130: 461–470. Mehta R, Kellum J, Shah S et al. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 2007; 11: R31. Zimmerman R, Ezeanuna P, Conner Kane J et al. Remote ischemic preconditioning prevents acute kidney injury during cardiopulmonary bypass. American Society of Nephrology, Kidney Week. 27 October–1 November 2009. San Diego, USA.

9