The Association Between Pulsatile Cardiopulmonary Bypass and Acute Kidney Injury After Cardiac Surgery: A Before-and-After Study

ARTICLE IN PRESS Journal of Cardiothoracic and Vascular Anesthesia 000 (2019) 16

Contents lists available at ScienceDirect

Journal of Cardiothoracic and Vascular Anesthesia journal homepage: www.jcvaonline.com

Original Article

The Association Between Pulsatile Cardiopulmonary Bypass and Acute Kidney Injury After Cardiac Surgery: A Before-and-After Study Tim G. Coulson, FANZCA, PhD*,y,z,1, Eve McPhilimey, MSc, ACP*, Florian Falter, MD, FRCA, FFICM, PhD*, Yasir Abu-Omar, MBChB, DPhil (OXON), FRCS*, Andrew A. Klein, MBBS FRCA, FFICM* *

Royal Papworth Hospital, Cambridge, UK y Austin Health, Melbourne, Australia z Centre for Integrated Critical Care, University of Melbourne, Melbourne, Australia

Objectives: To investigate the association between pulsatile perfusion and cardiac surgeryassociated acute kidney injury. Design: An uncontrolled, retrospective before-and-after study. Setting: Single tertiary hospital. Participants: A total of 2,489 patients undergoing cardiac surgery with cardiopulmonary bypass (CPB). Interventions: Pulsatile versus nonpulsatile perfusion. Measurements and Main Results: Data for nonpulsatile perfusion was collected from April 1, 2016, to March 31, 2017 (n = 1,223). A practice change to universal pulsatile CPB occurred on April 3, 2017. Data for pulsatile perfusion was collected from May 1, 2017, to June 30, 2018 (n = 1,266). The primary outcome was the incidence of acute kidney injury (AKI) after cardiac surgery. Multivariable analysis was carried out to adjust for known confounders. Secondary outcomes included AKI stage, stroke, length of stay, and mortality. Subgroup analyses were carried out using prolonged CPB and chronic kidney disease. The primary outcome, incidence of AKI, did not differ between the nonpulsatile control group and the pulsatile group (23.9% v 25.4%, p = 0.392). The pulsatile group was not associated with AKI in the multivariable analysis (Odds ratio 1.09, p = 0.413). There were no differences in stages of AKI in the nonpulsatile group v pulsatile group (13.6% v 14.9%, 2.9% v 4.3%, and 7.4% v 6.1% for stages 1, 2, and 3, respectively, p = 0.12). There were no differences in subgroup analyses or secondary outcomes. Conclusions: There was no association found between kidney injury and pulsatile perfusion. It is likely that there is either no association between pulsatile perfusion and reduced kidney injury or that the difference is extremely small. Ó 2019 Elsevier Inc. All rights reserved. Key Words: pulsatile; perfusion; acute kidney injury; cardiac surgery

ACUTE KIDNEY INJURY (AKI) is common after cardiac surgery. Incidence depends on criteria used for diagnosis and the population studied, ranging from 5% to 42%.1,2 The most This research did not receive any specific grant for funding from either the public, commercial or not-for-profit sectors 1 Address reprint requests to Tim G. Coulson, Department of Anaesthesia, Austin Health, Heidelberg, Melbourne, VIC 3084. E-mail address: [email protected] (T.G. Coulson). https://doi.org/10.1053/j.jvca.2019.05.021 1053-0770/Ó 2019 Elsevier Inc. All rights reserved.

widely used definitions of AKI are the Acute Kidney Injury Network (AKIN) and risk, injury, failure, loss, end-stage kidney disease (RIFLE) definitions and are based on urine output or serum creatinine changes. The Kidney Disease: Improving Global Outcomes (KDIGO) criteria amalgamate these two to provide the current standard of diagnosis, and the best predictor of poorer outcomes. There are multiple influences on urinary output in the postcardiac surgery period. KDIGO criteria

ARTICLE IN PRESS 2

T.G. Coulson et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2019) 16

based on creatinine only therefore may be the most appropriate measure of AKI after cardiac surgery.1,3 AKI after cardiac surgery may be graded according to severity, from stage 1 (the least severe) to stage 3 (the most severe, including patients who are commenced on renal replacement therapy). All grades of AKI have been shown to be independently associated with worse long-term mortality.4 Even a small decline in renal function is associated with an increase in long-term mortality, which persists regardless of recovery before discharge.5 This association may represent unaccounted-for confounders.6 However, it also may be an indicator of other injurious processes in the perioperative time, or reflect a continuum between AKI and long-term renal injury.7 As such, AKI after cardiac surgery presents a potential target for therapy. AKI after cardiac surgery is a multifactorial process. The majority of risk factors are not modifiable. They include many components commonly used to predict risk of both long and short-term mortality in cardiac surgical patients (for example age, sex, hypertension, diabetes, and peripheral vascular disease). In contrast to other surgical procedures, cardiac surgery is reasonably unique in that in addition to the inflammatory response and embolic events seen in other vascular surgical procedures, there is also very often a requirement for extracorporeal circulation (cardiopulmonary bypass [CPB]). This results in additional inflammation, oxidative stress and hemolysis that may contribute to AKI.1 The majority of modern CPB circuits provide continuous, nonpulsatile flow by default. This nonpulsatile flow has been implicated in AKI after cardiac surgery. Pulsatile flow appears to provide better microvasular flow, reduced inflammatory response, and possibly improved organ function parameters when compared with nonpulsatile flow.8-11 There also may be a reduced need for hemofiltration in higher risk patients undergoing cardiac surgery.12 To date, most studies examining pulsatile CPB have been small or observational. In this study the authors aimed to investigate whether pulsatile CPB is associated with reduced rates of AKI after cardiac surgery in patients undergoing cardiac surgery using CPB. In April 2017, the authors’ institution decided to change the CPB protocol for CPB cases to pulsatile CPB. This decision was based on the favorable risk-benefit profile described earlier (improved microvascular flow, improved renal perfusion, and reduced renal injury) and seen in studies to date.8,12 There were no other changes in practice or surgical population between these 2 periods that the authors were aware of. The authors therefore subsequently performed a before-and-after study to compare outcomes between nonpulsatile and pulsatile CPB. The authors’ hypothesis was that pulsatile perfusion would be associated with reduced AKI rate. Methods Study Population The study took place at a single tertiary cardiac surgery center in the United Kingdom. Data were collected from April 1, 2016, to March 31, 2017. This group was defined as the nonpulsatile group. The change in perfusion to the routine use of

pulsatile CPB occurred on April 3, 2017. One month was allowed for the change in practice to become consistent during which no data was collected. Data from patients in the pulsatile group were collected from May 1, 2017, to June 30, 2018. The study was approved by the hospital audit committee. Inclusion criteria were patients undergoing coronary artery bypass grafting (CABG), valve repair or replacement, and combined CABG and valve repair or replacement requiring CPB. Organ transplantation, pulmonary thromboendarterectomy, major aortic surgery, and other procedures were excluded. Two surgeons were already using pulsatile CPB before the institutional change, therefore all data from these surgeons was excluded from the study both before and after the institutional change. Additional exclusions were patients with a body surface area (BSA) > 2.2 calculated using the Du Bois formula (for whom pulsatile CPB would be technically problematic because of the excessive circuit pressures needed to maintain adequate pump flow), patients undergoing minimally invasive extracorporeal circulation (as the circuit lacks a venous reservoir and blood-air interface) and patients who already had a balloon pump (which provided pulsatility during cross-clamp time). CPB Technique The CPB circuit in this study comprised of a hard-shell venous reservoir with either a Sorin 8F inspire oxygenator (LivaNova, Germany) or an Affinity oxygenator (Medtronic), half-inch pump, vacuum-assisted venous drainage device (Maquet, Germany), and St€ockert s5 heart-lung machine (St€ockert, Germany). Circuit prime consisted of 1,200 mL Hartmann’s Solution, 300 mL 10% mannitol, and 5,000 IU heparin. After 300 IU/kg heparinization, cannulation was performed and CPB was initiated with target flows of 2.4 L/min/m2. Cardiac arrest was achieved by 1,000 to 2,000 mL induction dose of cardioplegia (Harefield solution) in a 4:1 ratio with warm or cold blood as per surgical preference. Patients were cooled to 30˚C to 34˚C, or normothermia was maintained, and systemic arterial pressures were maintained between 50 and 80 mmHg according to clinician preference. Hemoglobin concentration was maintained at greater than 7 g/dL. Anesthesia was maintained using a combination of propofol and volatile anesthetic agents according to clinician preference. For patients in the pulsatile CPB group, pulsatile flow was instigated shortly after aortic cross-clamp application and terminated on cross-clamp removal. Pulsatile flow was generated by the pulse mode function on the arterial roller pump of the S5 heart-lung machine. Pulse frequency was set to 50 pulses per minute. The pulse width was set at 50%, which is the percentage of time within a pulse cycle that the blood is pumped at high speed. The base flow, which corresponds to the flow rate during the high-speed phase, was set at 40%. The greater the percentage base flow the lower the flow at the high-speed phase. The pulse settings used were based on “in-house” trials conducted in wet lab experiments that measured the circuit line pressure and flow rate during a simulated CPB case (Papworth perfusion team, 2016). The setting selected generated pulsatile flow at a wide

ARTICLE IN PRESS T.G. Coulson et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2019) 16

range of flow rates within the maximum threshold of the circuit line pressure and therefore could be used on most patients. Weaning from CPB and inotropic support was provided according to institutional protocol. Heparin was reversed after separation using protamine. Once adequate hemostasis and chest closure was achieved, patients were transferred to the intensive care unit.

3

significant. A power of 80% and a significance of 0.05 were chosen to determine sample size. This gave a required population of 1,134 patients in each group. The primary outcome and subgroup analyses of categorical variables were carried out using the chisquare test. Logistic regression was used to control for confounding factors. A p value of <0.05 was considered significant. Type I error was accounted for using Bonferroni correction. All analyses were carried out using Stata 12.0.16

Outcomes The primary outcome of the study was the incidence of AKI after cardiac surgery in the pulsatile CPB group versus the nonpulsatile CPB group, as defined using the KDIGO criteria (Table 1). A multivariable model was constructed to account for known confounders.14,15 Variables included age, sex, New York Heart Association status, ejection fraction, chronic obstructive pulmonary disease, diabetes mellitis, previous cardiac surgery, case priority, procedure type, preoperative creatinine, CPB time, extra-arterial arteriopathy, and hypertension. AKI after cardiac surgery was diagnosed on laboratory criteria (creatinine change) alone.3 Diagnosis of AKI was limited to 7 days after surgery.1 A creatinine rise after surgery (from the individual patient baseline value) consistent with AKI was flagged automatically by laboratory software, and creatinine levels were checked against preoperative creatinine to generate an AKI stage. A confirmatory manual check was carried out on these results, comparing them to manually entered preoperative creatinine levels. Secondary outcomes included CSA-AKI stages, stroke rate, mortality, and length of stay.14 A subgroup analysis was carried out to compare the incidence of AKI in patients with Chronic Kidney Disease (CKD) stage 3 or above and in patients with more prolonged CPB (>120 minutes and >180 minutes). Two surgeons were already using pulsatile perfusion. AKI rates were compared between the 2 study periods for this subset to assess for temporal trends in AKI rate unrelated to the practice change.

A total of 4,067 patients underwent cardiac surgery during the study period. Of these, 507 were excluded owing to a body surface area > 2.2 m2 or no CPB used. Six hundred fifteen patients who were operated on by surgeons already routinely using pulsatile CPB were also excluded. This left 2,945 patients. A further 336 patients were excluded who underwent aortic surgery or other major cardiac surgery not classified as CABG or valvular procedures. One hundred twenty patients were excluded who had minimally invasive extracorporeal circulation, preoperative intra-aortic balloon pump (IABP), or extracorporeal membrane oxygenation. The final combined database contained 2,489 patients, of whom 1,223 were in the nonpulsatile CPB group and 1,266 were in the pulsatile CPB group (Fig 1). Patient characteristics for each group are shown in Table 2. The only difference was a lower rate of diabetes in the pulsatile group. The primary outcome, incidence of AKI after cardiac surgery, did not differ between the nonpulsatile control group and the pulsatile group (23.9% v 25.4%, p = 0.392), and there was no association between the pulsatile group and AKI in the multivariable model (Tables 3 and 4). There were no differences in stages of CSA-AKI in the nonpulsatile group versus pulsatile group

All general cardiac surgical paents in study period N=4067

Statistical Methods A power calculation was carried out using a historical AKI rate of approximately 25% using institutional audit data. An absolute reduction in CSA-AKI of 5% was deemed to be clinically Table 1 KDIGO Criteria for Diagnosing AKI in Adults in This Study (Based on Serum Creatinine Concentration Only)21 AKI Stage

Serum Creatinine

1

1.5-1.9 times baseline OR >> 26.5 umol/L increase 2.0-2.9 times baseline 3.0 times baseline OR Increase to >>353 umol/L OR Initiation of renal replacement therapy

2 3

Results

Abbreviations: AKI, acute kidney injury; KDIGO, Kidney Disease Improving Global Outcomes.

BSA > 2.2 or no cardiopulmonary bypass (CPB) used N = 507

All CPB paents N = 3560

Parcipang surgeons only N = 2945

Surgeons already using pulsale CPB excluded N=615

Aorc and other surgeries excluded N=336 CABG, VR and combined procedures N = 2609 IABP and ECMO paents removed N=120 Final database N = 2489

Fig 1. Flow diagram describing patient population and exclusions.

ARTICLE IN PRESS 4

T.G. Coulson et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2019) 16

Table 2 Baseline Characteristics in Each Group Nonpulsatile Bypass (n = 1,223) Age (y) Sex, female Procedure type: CABG VR CABG + VR Urgency: Elective surgery Urgent Emergency surgery Ejection Fraction: Good Moderate Poor NYHA Status: 1 2 3 4 IDDM Preoperative creatinine (umol/L) Bypass time (min) EuroSCORE I (logistic predicted mortality risk) Previous cardiac surgery Cleveland clinic score

Table 3 CSA-AKI and Other Outcomes by Group Pulsatile Bypass (n = 1,266)

p Value

72 (64-78) 391 (32.0%)

72 (64-78) 417 (32.9%)

0.72 0.61

527 (43.1%) 489 (40.0%) 207 (16.9%)

541 (42.7%) 520 (41.1%) 205 (16.2%)

0.82

819 (67.0%) 379 (31.0%) 25 (2.04%)

828 (65.4%) 398 (31.4%) 40 (3.16 %)

0.199

923 (75.5%) 264 (21.6%) 36 (2.94%)

924 (73.0 %) 298 (23.5%) 44 (3.48%)

0.347

243 (19.9%) 584 (47.8%) 344 (28.1%) 52 (4.25%) 84 (6.87%) 82 (70-96)

294 (23.2%) 575 (45.4%) 357 (28.2%) 40 (3.16%) 61 (4.82%) 84 (70-100)

0.113

0.029 0.078

89 (73-112) 4.81 (2.52-8.67)

87 (70-112) 4.63 (2.44-8.64)

0.283 0.869

52 (4.25%) 4 (3-5)

44 (3.48%) 4 (3-5)

0.315 0.752

NOTE. Median (interquartile range) or n (%). Abbreviations: CABG, coronary artery bypass graft; IDDM, insulin dependent diabetes mellitis; NYHA, New York Heart Association; VR, valve repair/ replacement.

(Table 3). Subgroup analyses in patients with CKD stage 3 or greater, and more prolonged perfusion times showed no difference in the incidence of AKI after cardiac surgery between the groups. There was no difference in need for renal replacement therapy among patients with preexisting CKD3 or greater (Table 3). There was also no difference in stroke rate or mortality (Table 3). There was a small difference in length of stay that was insignificant after Bonferroni correction. There were 375 cases undertaken in the group of surgeons already using pulsatile perfusion routinely. AKI rate was higher in this group compared with the main group of surgeons in the study (n = 146 of 375 or 38.9% v n = 613 of 2489 or 24.6%, p < 0.001). There was no difference in AKI rate between the 2 study periods for these surgeons (42.9% v 35.2%, p = 0.13).

Discussion Main Findings In this before-and-after study, there was no association between the introduction of pulsatile CPB as standard of care and rates of AKI after cardiac surgery. Adjustment for possible confounders also demonstrated no evidence of benefit.

Outcome

Nonpulsatile

Pulsatile

All patients (n = 2,489): AKI (any stage) No AKI AKI (stage 1) AKI (stage 2) AKI (stage 3) Preexisting CKD stage 3 (n = 602): AKI (any stage) AKI (stage 2 or 3) Postoperative renal replacement therapy Perfusion time > 2 h AKI (any stage) Perfusion time > 3 h AKI (any stage) Other outcomes: Postoperative stroke 30-day mortality Median length of hospital stay (IQR)

Total n = 1,223 292 (23.9%) 931 (76.1%) 166 (13.6 %) 36 (2.9%) 90 (7.4%) Total n = 287

Total n = 1,266 321 (25.4%) 945 (74.6%) 189 (14.9%) 55 (4.3%) 77 (6.1%) Total n = 315

131 (45.6%) 61 (21.3%) 47 (16.4%)

141 (44.8%) 63 (20.0%) 36 (11.4%)

Total n = 244 77 (31.6%) Total n = 47 23 (48.9%)

Total n = 247 87 (35.2%) Total n = 44 17 (38.6%)

19 (1.6%) 18 (1.5%) 9 (7-14)

13 (1.0%) 17 (1.3) 9 (6-13)

p Value

0.392 0.120

0.828 0.704 0.079

0.389 0.323 0.244 0.785 0.027

NOTE. Median (interquartile range) or n (%). Abbreviations: CKD, chronic kidney disease; CSA-AKI, cardiac surgery associated acute kidney injury; IQR, interquartile range.

Findings in Context No prior large studies have been conducted examining the association between pulsatile perfusion and renal outcomes. In recent years, 2 meta-analyses have been published examining the effects of pulsatile perfusion on creatinine levels, biomarkers, and renal injury.17,18 The first included pediatric and adult patients and found no difference except for lactate levels and differences in creatinine clearance.18 The second included adult patients only and some additional studies. Overall, they noted reduced acute renal injury and improved creatinine clearance. The majority of these studies included 50 or fewer patients. The studies that appeared to drive the results of this meta-analysis included 2 of the larger studies by Serraino and Onorati.8,13 Notably, both Serraino and Onarati’s studies used an IABP to generate pulsatility. The studies were therefore limited to patients who already had an indication for an IABP. These patients were excluded from the present study. They also both describe very similar results from the same academic department and same time period, and it is unclear to what extent these included the same patients. Furthermore, renal injury findings were driven by the subgroup who had stage 3 CKD. Patients in the nonpulsatile group with CKD3 had a 70% AKI rate, much higher than that found in this study. It appears likely that the differences in both the population studied and the intervention delivered are responsible for the differences between the authors’ study and meta-analyses. Other possibilities include the potential for confounding owing to the authors’ study design and limitations (discussed later) and the possibility of publication bias and duplication affecting the

ARTICLE IN PRESS T.G. Coulson et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2019) 16 Table 4 Logistic Model Using AKI as Dependent Variable and other Risk Factors as Independent Variables Outcome

Variable

Odds Ratio

Lower CI

Upper CI

p Value

5

by confounding factors. Similarly, there may be a subset of patients at high risk of AKI who would gain a small benefit from pulsatile perfusion, but too few of these patients were present in this study to detect this difference. Strengths and Limitations

AKI (any stage): Pulsatile CPB 1.09 Age 1.04 Female 1.10 NYHA status (reference 1): 2 1.09 3 1.35 4 1.51 EF grade (reference “good”): Moderate 1.27 Poor 1.52 COPD 1.06 Diabetes (reference none): Oral therapy 1.53 Insulin therapy 2.81 Previous cardiac 1.94 surgery Surgical priority (reference elective): Urgent 1.20 Emergency 2.00 Preoperative 1.01 creatinine CPB time 1.004 Extra-arterial 1.62 arteriopathy Hypertension 1.24

0.89 1.02 0.87

1.33 1.05 1.38

0.413 <0.001 0.416

.82 1.00 0.88

1.43 1.83 2.58

0.563 0.054 0.133

1.01 0.91 0.76

1.60 2.53 1.49

0.043 0.109 0.718

1.14 1.90 1.20

2.06 4.16 3.17

0.004 <0.001 0.007

0.96 1.08 1.01

1.50 3.71 1.02

0.117 0.028 <0.001

1.002 1.18

1.007 2.24

0.002 0.003

1.00

1.54

0.054

NOTE: Grades according to EuroSCORE1 grading (n = 2,422). Abbreviations: AKI, acute kidney injury; COPD, chronic obstructive pulmonary disease; CPB, cardiopulmonary bypass; EF, ejection fraction; NYHA, New York Heart Association status;

meta-analyses. Since these meta-analyses, 1 small observation study has shown no significant difference in outcome,12 and another found an improvement in creatinine clearance.19

Implications of Findings To the authors’ knowledge, this study represents the largest comparison of renal outcomes in pulsatile versus nonpulsatile perfusion to date. A number of possibilities exist. First, that pulsatile perfusion does not affect renal outcomes, either because it has no beneficial effect or because its beneficial effects are balanced by negative effects such as secondary renal injury associated with hemolysis. Second, the authors’ technique for pulsatile perfusion delivery (using roller pumps) was not sufficient to improve renal hemodynamics and therefore did not improve outcomes.20 Interestingly, quantification of pulsatility is frequently not described in currently available studies. Third, there is a difference, but it is too small to measure in a study of this size, or this difference was outweighed

Strengths of this study include the large number of patients, pragmatic design, and the use of standardized endpoints.3 As a retrospective study, the authors cannot provide a clear cause and effect, and it is possible that unknown confounders contributed to the findings. (The authors were only able to adjust for known and measurable confounders.) However, the before and after design helped maintain similarities between the groups, including the same surgeons and a very similar group of patients (there were no changes in referral patterns during the study), and the authors were able to adjust for the major known confounders (there were some differences in diabetes rate, but these were accounted for in the multivariable analysis). A large number of patients were excluded from the intervention, and these results cannot be extrapolated to those patients. As discussed in the previous paragraph, the results may have been influenced by the inability to generate a sufficient pulse pressure using the authors’ methodology. The authors were unable to retrospectively quantify the pulse pressure generated in each patient. The pulse generated by the arterial roller pump may have been dampened by the oxygenator, nonetheless pulsatility was confirmed using invasive arterial pressure measurements. The authors were unable to compare other biomarkers of kidney injury, however, given that there was no difference in clinical outcome measures it is unlikely that a difference in biomarkers would prove to be important. It is possible that temporal trends in AKI rate affected the outcome. For example, an increasing overall rate in AKI during the study period could confound the ability to detect a reduced AKI rate as a result of the practice change. However, the analysis of the subset of surgeons who already were using pulsatile perfusion did not suggest this was the case. It should be noted that the surgeons already using pulsatile CPB had a higher rate of AKI than the other surgeons, raising the possibility of differences in the surgical populations. Conclusions In this before-and-after study of pulsatile perfusion in cardiac surgery there was no association between the universal introduction of pulsatile perfusion and AKI rate. Given the limitations of the study design the authors cannot determine cause and effect. However, it appears likely that pulsatile flow generated in this manner has no effect on renal function or has such a small effect that it was not detectable in a study of this size. Conflict of interest The authors have no conflicts of interest to disclose.

ARTICLE IN PRESS 6

T.G. Coulson et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2019) 16

References 1 Wang Y, Bellomo R. Cardiac surgery-associated acute kidney injury: Risk factors, pathophysiology and treatment. Nature Reviews Nephrology 2017; 13:697–711. 2 Vives M, Wijeysundera D, Marczin N, et al. Cardiac surgery-associated acute kidney injury. Interact Cardiovasc Thorac Surg 2014;18:637–45. 3 McIlroy DR, Bellomo R, Billings FT IV, et al. Systematic review and consensus definitions for the Standardised Endpoints in Perioperative Medicine (StEP) initiative: Renal endpoints. Br J Anaesth 2018;121: 1013–24. 4 Hobson CE, Yavas S, Segal MS, et al. Acute kidney injury is associated with increased long-term mortality after cardiothoracic surgery. Circulation 2009;119:2444–53. 5 Loef BG, Epema AH, Smilde TD, et al. Immediate postoperative renal function deterioration in cardiac surgical patients predicts in-hospital mortality and long-term survival. J Am Soc Nephrol 2005;16: 195–200. 6 Parikh CR, Puthumana J, Shlipak MG, et al. Relationship of kidney injury biomarkers with long-term cardiovascular outcomes after cardiac surgery. J Am Soc Nephrol 2017;28:3699–707. 7 Chawla LS, Eggers PW, Star RA, Kimmel PL. Acute kidney injury and chronic kidney disease as interconnected syndromes. N Engl J Med 2014;371: 58–66. 8 Serraino GF, Marsico R, Musolino G, et al. Pulsatile cardiopulmonary bypass with intra-aortic balloon pump improves organ function and reduces endothelial activation. Circ J 2012;76:1121–9. 9 O’Neil MP, Fleming JC, Badhwar A, et al. Pulsatile versus nonpulsatile flow during cardiopulmonary bypass: Microcirculatory and systemic effects. Ann Thorac Surg 2012;94:2046–53.

10 Siepe M, Goebel U, Mecklenburg A, et al. Pulsatile pulmonary perfusion during cardiopulmonary bypass reduces the pulmonary inflammatory response. Ann Thorac Surg 2008;86:115–22. 11 Onorati F, Santarpino G, Tangredi G, et al. Intra-aortic balloon pump induced pulsatile perfusion reduces endothelial activation and inflammatory response following cardiopulmonary bypass. Eur J Cardiothorac Surg 2009;35:1012-9. 12 Farid S, Povey H, Anderson S, et al. The effect of pulsatile cardiopulmonary bypass on the need for haemofiltration in patients with renal dysfunction undergoing cardiac surgery. Perfusion 2016;31:477–81. 13 Onorati F, Santarpino G, Rubino AS, et al. Body perfusion during adult cardiopulmonary bypass is improved by pulsatile flow with intra-aortic balloon pump. Int J Artif Organs 2009;32:50–61. 14 Thakar CV. A clinical score to predict acute renal failure after cardiac surgery. J Am Soc Nephrol 2004;16:162–8. 15 Karkouti K, Wijeysundera DN, Yau TM, et al. Acute kidney injury after cardiac surgery. Circulation 2009;119:495–502. 16 StataCorp. Stata Statistical Software: Release 12. College Station, TX: Stata Corp LP; 2011. 17 Nam MJ, Lim CH, Kim H-J, et al. A meta-analysis of renal function after adult cardiac surgery with pulsatile perfusion. Artif Organs 2015;39:788–94. 18 Seivert A, Sistino J. A meta-analysis of renal benefits to pulsatile perfusion in cardiac surgery. J Extra Corpor Technol 2012;44:10–4. 19 Milano AD, Dodonov M, Van Oeveren W, et al. Pulsatile cardiopulmonary bypass and renal function in elderly patients undergoing aortic valve surgeryy. Eur J Cardiothorac Surg 2014;47:291–8. 20 Evans RG, Lankadeva YR, Cochrane AD, et al. Renal haemodynamics and oxygenation during and after cardiac surgery and cardiopulmonary bypass. Acta Physiol 2017;222;e12995-15. 21 Kellum JA, et al. KDIGO clinical practice guideline for acute kidney injury. Kidney Int Suppl 2012;2:1–138.