Fluid Balance and Acute Kidney Injury

MISSING IMAGE

Material Information

Title:
Fluid Balance and Acute Kidney Injury a Prospective Observational Study
Physical Description:
1 online resource (32 p.)
Language:
english
Creator:
Kambhampati,Ganesh
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Medical Sciences, Clinical Investigation (IDP)
Committee Chair:
Brantly, Mark L
Committee Members:
Nelson, David R

Subjects

Subjects / Keywords:
Clinical Investigation (IDP) -- Dissertations, Academic -- UF
Genre:
Medical Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Positive fluid balance (FB) has been linked to adverse clinical outcomes in patients with acute kidney injury (AKI). However, it is unclear whether FB is the cause or result of AKI. We therefore performed a prospective, observational study to investigate the relationship of FB and AKI. Adult patients undergoing non-transplant cardiovascular surgery with estimated glomerular filtration rate>30mL/min/1.73m2 were included. Patients were divided into quartiles based on FB in the first 24 hours from initiation of surgery. Incidences of AKI, urine Neutrophil Gelatinase-Associated Lipocalin (NGAL) and IL-18 concentrations, serum cytokine concentrations and serum creatinine (SCr) levels were compared between the lowest (Low-FB) and highest (High-FB) quartile groups. AKI was defined per Acute Kidney Injury Network criteria. One hundred patients were included for analyses. Patients in the High-FB were significantly older, had higher baseline SCr, lower mean arterial pressures (MAP) and >92% of patients in each group were on angiotensin converting enzyme inhibitors or angiotensin receptor blockers preoperatively. The major finding was that the High-FB group had a significantly (six-fold) increased risk for AKI (adjusted OR 6.48, CI95%1.37-30.51, p=0.018) and was associated with significantly higher urine NGAL (p=0.038), IL-8 (p=0.01), MCP-1 (p=0.03) and TNF-alpha (p=0.003) concentrations. An important observation was that positive FB occurred early in the intraoperative period and continued into the initial ICU period. Positive FB preceded the development of AKI. The High-FB group received significantly more blood product transfusions (p=<0.001), medications (p<0.001) and fluids (p<0.001) and had significantly longer duration of surgery (p=0.003) and time on cardiopulmonary bypass machine (p=0.018). Urine output was not significantly different (p=0.483) between groups in the intraoperative period. Both groups demonstrated a 35-40% reduction in intraoperative mean arterial pressure from baseline values, but significant differences between them were not observed (p=0.954). High-FB group had longer ICU (p=0.001) and hospital stay (p=0.048). Positive FB in the first 24 hours from initiation of surgery may be an independent risk factor for postoperative AKI. Positive FB was associated with increased biomarkers of AKI and mediators of inflammation. Positive FB occurred early in the intraoperative period and preceded the development of AKI. The data is suggestive of the causal role of FB in AKI; however, more studies will be required to confirm this relationship.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Ganesh Kambhampati.
Thesis:
Thesis (M.S.)--University of Florida, 2011.
Local:
Adviser: Brantly, Mark L.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0043422:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 FLUID BALANCE AND ACUTE KIDNEY INJURY: A PROSPECTIVE OBSERVATIONAL STUDY By GANESH KAMBHAMPATI A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Ganesh Kambhampati

PAGE 3

3 To my wife, mentors, colleagues, coinvestigators and friends who guided, supported and participated in this clinical trial and helped me successfully accomplish my goal on time

PAGE 4

4 ACKNOWLEDGMENTS I thank the mentor of my research project Dr. A. Ahsan Ejaz for his mentoring, the cofellows in the Division of Nephrology for their research assistance, the participants in my clinical trial for their honest and open participation, and the Division of Nephrology for its generous support. I thank my d ivision chief and T32 m entor Dr. Mark Segal for guiding me in finding a research project and assist me in building my career in early days of my fellowship training. I thank my wife for her loving encouragement, which motivated me to complete my study.

PAGE 5

5 TABLE OF CON TENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 6 LIST OF FIGURES .......................................................................................................... 7 ABSTRACT ..................................................................................................................... 8 CHAPTER 1 INTRODUCTION .................................................................................................... 10 Study Aim ............................................................................................................... 10 Background ............................................................................................................. 10 Study Design and Inclusion Criteria ........................................................................ 11 Study Protocol ........................................................................................................ 12 Statistical Methods .................................................................................................. 13 2 RESULTS ............................................................................................................... 15 Baseline Patient Characteristics ............................................................................. 15 Intraoperative Parameters ...................................................................................... 16 3 OUTCOMES ........................................................................................................... 20 Fluid Balance and Biom arkers of Acute Kidney Injury ............................................ 20 Fluid Balance and the Incidence of Acute Kidney Injury by Acute Kidney Injury Network Criteria ................................................................................................... 21 Fluid Balance and Cytokine Profiles ....................................................................... 22 Fluid Balance, Peak Serum Creatinine and Glomerular Filtration Rate .................. 22 Fluid Balance, Hospital Complications, Ventilation Support and Length of Hospital Stay ....................................................................................................... 23 Fluid Ba lance, Renal Replacement Therapy and Thirty day Mortality .................... 23 4 DISCUSSION ......................................................................................................... 27 REFERENCES .............................................................................................................. 30 BIOGRAPHICAL SKETCH ............................................................................................ 32

PAGE 6

6 LIST OF TABLES Table page 2 1 Patient characteristics. ....................................................................................... 17 3 1 Predictors of acute kidney injury by univariate analysis. ..................................... 24

PAGE 7

7 LIST OF FIGURES Figure page 2 1 Intake volumes in lowest vs. highest quartiles. ................................................... 18 2 2 Urine output in lowest vs. highest quartiles. ....................................................... 18 2 3 Fluid balance in lowest vs. highest quartiles. ...................................................... 19 2 4 Perioperative mean arterial pressures. ............................................................... 19 3 1 Receiver operating characteristics curves. ......................................................... 25 3 2 Serum creatinine levels in lowest vs highest quartiles. ...................................... 25 3 3 Fluid balance and resource utilization. ............................................................... 26

PAGE 8

8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science FLUID BALANCE AND ACUTE KI DNEY INJURY: A PROSP ECTIVE OBSERVATIONAL STUDY By Ganesh K ambhampati August 2011 Chair: Mark Brantly Major: Medical Sciences Clinical and Translational Science Positive fluid balance (FB) has been linked to adverse clinical outcom es in patients with acute kidney injury (AKI). However it is unclear whether FB is the cause or result of AKI. We therefore performed a prospective, observational study to investigate the relationship of FB and AKI. Adult patients undergoing nontranspl ant cardiovascular surgery with estimated glomerular filtration rate>30mL/min/1.73m2 were included. Patients were divided into quartiles based on FB in the first 24 hours from initiation of surgery. Incidences of AKI, u rine Neutrophil Gelatinase Associated Lipocalin (NGAL) and IL 18 concentrations, serum cytokine concentrations and serum creatinine (SCr) levels were compared between the lowest (Low FB) and highest (HighFB) quartile groups. AKI was defined per Acute Kidney Injury Network criteria. One hundr ed patients were included for analyses. Patients in the HighFB were significantly older, had higher baseline SCr, lower mean arterial pressures (MAP) and >92% of patients in each group were on angiotensin converting enzyme inhibitors or angiotensin receptor blockers preoperatively. The major finding was that the HighFB group had a significantly (six fold) increased risk for AKI (adjusted OR 6.48, CI95%1.37 30.51, p=0.018) and was

PAGE 9

9 associated with significantly higher urine NGAL (p=0.038), IL8 (p=0. 01), MCP 1 (p=0.03) and TNF alpha (p=0.003) concentrations. An important observation was that positive FB occurred early in the intraoperative period and continued into the initial ICU period. Positive FB preceded the development of AKI. The HighFB group received significantly more blood product transfusions (p=<0.001), medications (p<0.001) and fluids (p<0.001) and had significantly longer duration of surgery (p=0.003) and time on cardiopulmonary bypass machine (p=0.018). Urine output was not significantl y different (p=0.483) between groups in the intraoperative period. Both groups demonstrated a 3540% reduction in intraoperative mean arterial pressure from baseline values, but significant differences between them were not observed (p=0.954). HighFB grou p had longer ICU (p=0.001) and hospital stay (p=0.048). Positive FB in the first 24 hours from initiation of surgery may be an independent risk factor for postoperative AKI. Positive FB was associated with increased biomarkers of AKI and mediators of inflammation. Positive FB occurred early in the intraoperative period and preceded the development of AKI. The data is suggestive of the causal role of FB in AKI ; however more studies will be required to confirm this relationship.

PAGE 10

10 CHAPTER 1 INTRODUCTION Study Aim The primary aim of this study was to investigate the relationship between fluid balance and acute kidney injury using clinical parameters, biochemical (serum creatinine) and biomarkers of acute kidney injury (urine NGAL and IL18). Other endpoints for the study included 2day and hospital peak serum creatinine, discharge serum creatinine, postoperative complication rates, length of intensive care unit and hospital stay, incidence of 30day dialysis and/or all cause mortality, and inflammatory response profile. A cute kidney injury was defined as an absolute increase in serum crea tinine > 0.3mg/dL from baseline within 48 hours after surgery in accordance with the Acute Kidney Injury Networks criteria.9 Twenty four hour fluid balance is defined as the difference between intake and output within 24 hours from the start of surgery (t ime 0hr 24hr). Background Positive fluid balance in critical care patients has been proposed to be a major determinant of adverse outcomes in patients with acute kidney injury (AKI) In post hoc analysis, a 10% increase in body weight from baseline1 or pos itive fluid balance during the patient's intensive care stay2 was associated with increased 60day mortality in patients with acute kidney injury. The median cumulative fluid balance after the onset of acute kidney injury has been shown to increase from 2. 7 liters on day 2 to 6.5 liters on day 7 in critically ill patients, suggesting severity progression of AKI .3 Positive fluid balance has also been shown to be an independent risk factor for adverse outcomes in patients with acute lung injury4,5 where the proposed mechanisms involve tissue edema,

PAGE 11

11 impaired oxygenation and metabolite diffusion, impaired capillary blood flow and lymphatic drainage, and disturbed cell cell integrity that contribute to organ dysfunction. However, it is still unclear from reported studies whether positive fluid balance was the result of acute kidney injury or contributed to acute kidney injury that culminated in adverse outcomes. We therefore investigated the relationship of fluid balance and acute kidney injury in a prospective, observational study in patients undergoing cardiovascular surgery where rapid changes in fluid volume during surgery offer unique opportunities to study this relationship. Study Design and Inclusion Criteria This is a prospective, observational study desig ned to investigate the relationship between fluid balance and acute kidney injury. Acute kidney injury was ascertained by traditional (serum creatinine [ SCr ] ) and nontraditional biomarkers urine neutrophil gelatinase associated lipocalin, ( urine NGAL ) and IL 18 of kidney injury. Other outcomes studied were peak serum creatinine, postoperative complications, length of hospital stay and inflammatory response profiles in patients undergoing cardiovascular surgery. We set the criteria for inclusion into the st udy based on previous reports6 to maximize the likelihood of seeing an effect. Specifically, we limited recruitment of subjects to older than 18 years of age undergoing cardiovascular surgery (thoracic aortic aneurysm, cardiac valve or coronary artery bypass graft surgery) and with estimated glomerular filtration rate > 30ml/min/1.73m2. Patients with a prior history of organ transplant, on preoperative intraaortic balloon pump or on natriuretic peptides were excluded from study participation. The study was performed by the Division of Nephrology, Transplantation and Hypertension and the Division of Thoracic and

PAGE 12

12 Cardiovascular Surgery at the University of Florida (UF) in Gainesville, FL, was approved by the UF Institutional Review Board and conforms to regulations of the UF ethics committee on h uman r esearch Patients were recruited from the cardiovascular surgery clinics. The study was registered at www.ClinicalTrials.gov unique identifier NCT01168583. Written informed consents were obtained from all patients. Study P rotocol Baseline clinical data on demographics, comorbid conditions, medications, vital signs, and pertinent laboratory tests were obtained by patient interviews and review of electronic medical records. All patients received routine postoperative supportive care for their medical and surgical problems, including care for acute kidney injury, optimization of fluid and nutritional status and inotropic support as required. Cardiopulmonary bypass was utilized in all patients using standard cardiopulmonary bypass protocol. The standard cardiopulmonary bypass prime consisted of 1400mL of Plasmalyte, 5000units of heparin, 12.5gm of mannitol, 50meq of sodium bicarbonate, 25gm of albumin in 100mL of normal saline, and 250mg of methylprednisolone. The postoperative fluid protocol included the administration of up to 1 liter of normal saline for cardiac index <2, unstable systolic blood pressure <80mmHg with left atrial pressure or pulmonary artery diastolic pressure < 15mmHg. Diuretics could be administered for volume management per established protocol of the cardiovascular surgery intensive care unit that included central venous pressure parameters, physical, radiological or physiological index evidence for pulmonary congestion and inadequate oxygenation. The need for renal replacement therapy was determined independently by the patients treating nephrologists per current standard of care criterion, which included blood urea nitrogen > 80mg/dL, electrolyte or acidbase disorders not responding to medical

PAGE 13

13 management, and diuretic unresponsiveness with urine output < 0.5m l /kg/hour and refractory volume overload as defined by central venous pressures > 15 mmHg. Estimated glomerular filtration rate (GFR) was cal culated using the National Kidney Foundations MDRD Calculator (2000 2001). Intraoperative data on fluid volume and surgical procedure were obtained from the standardized anesthesia operative room records. Postoperative data were collected daily at patient bedside from standardized intensive care unit flow charts and medical records. Serum and urine samples were collected at 24 hours from start of surgery and stored at 80OC. Urine NGAL measurements were performed in the laboratory of Dr. Prasad Devarajan at the Cincinnati Children's Hospital Medical Center Biomarker Laboratory, Cincinnati, Ohio using a previously published ELISA method.7 Urine IL 18 measurements were performed in the laboratory of Dr. Charles Edelstein at the University of Colorado Heal th Sciences Center, Denver using a previously published ELISA method.8 Inflammatory cytokines were quantified by enzymelinked immunosorbent assay using multiplex human cytokine/chemokine kit. The following thirty nine serum and urine biomarkers were measured: EGF, Eotaxin, FGF 2, Flt e ligand, Fractalkine, G CSF, GM CSF, GRO, IL 10, IL12(p40), IL12(p70), IL13, IL 15, IL 17, Il 1a, IL 1b, Il 1ra, IL 2, Il 3, IL 4, IL 5, IL 6, IL 7, IL 8, IL 9, INF alpha2, INF gamma, IP 10, MCP 1, MCP 3, MDC, MIP 1a, MIP 1b, sCD40L, sIL2Ra, TGF alpha, TNF alpha, TNF beta, VEGF. Statistical M ethods Patients were divided into quartiles based on 24hour fluid balance status. Outcome variables were calculated for each quartile and compared between quartiles. Comparisons bet ween groups were assessed using parametric tests for normally

PAGE 14

14 distributed variables (chi squared t test) and nonparametric tests (MannWhitney U test) for variables which are not normally distributed as appropriate. Results are presented as mean + standard error of mean with P value. Median values are provided for several variables. P value < 0.05 was considered statistically significant. All analyses were conducted using SPSS version 18, Chicago Ill

PAGE 15

15 CHAPTER 2 R ESULTS Baseline P atient C haracteristics One hundred patients were enrolled and all of them completed the study. Most of the patients were Caucasian males and most of them were chronically taking angiotensionconverting enzyme inhibitor or angiotensin receptor blocker medications (Table 2 1 ). Renal and cardiac functions appeared to be well preserved. Preoperative cardiac catheterizations were performed in 39% of patients within 72 hours of planned surgery. The three types of surgical procedures ( thoracic aortic aneurysm, cardiac valve or coronary artery bypass graft surgery ) were performed in relatively equal proportions. The mean intake, output and fluid balance for the first 24 hours from the start of surgery (time 0hr 24hr) for the full cohort are as follows: intraoperative (O R) intake, blood products 1169.3 + 102.5mL, medications/fluids 1580.5+ 105.7mL, total OR intake 2749.8+ 159.3mL; OR output, blood loss7.50 + 6.6mL, urine output1079.8+ 62.9mL, total OR output 1087.3+ 63.2mL, OR fluid balance 1662.5+ 145.9mL; total postoperative (end of surgery until 24hour) intake 2629.4+ 124.1mL, urine output 1101.2+ 54.5mL, drainage and other outputs 640.8+ 51.4mL, total postoperative output 1742.1 + 68.7mL, postoperative fluid balance 887.3+ 112.8mL; 24 hr fluid balance mean 2567.2+ 202.2mL, median 2268.5mL, mode 2706.0mL, 25th percentile 936.2mL, 50th percentile 2268.5mL and 75th percentile 3898.2mL. To investigate the relationship between fluid balance and study outcomes, patients were divided into quartiles based on 24hr fluid balance status. Outcomes in the highest quartile group were then compared with the lowest quartile group. Baseline characteristics of the groups are exhibited in Table 2 1 The highest quartile group was older, had lower mean arterial pressures and higher serum

PAGE 16

16 creatinine levels, although estimated glomerular filtration rate was not significantly different compared to the lowest quartile group. There were no differences in the types of surgery performed or in preoperative exposure to contrast agents. Patients in the highest quartile received significantly more blood products and fluids in the intraoperative, postoperative and intensive care unit stay period (Figure 2 1 ) Urine output s were not different between the groups in the intraoperative period, but thereafter was significantly reduced in the highest quartile group in the postoperative and intensive care unit period (Figure 2 2 ). Differences in fluid balance emerged early in the intraoperative period between the groups and continued to increase i n the intensive care unit period (Figure 2 3 ). More patients in this group required diuretic therapy in the immediate postoperative period than in the lowest quartile group (lowest quartile 32% vs. highest quartile 100%, p<0.001). Intraoperative P arameters The mean duration of surgery for the full cohort was 403.5+ 12.5min, mean cardiopulmonary bypass time was 167.0+ 7.3min and mean aortic crossclamp time (N=91) was 105.2+ 5.5min. Investigation of events in the intraoperative period demonstrated that the highest quartile group had a longer duration of surgery (lowest quartile 363.0+ 17.2min vs. highest quartile 467.6+ 27.9, p=0.003) and prolonged time on heart lung perfusion machine (cardiopulmonary bypass time, lowest quartile143.4+ 44.7min vs. hig hest quartile192.6+ 17.7min, p=0.018). Aortic cross clamp time was not significantly different (lowest quartile 95.5+ 8.9min vs. highest quartile 115.8+ 13.0, p=0.208) between the groups. A large reduction in mean arterial pressure from baseline values was no ted in the intraoperative period; however significant differences between the groups were not evident ( Figure 24 ).

PAGE 17

17 Table 21. Patient characteristics. Variables Full cohort N=100 24 hour fluid balance Lowest quartile Highest quartile p value* Demographics Age (years) Male gender (%) Race (%): Caucasian African American Others 61.4+ 1.4 60 85 8 7 56.8+ 2.4 56 76 16 8 66.4+ 1.9 64 76 12 12 0.003 0.773 0.842 Co morbid conditions (%) Hypertension Diabetes mellitus Coronary artery disease Congestive heart failure Previous cardiac surgery Peripheral vascular disease COPD 66 26 37 10 13 9 12 56 16 28 4 4 4 12 76 28 44 8 16 12 20 0.232 0.496 0.377 1.000 0.349 0.609 0.702 Medications (%) ACE inhibitors/ARBs Diuretics Beta blockers Calcium channel blockers 93 35 47 19 92 40 44 20 96 36 60 28 1.000 1.000 0.259 0.742 Cardiac function NYHA class M ean arterial pressure (mmHg) LVEF (%) 1 79.6+ 1.0 52.7 + 1.6 1 84.3+ 2.1 50.9 + 4.1 1 76.1+ 2.1 52.7 + 2.7 0.009 0.709 Renal function BUN (mg/dL) S erum creatinine (mg/dL) E stimated glomerular filtration rate (mL/min/1.73m 2 ) 18.4+ 0.7 0.9 + 0.0 85.9+ 3.1 18.1+ 1.6 0.8 + 0.0 84.4+ 4.0 19.4+ 1.2 1.0 + 0.1 93.2+ 6.1 0.495 0.046 0.239 Surgical procedures (%) TAA Valves CABG 28 40 32 32 32 36 32 36 32 0.943 Preoperative cardiac catheterization ( %) 39 40 52 0.571 *p value calculated for lowest vs. highest quartiles; LVEF: Left ventricular ejection fraction; mean arterial pressure: mean arterial pressure; TAA: thoracic aortic aneurysm; valves: cardiac valves; CABG: coronary artery bypass graft; BUN: blood urea nitrogen; COPD: chronic obstructive pulmonary disease; ACE inhibitors: angiotensin converting enzyme inhibitors, ARB: angiotensin receptor blocker; NYHA: New York Heart Association.

PAGE 18

18 Figure 21. Intake volumes in lowest vs. highest quartiles. Figure 22. U rine output in lowest vs. highest quartiles.

PAGE 19

19 Figure 23. Fluid balance in lowest vs. highest quartiles Figure 24. Perioperative mean arterial pressures

PAGE 20

20 CHAPTER 3 OUTCOMES Fluid B alance and B iomarkers of A cute K idney I njury Early acute kidney injury was investigated by two different urinary biomarkers of kidney injury IL 8 an d NGAL The mean and median values of urine IL18 concentrations respectively were as follows: full study cohort, 249.6 + 50.3pg/mL and 90.7pg/mL; lowest quartile, 244.2+ 120.3pg/mL and 89.9pg/mL and highest quartile, 380.4+ 144.8pg/mL and 95.2pg/mL. The re was no significant differen ce in IL 18 between the two groups (p=0.741). As expected, p atients with acute kidney injury had higher urine IL18 levels compared to those wit hout acute kidney injury (546.5+ 163.7pg/mL vs. 139.8+ 23.9pg/mL, p=0.021). There were no significant differences in urine IL18 levels between patients with acute kidney injury (n=27) and patients in the highest quartile group (n=25): acute kidney injury 546.5+ 163.7pg/mL vs. highest fluid balance 380.4+ 144.8pg/mL, p=0.454). However, the highest quartile group had higher urine IL18 levels than the no acute kidney injury group (highest quartile380.4+ 144.8pg/mL vs. no acute kidney injury 139.8+ 23.9pg/mL, p=0 .011). Significantly higher urine NGAL concentrations were noted in the highest quartile group (lowest quartile mean 42.8+ 19.7ng/mL, median 9.5ng/mL vs. highest quartile mean 107.0+ 66.4ng/mL, median 22.1ng/mL; p=0.038) in comparison to the lowest quartil e group. Patients with acute kidney injury had higher urine NGAL concentration than those with no acute kidney injury (107.8 + 61.4ng/mL vs. 34.0+ 7.5ng/mL, p=0.018). Since the earliest evidence of positive fluid balance was observed in the intraoperative period, we investigated the correlation of blood product transfusion, medication/fluids, total intake and fluid balance in this period with urine NGAL as the

PAGE 21

21 biomarker for kidney injury. In the full cohort, significant correlation was demonstrated between transfusion volume (r=0.281, p=0.005), total intake volume (r=0.274, p=0.006), and fluid balance (r=0.254, p=0.011) with urine NGAL concentrations. When the analysis was restricted to only the highest quartile group, transfusion volume was the only variabl e that significantly correlated with urine NGAL (r=0.397, p=0.050), however transfusion volume did not show any correlation with serum creatinine, the traditional marker of acute kidney injury. The r eceiver o perating characteristics (ROC) curves for f luid balance, urine NGAL and IL18 to predict acute kidney injury vs. no acute kidney injury are shown in Figure 3 1 The overall ability to predict acute kidney injury vs. no acute kidney injury was not significantly different between fluid balance, urine NGAL and IL 18 (AUC for fluid balance: 0.669+ 0.064, CI95% 0.5450.794, p=0.010; IL18: 0.654+ 0.065, CI95% 0.5260.782, p=0.018; NGAL: 0.624+ 0.065, CI95% 0..497 0.751, p=0.058). Fluid Balance a nd t he Incidence o f Acute Kidney Injury b y Acute Kidney Injury Network Criteria The overall incidence of acute kidney injury in the full cohort was 27%. Most (97%) of the patients had S tage 1 acute kidney injury defined as s erum creatinine increase 2.0 fold from baseline.9 The highes t quartile group was associated with a significantly higher incidence of acute kidney injury compared to the lowest quartile group (lowest quartile16% vs. highest quartile52%, p=0.016). In the univariate analysis performed to determine the odds ratio (OR) for acute kidney injury, age, diabetes status, cardiopulmonary bypass time >200min and 24hour fluid balance were significant predictors of acute kidney injury (Table 3 1 ). The highest quartile group had a 6fold increased risk for acute kidney injury (OR 6.0, CI95%

PAGE 22

22 1.7 21.3, p=0.006). When all risk factors in Table 3 2 were included, without selection, in the multivariate model, the highest quartile group was again associated with a 6fold increased risk for acute kidney injury (OR 6.48, CI95%1.37 30.51, p=0.018) compared to the lowest quartile group. The incidence of acute kidney injury during the entire hospital stay was also higher in the highest quartile group compared to the lowest quartile group (lowest quartile 20% vs. highest quartile 60%, p=0.009) associated with a six fold increased risk for in hospital acute kidney injury (OR 6.0, CI95% 1.7 21.3). Intraoperative urine output or fluid balance per se was not a significant predictor of acute kidney injury. Fluid B alance and C ytokine P rofiles Only f our of the 39 serum and urine biomarkers that were measured were significant ly different between the groups : serum IL8 (lowest quartile 27.1pg/mL vs. highest quartile 41.3pg/mL, p=0.010), IL10 (lowest 21.7pg/mL vs. highest quartile 47.5pg/mL, p=0.009), M CP 1 (lowest quartile 333.0pg/mL vs. highest quartile 527.0pg/mL, p=0.030) and TNF alpha (lowest quartile 6.9pg/mL vs. highest quartile 11.2pg/mL, p=0.003). IL1a/IL1ra ratios were not significantly different between the groups (p=0.797). In the urine, si gnificant differences were observed in IL1ra (lowest quartile 52.9pg/mL vs. highest quartile 143.0pg/mL, p=0.031), and IL 10 (lowest quartile 9.9pg/mL vs. highest quartile 13.7pg/mL, p=0.048). Fluid B alance, P eak S erum C reatinine and G lomerular F iltratio n R ate Significant differences in 48hour peak, hospital peak and discharge serum creatinine were noted between the groups (Figure 3 2 ). However, the discharge estimated glomerular filtration rates were not significantly different between the groups (lowest quartile78.6+ 7.2mL/min/1.73m2 vs. highest quartile87.4+ 8.4mL/min/1.73m2

PAGE 23

23 (p=0.428). S erum creatinine values were noted to return to baseline values at discharge (p=0.434). Fluid B alance, H ospital C omplications V entilation S upport and L ength of H ospital S tay The incidence of postoperative complications such as cardiac arrhythmias, pacemaker insertions, pleural effusions, pneumonias, pneumothorax, tracheostomy, intra aortic balloon pump placement, poor wound healing, gastrointes tinal bleeding, stokes and infections were significantly higher in the highest quartile group (lowest quartile 8% vs. highest quartile 56%, p<0.001). These patients also spent significantly more time in the intensive care unit and had a significantly longer duration of hospital stay (Figure 3 3). Time on mechanical ventilation support was not different between the groups (p=0.65). Fluid B alance, R enal R eplacement T herapy and Thirty day M ortality Two patients in the highest quartile group and none in the lowest quartile group required renal replacement therapy (p=0.490). Thirty day mortality rates were not significantly different between the groups (lowest quartile 0% vs. highest quartile 4%, p=1.000).

PAGE 24

24 Table 31. Predictors of acute kidney injury by univariate analysis. Variables Odds r atio CI 95% p value Demographics Age (years) Race Gender 0.97 0.64 1.24 0.940.99 0.311.30 0.54 2.84 0.038 0.218 0.611 Type of surgery Thoracic aortic aneurysms Cardiac valves Coronary artery bypass grafts 0.82 0.90 1.34 0.332.03 0.392.05 0.57 3.14 0.675 0.802 0.505 Co morbidities Diabetes mellitus Hypertension Chronic obstructive pulmonary disease Peripheral vascular disease Coronary artery disease Previous cardiac surgery 3.55 2.31 1.64 2.10 1.89 0.66 1.4 9.0 0.945.7 0.495.5 0.538.3 0.834.35 0.19 2.31 0.008 0.068 0.424 0.294 0.132 0.516 Cardiac function Ejection fraction <45% 1.26 0.44 3.64 0.670 Renal function Estimated glomerular filtration rate (ml/min/1.73m 2 ) 0.99 0.98 1.00 0.350 Intraoperative Duration of surgery cardiopulmonary bypass time>200 minutes 0.99 2.85 0.991.00 1.13 7.16 0.280 0.026 Fluid balance 24hour fluid balance Lowest vs. Highest quartile 1.00 5.69 0.991.00 1.51 21.41 0.004 0.010

PAGE 25

25 Figure 31. Receiver operating characteristics curves Figure 32. Serum creatinine levels in lowest vs. highest quartiles.

PAGE 26

26 Figure 33. Fluid balance and resource utilization.

PAGE 27

27 CHAPTER 4 DISCUSSION Several retrospective studies have linked positive fluid balance with adverse clinical outcomes in patients who had developed acute kidney injury,1 3 however the temporal relationship of fluid balance and acute kidney injury remains unclear. We investigated the relationship between fluid balance and acute kidney injury in a prospective manner in patients undergoing cardiac surgery using standard surgical protocols. This setting allowed for relative timelines to be established between events and outcomes. The major finding of the study was that positiv e fluid balance in the first 24 hours from initiation of surgery was an independent risk factor for postoperative acute kidney injury. The risk was six fold hi gher in the highest quartile vs. lowest quartile group. Positive fluid balance was also associated with increased urine biomarkers of acute kidney injury and high serum creatinine levels during hospital stay. An important observation was that positive fluid balance occurred early in the intraoperative period and preceded the rise in serum creatinine. We did not evaluate fluid balance status beyond postoperative day 2 due to reported inaccuracies in multi day intakes and outputs tracking in the intensive care unit setting.10 However others have reported increasing cumulative positive fluid balance with progression severity of acute kidney injury.3 Patients in the highest quartile group received more blood product transfusions, medications and fluids in the intraoperative period than the lowest quartile group. Perioperative red blood cell transfusions have been independently linked to acute kidney injury.11 Urine output was not different between the groups. We inve stigated other preand intraoperative events that could possibly have precipitated the increased intake and found that the highest quartile group included older patients

PAGE 28

28 with higher baseline serum creatinine, lower baseline mean arterial pressure, longer duration of surgery and prolonged cardiopulmonary bypass time than the lowest quartile group. Both groups demonstrated a 3540% reduction in intraoperative mean arterial pressure from baseline values, but significant differences between them were not obser ved. Blood loss was not different between the groups. It is noteworthy that volume and method of intraoperative fluid administration, whether due to preferences of the surgical team, complexity of surgery or other issues, may also be an important determina nt of outcomes. Indeed, a recent study in children has reported an increase in 48hour mortality in critically ill children treated with fluid boluses,12 and also conservative strategy of fluid management in patients with acute lung injury has been shown t o improve lung function and shorten the duration of mechanical ventilation and intensive care stay without increasing nonpulmonary organ failures.5 Urine NGAL and mediators of inflammation were elevated in the highest quartile group, suggesting an association between fluid balance, inflammation and acute kidney injury. The possibility that intraoperative acute kidney injury due to longer cardiopulmonary bypass time and complexity of surgery was the initiating event in this study cannot be entirely excluded. However, intraoperative biomarkers of acute kidney injury and inflammatory mediators were not available to make that determination. Serum and uri ne samples were collected at 24 hours following initiation of surgery to separate the effects of positive fluid balance on inflammation and kidney injury from that of surgery related injury. In light of the findings of the study that positive fluid balance occurs as early as in the intraoperative period, the focus of the time period of investigation may need to b e modified in future studies.

PAGE 29

29 Another significant observation was that at 24 hours from start of surgery, there were no differences between the overall ability of fluid balance, urine NGAL and IL18 to predict acute kidney injury vs. no acute kidney inj ury. That is, simple clinical observation may give similar information as urine NGAL and IL18 in this setting and time period. However, it needs to be emphasized that urine NGAL and IL18 are reliable biomarkers that can detect acute kidney injury within 2 6 hours after renal injury.7,13 Postoperative complication rates were higher in the highest quartile vs. lowest quartile group, and is in accordance with other published reports of the adverse effects of positive fluid balance in surgical patients.1416 Intensive care unit and hospital lengths of stay were longer but no differences were observed in the need for dialysis or duration of mechanical ventilation confirming previous findings. The strengths of the study was its setting, observations of events and outcomes during the course of routine surgical procedures, and association of operative and clinical variables with traditional and nontraditional biomarkers of acute kidney injury and inflammatory mediators. The limitations of the study were the lack of intraoperative biomarker and cytokine data and relatively small sample size. However, our data is hypothesis generating and highlights the need for investigations regarding fluid balance and acute kidney injury to be performed much earlier than after t he onset of acute kidney injury. We conclude that positiv e fluid balance in the first 24 hours from initiation of surgery may be an independent risk factor for postoperative acute kidney injury. Positive fluid balance is an excellent and simple predic tive marker that precedes the rise in serum creatinine.

PAGE 30

30 REFERENCES 1. Bouchard J, Soroko SB, Chertow GM, Himmelfarb J, Ikizler TA, Paganini EP, Mehta RL; Program to Improve Care in Acute Renal Disease (PICARD) Study Group. Fluid accumulation, survival and recovery of kidney function in critically ill patients with acute kidney injury. Kidney Int. 2009; 76:422427. 2. Payen D, de Pont AC, Sakr Y, Spies C, Reinhart K, Vincent JL; Sepsis Occurrence in Acutely Ill Patients (SOAP) Investigators. A positive fluid balance is associated with a worse outcome in patients with acute renal failure. Crit Care. 2008;12:R74. 3. Macedo E, Bouchard J, Soroko SH, Chertow GM, Himmelfarb J, Ikizler TA, Paganini EP, Mehta RL; Program to Improve Care in Acute Renal Disease Study Fluid accumulation, recognition and staging of acute kidney injury in critically ill patients. Crit Care. 2010;14:R82. 4. Alsous F Khamiees M DeGirolamo A Amoateng Adjepong Y Manthous CA .Negative Fluid Balance Predicts Survival in Patients With Sept ic Shock. Chest. 2000; 117:17491154. 5. National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network, Wiedemann HP, Wheeler AP, Bernard GR, Thompson BT, Hayden D, deBoisblanc B, Connors AF Jr, Hite RD, Harabin AL. Comparison of Two FluidManagement Strategies in Acute Lung Injury N Engl J Med. 2006; 354:25642575. 6. Ejaz AA, Martin TD, Johnson RJ, Winterstein AG, Klodell CT, Hess PJ Jr, Ali AK, Whidden EM, Staples NL, Alexander JA, House Fancher MA, Beaver TM. Prophylactic nesiritide does not prevent dialysis or all cause mortality in patients undergoing highrisk cardiac surgery. J Thorac Cardiovasc Surg. 2009;138:959964. 7. Bennett M, Dent CL, Ma Q, Dastrala S, Grenier F, Workman R, Syed H, Ali S, Bar asch J, Devarajan P. Urine NGAL predicts severity of acute kidney injury after cardiac surgery: a prospective study. Clin J Am Soc Nephrol. 2008; 3:665673. 8. Edelstein CL. Biomarkers of acute kidney injury. Adv Chronic Kidney Dis 2008; 15:222234. 9. Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, Levin A; Acute Kidney Injury Network. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care. 2007;11:R31. 10. Eastwood GM. Evaluating the rel iability of recorded fluid balance to approximate body weight change in patients undergoing cardiac surgery. Heart Lung. 2006; 35:2733.

PAGE 31

31 11. Karkouti K, Wijeysundera DN, Yau TM, Callum JL, Cheng DC, Crowther M, Dupuis JY, Fremes SE, Kent B, Laflamme C, Lam y A, Legare JF, Mazer CD, McCluskey SA, Rubens FD, Sawchuk C, Beattie WS. Acute kidney injury after cardiac surgery: focus on modifiable risk factors. Circulation. 2009; 119:495502. 12. Maitland K, Kiguli S Opoka RO Engoru C Olupot Olupot P Akech SO Nyeko R Mtove G Reyburn H Lang T Brent B Evans JA Tibenderana JK Crawley J Russell EC Levin M Bab iker AG Gibb DM ; the FEAST Trial Group Mortality after Fluid Bolus in African Children with Severe Infection. N Engl J Med. 2011 May 26. [Epub ahead of print] 13. Parikh CR, Abraham E, Ancukiewicz M, Edelstein CL. Urine IL 18 is an early diagnostic marker for acute kidney injury and predicts mortality in the intensive care unit. J Am Soc Nephrol. 2005; 16:30463052. 14. Benes J, Chytra I, Altmann P, Hluchy M, Kasal E, Svitak R, Pradl R, Stepan M. Intraoperative fluid optimization using stroke volume variation in high risk surgical patients: results of prospect ive randomized study. Crit Care. 2010; 14:R118. 15. McArdle GT, McAuley DF, McKinley A, Blair P, Hoper M, Harkin DW. Preliminary results of a prospective randomized trial of restrictive versus standard fluid regime in elective open abdominal aortic aneurys m repair. Ann Surg. 2009; 250:2834. 16. Kalus JS, Caron MF, White CM, Mather JF, Gallagher R, Boden WE, Kluger J. Impact of fluid balance on incidence of atrial fibrillation after cardiothoracic surgery. Am J Cardiol. 2004; 94:14231425.

PAGE 32

32 BIOGRAPHICAL SKETCH Ganesh Kambhampati was born in Hyderabad, India, and is t he younger of two children; he grew up primarily in Hyderabad, India. He graduated with a medical degree from Osmania Medical College in India and subsequently moved to Bronx, New Y ork. He completed his residency in i nternal m edicine at Albert Einstein College of Medicine and decided to pursue his career in n ephrology He then moved to Gainesville, FL, where he is currently a fellow in the College of Medicine, Department of Medicine, D ivision of Nephrology at the University of Florida. He completed his Master of Science in m edical sciences with a concentration in clinical and translational research in August 2011. Ganesh has been married to his wife, Supriya, for 5 years.