Water balance , acute kidney injury and mortality of intensive care unit patients

Early volume expansion guided by the optimization parameters of microcirculatory perfusion remains a recommendation to reduce the incidence of AKI ‘and minimizes mortality in septic patients. Volume expansion is also a preventive measure for nephrotoxic AKI, as in the iodine contrast-induced AKI, to aminoglycosides, amphotericin B, rhabdomyolysis and tumor lysis syndrome, among other situations. In turn, the maintenance of ad libitum fluid infusion in critically ill patients with or without acute kidney failure, can lead to accumulated FB+ and increase morbidity and mortality. FB+ can also delay AKI diagnosis, by serum creatinine hemodilution. FB+ could be a manifestation of the inability to adequately control fluid and salt balance in critically ill and inflamed patients by changes in the kidney medullary microcirculation function integrity. In conclusion, FB+ may be an early AKI biomarker and an independent risk factor for mortality in ICU patients. Randomized clinical trials are needed to evaluate the complex interrelationship between FB+, AKI and death.

Acute kidney injury (AKI) has a high hospital incidence and is associated to significant morbidity and mortality.Sepsis, major surgery and low cardiac output are the main cause of AKI worldwide.In the majority of these situations, volume expansion is part of both prevention and therapeutic management, restoring peripheral perfusion and attenuating drug nephrotoxicity.Early and aggressive volume resuscitation in septic patients halts tissue ischemia and is associated with higher survival.However, a liberal fluid infusion strategy after six hours can cause fluid overload.Fluid overload has been associated with morbidity and mortality increase in critically ill patients.Herein, we present a review of the main studies that assessed the effects of net fluid balance/fluid overload on the morbidity and mortality of critically ill patients.We suggest that positive water balance may be used as a potential early biomarker of AKI in these patients.][8][9][10] AKI prevention involves the identification of its main causes.In the clinical context of ICU patients, AKI is predominantly multifactorial in etiology: ischemic and/or nephrotoxic. 11][14][15] In 2001, Rivers et al. 14 demonstrated in a randomized controlled study that early aggressive fluid resuscitation -driven by protocol with defined goals -significantly decreased mortality in severe sepsis.Despite numerous criticisms during these last 13 years after its publication, 15 this study markedly influenced the treatment of septic patients, having been adopted as a key recommendation in all publications from the Surviving Sepsis Campaign. 16urrently, there are three randomized, controlled clinical trials in progress, aiming to reproduce Rivers's findings: one in the U.S. (Protocoled Care for Early Septic Shock -Process), another in Australia (Australasian Resuscitation in Sepsis Evaluation randomized Controlled Trial -ARISE) and the third in England.Although Rivers et al. paper did explore the impact of volume resuscitation in the prevention of AKI, Lin et al. 17 showed in a later study, that early fluid resuscitation, similar to that used in the Rivers's protocol, reduced the incidence of AKI (55% to 39%, p = 0.015) in patients with septic shock.
Thus, adequate fluid resuscitation within the first six hours of hospital care in septic patients seems to be associated with prevention of tissue ischemia (including renal ischemia) and increased survival.A positive fluid balance (FB+) for a short period of time in patients undergoing resuscitation with these protocols can be the cost to be paid to restore tissue perfusion.However, this liberal strategy of fluid infusion after the first six hours of care can cause a continuously positive FB.Indeed, this approach was not endorsed by Rivers; 14 however the situation is common in current clinical practice, 18 with deleterious consequences to the patient.
It is also possible that the goal-oriented fluid resuscitation benefits are not due only to the administration of larger volumes of fluids to patients who specifically so require, in accordance with pre-established hemodynamic parameters, but the earliness and suitability of this measure.Administration of fluids to the limited optimization of hemodynamic parameters set previously could result in less volume of fluid infusion, minimizing FB+. 19okarik et al. 20 piloted a study with 21 burn patients and showed that fluid response determination using LiDCO (Lithium Dilution Cardiac Output) during resuscitation enabled the administration of less crystalloid volume.The study randomized two groups of burn patients: Group 1, whose resuscitation was based on systolic pressure variation (SPV) measurements and pulse pressure variation (PPV) with LiDCO; and Group 2, whose resuscitation was based on commonly used formulas (modified Brooke/Parkland) for volume expansion in burn patients.Group 1 received significantly less crystalloid than Group 2 (5090 ± 680 ml and 7.820 ± 1.050 ml, respectively, p = 0.04). 20hus, goal-oriented fluid resuscitation does not necessarily imply larger infusions of fluid and excessively positive FB. 19,20sitive fluid bAlAnce And icu mortAlity: is fb+ hArmful?
2][23][24][25] Encapsulated organs such as the kidney and liver, have limited capacity to accommodate excess fluid, which can determine increased interstitial pressure with consequent impairment of blood flow and functional deterioration. 13,22Increased kidney interstitial pressure in patients with buildup FB+ cause blood hypoperfusion and decreased glomerular filtration.Moreover, the increase in intraabdominal pressure (IAP), central venous pressure (CVP) and kidney venous pressure (KVP) increases, in conditions of fluid overload substantially contributes to the decline in glomerular filtration rate, as suggested by experimental studies.Kidney venous pressure above 30 mmHg for 2 hours in intact porcine kidneys results in a significant decline in kidney plasma flow and glomerular filtration rate. 26KVP elevation limited renal blood flow and urine formation more than arterial blood hypoperfusion. 26,27In patients with cardiorenal syndrome, with or without low cardiac output, elevation of CVP, KVP IAP may contribute to progressive kidney dysfunction. 27nd in patients undergoing elective cardiac surgery, having a high CVP is a strong predictor of postoperative AKI, regardless of low heart output. 27,28nder physiological conditions, IAP ranges from sub atmospheric to 0 mmHg.Prolonged elevated IAP or above 12 mmHg defines intraabdominal hypertension.Levels of IAP between 10 and 15 mmHg cause a reduction in mesenteric perfusion with ischemia, inflammation, and more swelling, which in turn enhances the intraabdominal pressure, closing the cycle.When the IAP achieves 20 mmHg, there is an increase in the chances of systemic clinical consequences, such as cardiovascular disorders (increased CVP, decreased venous return and cardiac output), kidney disorders (decreased renal plasma flow, increased KVP, renal microvascular congestion, increased renal vascular resistance, elevated catecholamines, angiotensin II, inflammatory cytokines and renal capsule hypertension), lung (reduced chest wall compliance, increased KVP, increased dead space, hypercapnia and hypoxia) and cerebral (increased intracranial pressure and decreased cerebral perfusion pressure), configuring an abdominal compartment syndrome. 29,30rimarily, the abdominal compartment syndrome was described in patients undergoing abdominal surgery or trauma victims; however, currently, it is increasingly diagnosed in patients undergoing massive fluid resuscitation, multiple transfusions, hypothermia or coagulation disorders, i.e., patients with successive FB+.There is evidence that intraabdominal hypertension present at ICU admission is associated with increased risk of severe organ dysfunction (including AKI) and the highest mortality. 29,30nother no less important aspect, which may contribute to the adverse outcomes associated with FB+, is the type of solution used for fluid resuscitation.][33] The use of 20% albumin has been associated with lower incidence of AKI and reduced mortality in cirrhotic patients with spontaneous bacterial peritonitis, but its routine use ends up being limited by the higher cost of crystalloid. 34Thus, the Kidney Disease Improving Global Outcomes (KDIGO) Clinical Practice Guideline for AKI in its latest edition, suggests that isotonic crystalloids are preferable, instead of synthetic or non-synthetic colloids for volume expansion in patients at risk of developing AKI or AKI installed in the absence of a hemorrhagic shock. 32owever, crystalloids are not without risk.6][37] Chowdhury et al. 36 led a crossover study in healthy volunteers, randomized to receiving alternating 2 liters of isotonic 0.9% saline (containing 150 mmol/l chlorine) and 2 liters chArt 1 PotentiAl side effects of A Positive fluid bAlAnce (fb+) of Plasmalyte (containing 98 mmol/l chloride) IV.Upon receiving isotonic saline 0.9%, individuals took longer to have the first urination (from 90 to 142 min, p = 0.006), reduced urine volume of 833 ml to 523 ml (p = 0.002) and reduced kidney plasma flow velocity and kidney cortical perfusion assessed by MRI. 36Yunos et al. 37 conducted a prospective, controlled pilot study in critically ill patients, comparing the liberal strategy of fluid infusion containing chlorine (as gelatin and 0.9% saline) with the use of balanced solutions or with less chlorine content such as Hartmann's solution (containing 109 mmol/l chlorine) and Plasmalyte.Fluid handling with balanced solutions reduced the frequency of hyperchloremia, metabolic acidosis and was associated with significant reduction in the incidence of AKI and need for supportive kidney therapy. 37Therefore, not only the amount of fluid infused in the volume resuscitation, but also its composition, can interfere with unfavorable outcomes associated with FB+.[40][41][42][43][44][45][46][47][48][49][50][51][52][53][54] Bouchard et al. 49 evaluated the adult population with AKI in the PICARD study and found that, at the time of AKI diagnosis, the percentage of fluid accumulation in relation to the patient's weight upon ICU admission was lower in survivors than in nonsurvivors (p = 0.01).However, this difference was not statistically significant after adjustment for APACHE III (p = 0.12) scores.When the rate of fluid accumulation from all patients was greater than 10%featuring fluid overload, mortality at 30 and 60 days rose from 25 to 37% (p = 0.02) and 35 to 48% (p = 0.01), respectively.The risk of death associated with fluid overload in dialysis patients was 2.07 (95% CI 1.27 to 3.37), adjusted for severity of illness and dialysis modality.In patients not on dialysis with fluid overload, the risk of death was 3.14 (95% CI 1.18 to 8.33).Patients who kept fluid accumulation during hospitalization had higher mortality, which was proportional to fluid buildup (p < 0.0001).Patients in whom dialysis was able to correct fluid overload had lower mortality than those who remained with fluid overload after dialysis (35% vs. 56% respectively, p = 0.002).The adjusted risk of death associated with fluid overload at the end of dialysis was 2.52 (95% CI 1.55 to 4.08).Patients with fluid overload at the time that creatinine was higher were less likely to recover kidney function (35% vs. 52% respectively,  p = 0.007).This latter finding suggests that, after AKI installing, the administration of excess fluid exerts no protective effect on kidney recovery.However, it is not possible to say that the fluid overload per se was the cause of delayed kidney recovery, or if patients with fluid overload already had a more severe AKI, and therefore also had a later recovery. 37y means of a multicenter prospective observational study with 296 patients from 17 ICUs, Vaara ST et al., 53 also reported that fluid overload (defined as fluid accumulation above 10% of the baseline weight) at the time of the Renal Support Therapy (RST) indication, was associated with a higher risk of death at 90 days (OR 2.6), after adjustment for severity of illness, RST onset time, RST modality and sepsis. 536][47][48][49][50][51][52][53] None of these studies can clarify whether patients with FB+ have higher mortality for being more severe (and therefore require greater investment on drugs, antibiotics, sedatives, parenteral nutrition etc.) or whether the FB+ has a really independent pathophysiological involvement in the death process.Alternatively, we can speculate that FB+ is an early biomarker of AKI, which is associated with mortality (Figure 1).There are only a few randomized clinical trials that studied the association between mortality and FB+.Wiedemann et al. 38 randomized 1,000 patients with acute respiratory distress syndrome (ARDS) into two fluid management strategies, one conservative and the other liberal.This is the largest prospective, controlled study evaluating the impact of fluid overload on the duration of mechanical ventilation and mortality in ARDS.The conservative strategy of fluid infusion reduced the duration of mechanical ventilation, without worsening kidney function; however, mortality was similar in both grupos. 38he largest randomized controlled trial designed to evaluate the impact of strategies for fluid resuscitation on mortality was developed in six clinical centers in Africa. 54The authors studied septic children (< 12 years of age) with signs of peripheral hypoperfusion who were randomized to resuscitation with a bolus of saline (20 ml/ kg IV at 1 hour), albumin bolus (20 ml/kg IV at 1 hour) or no bolus (control).The three groups equally received 2.5 to 4 ml/kg/h as maintenance fluids, blood transfusions if hemoglobin was below 5 g/dL and appropriate antibiotics.Children with gastroenteritis, severe malnutrition and shock of noninfectious causes were excluded.Children with severe hypotension were randomized to saline bolus or albumin, with no control group.The 48-hour mortality was higher in the groups receiving intravenous bolus of 0.9% saline or albumin compared to the group who received bolus (11% both groups versus 7% bolus, RR 1.45, p = 0.003). 54No FB data from each group was shown.Furthermore, the study was conducted in the pediatric population and health centers without advanced life support, making it difficult to generalize these results for adult patients admitted to intensive care units.
The morbid effects of fluid overload in the various organs may be responsible for the association between mortality and FB+.However, another aspect that could also contribute to higher mortality in patients with fluid overload is reduced plasma concentrations of hydrophilic antibiotics by increasing the volume of their distribution. 55As will be discussed below in the topic, it is known that inflamed/septic patients develop endothelial damage with increased capillary permeability predisposing to fluid shifts from the intravascular compartment to the interstitium.Hydrophilic antibiotics such as aminoglycosides, beta-lactam and glycopeptides may follow this fluid diversion, resulting in subtherapeutic plasma concentrations. 55Thus, the complex association between FB+ and mortality has not been clearly elucidated.FB+ would be a severity marker of cardio/kidney/microvascular severity or is it an independent factor for higher mortality in critically ill patients?(Figure 1).

positive fluid bAlAnce And Aki diAgnosis: does fb+ interfere with the Aki diAgnosis?
Currently, AKI diagnosis is based on the RIFLE criteria (Risk, Injury, Failure, Loss, End stage Kidney Disease), AKIN (Acute Kidney Injury Network) and KDIGO (Kidney Disease Improving Global Outcomes), which are based on serum creatinine increase and reductions in urinary volume. 56,57These criteria were constructed from the observation that small increases in serum creatinine 57 and diuresis reductions 58 imply a worse prognosis for patients.The three classification systems are divided into severity stages and there is growing evidence of a correlation between the AKI stage by RIFLE and AKIN criteria and mortality. 59luid balance was not included in the diagnosis of AKI by any of these currently adopted criteria, a fact which leads us to reflect on renal physiology.
It is known that a patient with normal renal function and usual diet (generating 800 mOsm of excreta per day) is able to eliminate up to 16 liters of maximally dilute urine (50 mOsm/L), for physiological suppression of antidiuretic hormone (ADH). 60In turn, when a critically ill patient develops AKI diagnosed by standard criteria and receives large fluid infusions (antibiotics, sedation, vasoactive drugs etc.), it is easy to understand why he develops FB+, since there is obvious impairment in their ability to excrete the sodium overload received.However, critically ill patients without a diagnosis of AKF by the RIFLE, AKIN or KDIGO criteria also evolve with positive FB.Why they could not adequately increase the excretion of free salt and water, in response to sodium overload?
There are medical conditions in the ICU that may predispose FB+, even in patients without AKI.Hypotension, low cardiac output, vasodilation of sepsis (albeit without hypotension), hypoalbuminemia, mechanical ventilation (MV) are able to activate the neuroendocrine system (antidiuretic hormone, sympathetic nervous system, renin-angiotensin-aldosterone system), with tubular retention of free water and sodium. 12,13ecause these patients usually receive large infusions of water and salt, the result is positive FB and edema.4]25 Other pathophysiological mechanisms that mediated the development of FB+ in critically ill patients without a diagnosis of AKI by current criteria, possibly involve two molecular systems: the angiopoietin/Tie (Ang/Tie) 61,62 system and the endothelial glycocalyx. 63,64The angiopoietin/Tie (Ang/Tie) system participates in major endothelial functions: angiogenesis, fluid homeostasis maintenance, electrolytes and protein transport by endothelial cells and inflammation/coagulation triggered by endothelial injury.These three endothelial functions are altered in multiple organ dysfunction syndrome (MODS): blood flow regulation is changed, vascular permeability becomes higher with fluid and cell extravasation into the surrounding tissues and the mechanisms of inflammation and coagulation are activated.Experimental and clinical studies suggest that the Ang/Tie system participates in the vascular barrier dysfunction observed in critically ill patients. 61,62ngiopoietin 1 (Ang-1), a glycoprotein produced by pericytes causes vascular stability by binding to the transmembrane Tie-2 receptor of endothelial cells.Rather, angiopoietin 2 (Ang-2) released by the Weibel-Palade corpuscles of endothelial cell cytoplasm by inflammatory stimuli, antagonizes the stabilizing effect of Ang-1 by binding competitively to the same receptor Tie-2, weakening the intercellular junctions, predisposing to fluid leakage, inflammation and coagulation.Patients with septic shock have high plasma levels of Ang-2, that correlate with FB+, pulmonary dysfunction and mortality. 62t is also possible that the endothelial glycocalyx is reduced in situations of ischemia and/or inflammation, such as sepsis or systemic inflammatory response syndrome (SIRS). 63Glycocalyx is a structure composed of glycosaminoglycans and proteoglycans, which covers the endothelial surface facing the capillary lumen, forming a barrier against the passage of macromolecules and limiting the adhesion of inflammatory cells.The reduction of its thickness, in situations of ischemia/inflammation, causes an increase in vascular permeability and leukocyte adhesion, contributing to leakage of fluid and macromolecules to the interstitium. 63,64herefore, ischemic and inflammatory attacks, so frequent in critically ill patients cause endothelial dysfunction mediated by changes in the angiopoietin 2/Tie-2 system and endothelial glycocalyx stripping, contributing to the diversion of fluid and macromolecules from the intravascular compartment to the interstitium and, again, activating the neuroendocrine system to retain fluids. 63,64n the other hand, critically ill patients may develop early and subtle AKI alone or together with events described above, manifested by the inability to properly control the fluid and sodium balance.The ability to concentrate and dilute urine depends on the functional integrity of the renal medullary microcirculation and may be impaired in the very early stages of kidney disease.While the cortical microcirculation is able to make its self-regulation through changes in tone of afferent and efferent arterioles up to certain limits for mean arterial pressure (MAP), the medullary microcirculation seems to be more dependent on MAP and renal perfusion pressure, more susceptible to ischemia and endothelial dysfunction in circulatory shock. 65hus, positive FB can potentially be an AKI biomarker, preceding the elevation of creatinine or decreased urine output.Moreover, fluid overload can cause hemodilution and underestimate the level of serum creatinine, delaying AKI diagnosis by the usual criteria. 18By analyzing data from the PICARD study, Macedo et al., proposed using the following formula to adjust serum creatinine for FB+: Adjusted creatinine = dosed creatinine x correction factor* * correction factor = ((weight upon admission x 0,6) + Σ daily built up FB)/(weight upon admission x 0.6).
The use of creatinine adjusted for this "correction factor" enabled the anticipation of the AKI diagnosis, in at least one day. 18his set of facts suggests that it is necessary to consider the possible inclusion of FB+ as an AKI criterion besides creatinine and urinary volume.Three observational studies, a retrospective with 90 patients, 66 another prospective with 100 patient 67 and the last: a secondary analysis of a prospective study involving 98 children suggested that FB+ is an early marker of renal dysfunction.All studies evaluated patients undergoing cardiovascular surgery 65,66 Dass et al. 66 conducted a post-hoc analysis of the Nesiritide Study, a randomized placebo-controlled clinical trial, from a single center, a study designed to assess the impact of prophylactic use of nesiritide in patients requiring renal support therapy and/or in the mortality of patients undergoing cardiovascular surgery.These authors investigated the hypothesis that FB+ in the first 24 hours after cardiac surgery could be an indicator of postoperative AKI.In this study, the incidence of AKI in patients with a median FB greater than +849 ml/24 hours was 80%, versus 25% in those patients with a median FB -1221 ml/24 hours (p = 0.001).
In a prospective observational study, Kambhampati et al. 67 divided the groups into FB quartiles, from the intraoperative up to 48 hours postoperatively.AKI incidence in the 4 th quartile (FB+ median of 5000 ml/24 hours) was 52%, compared with a 16% incidence of AKI in the 1 st quartile (FB+ median of 500 ml/24 h), p = 0.016.Multivariate analysis adjusted for confounding variables such as age, diabetes, hypertension, cardiopulmonary bypass time greater than 200 min, type of surgery and rate of basal glomerular filtration revealed that the higher FB+ quartile (4 th quartile) was associated with a significantly higher risk of AKI (OR 4.89; 95% CI; 1.38 to 24.1, p = 0.046). 67he third and most recent study evaluated whether fluid overload in the early postoperative period of cardiac surgery in 98 children would be associated with higher morbidity, including increased incidence of AKI. 68Early postoperative fluid overload was defined as FB+ greater than 5% of baseline weight, from the immediate postoperative period in the ICU until the first postoperative day.Fifty percent of children with fluid overload upon ICU admission developed AKI in the first postoperative day, whereas only 14.4% of those who did not have early fluid overload developed AKI (p = 0.023).Fluid overload preceded the onset of AKI, which occurred in any postoperative day.8% fluid overload (equivalent to 80 ml/kg) showed a specificity of 90% for the subsequent development of AKI RIFLE I or F. Fluid overload was also associated with a longer hospital stay (3.5 days more), two more days under inotropes, and higher prevalence of prolonged MV. 68 To date, there is no available evidence testing the hypothesis that FB+ is an early marker of AKI, also in non-surgical patients.

AbstrAct
Early volume expansion guided by the optimization parameters of microcirculatory perfusion remains a recommendation to reduce the incidence of AKI 'and minimizes mortality in septic patients.Volume expansion is also a preventive measure for nephrotoxic AKI, as in the iodine contrast-induced AKI, to aminoglycosides, amphotericin B, rhabdomyolysis and tumor lysis syndrome, among other situations.In turn, the maintenance of ad libitum fluid infusion in critically ill patients with or without acute kidney failure, can lead to accumulated FB+ and increase morbidity and mortality.FB+ can also delay AKI diagnosis, by serum creatinine hemodilution.
FB+ could be a manifestation of the inability to adequately control fluid and salt balance in critically ill and inflamed patients by changes in the kidney medullary microcirculation function integrity.
In conclusion, FB+ may be an early AKI biomarker and an independent risk factor for mortality in ICU patients.Randomized clinical trials are needed to evaluate the complex interrelationship between FB+, AKI and death.

Figure 1 .
Figure 1.Possible associations between acute kidney injury (AKI), positive fluid balance (FB+) and mortality.The FB+ may be associated with increased mortality for several pathophysiological pathways (blue arrows): to cause visceral and peripheral edema, making the organic functioning difficult; by dilution of hydrophilic antibiotics, reducing their effectiveness, or even by hemodilution of serum creatinine, delaying diagnosis of acute kidney injury (AKI).However, the FB+ is more often found in patients with higher clinical severity (orange arrow) and in patients with AKF (red arrow), both situations already independently associated with increased mortality.

tAble 1 studies
Assessing the imPAct of fluid bAlAnce (fb) in the moRbidity of inPAtients in An intensive cARe unit(icu) AKI: Acute kidney injury; CRT: Controlled randomized trial; ARDS: Acute respiratory distress syndrome; API: Acute pulmonary injury; MV: Mechanical ventilation; MOSD: Multiple organs and systems dysfunction.

tAble 2 studies
Assessing the imPAct of Positive fluid bAlAnce in the moRtAlity of in PAtients in An intensive cARe unit(icu) FO: Fluid overload or gain > 10% of the ICU admission weight; FB ml/24 hours = mean fluid balance in the 24 hours; AKI: Acute kidney injury.* Mean daily FB in the first 3 days of ICU admission; ** Mean daily FB during the entire ICU stay.