Intravenous Fluid Therapy in Traumatic Brain Injury and Decompressive Craniectomy

Abstract

The patient with head trauma is a challenge for the emergency physician and for the neurosurgeon. Currently traumatic brain injury constitutes a public health problem. Knowledge of the various supportive therapeutic strategies in the pre-hospital and pre-operative stages is essential for optimal care. The immediate rapid infusion of large volumes of crystalloids to restore blood volume and blood pressure is now the standard treatment of patients with combined traumatic brain injury (TBI) and hemorrhagic shock (HS). The fluid in patients with brain trauma and especially in patients with brain injur y is a critical issue. In this context we present a review of the literature about the history, physiology of current fluid preparations, and a discussion regarding the use of fluid therapy in traumatic brain injury and decompressive craniectomy.

Introduction

Traumatic  brain  injury (TBI) is a major public health  problem  and  a leading cause of  death and disability [1]. It is frequently accompanied by haemorrhagic shock (HS) [2-5]. The mechanism for adverse outcome in patients with combined TBI and HS may be due in part to the secondary ischemic injury of already vulnerable brain following loss of cerebral auto-regulation and or to adverse effects of TBI itself on  the normal  compensatory  response to  HS [6]. Rapid infusion as quickly as possible of large volumes of crystalloids to restore blood volume and blood pressure is now the standard treatment for patients with combined TBI and HS [7,8]. Perioperative fluid administration is an important aspect of surgical care, but is often poorly understood [9-13], and continues to be an empirical exercise, with nagging questions about its efficacy and complications [14].

Fluid therapy (FT), as the name implies is a treatment with fluids [15]. The final goal of fluid management is to optimize the circulatory system to ensure the sufficient delivery of oxygen to organs [16]. FT is needed for the following conditions:

1. Normal maintenance;

2. Blood or  fluid  loss due  to  wounds,  drains, induced diuresis etc;

3. Third space losses socalled fluid sequestration in tissue oedema or ileus;

4. Increased systemic requirements resulting from fever and hypermetabolic state.

TF should be tailored to match these requirements [17]. Intravenous (IV) fluids may be broadly classified into  colloid  and  crystalloid solutions.  They  have very different physical, chemical, and  physiological characteristics. Colloid solutions can be natural (albumin)  or synthetics (gelatins, dextrans, and hydroxyethyl starches).

Goal-directed fluid therapy (GDT) aimed at optimizing cardiac output  and oxygen delivery has been shown to improve the   outcome of  high-risk surgical patients [18].

Most   information   presented   herein   is  derived from the fluid management for the surgical patient in general, and those who were critically ill, such as trauma  patients.  Primary  studies  on  preoperative fluid therapy for decomprssive craniectomy (DC) are sparse.

The fluid management for patients undergoing elective major  surgery, e.g., neurotrauma   surgery, is  controversial  [19-21]. During  the  perioperative period, many pathophysiological changes occur that alter  the  normal  efficiency of  fluid  homeostasis. Despite this, perioperative fluid prescription is often poor, being based on an insufficient knowledge of water and electrolyte requirements and distribution [22-25].  Perioperatively, crystalloids, colloids  and blood components are required to meet the ongoing losses and for maintaining cardiovascular stability to sustain adequate tissue perfusion [26,27].

Intravenous  fluids maintain  hydration  while patients are unable to drink and replace losses that occur as a result of surgery [28]. Severity of illness, magnitude  and  duration  of surgery, comorbidities and the host response to injury, influence the perioperative  fluid  needs.

Although  the  principle goal of fluid administration is to maintain adequate tissue perfusion and the perils of over and under- resuscitation  are  well documented,  there  are  no standards of care guiding for perioperative fluid administration [28]. The aim of this work is to review the current  topics of fluid management in patients with traumatic  brain injury and the candidates for decompressive craniectomy.

History of Modern Fluid Therapy

 The  intravenous  fluid therapy  (IVF) first gained importance   in  the  treatment   of cholera  in  the 1830s [29-34], with the reports of William Brooke O’Shaughnessy on his  terminal  cholera  patients’ blood observations [29]. Their blood were thick and obscure, thus concluded that there was water deficit in those patients [31]. Aiming to replace the corporal fluid loss, the 0.9% physiologic saline solution was used  for surgical  patients.  The  perioperative  use of IVF to compensate  for the  injurious effects of anaesthesia began  in1880s. Clinical improvements were consequently noted, though the adverse effects of saline were observed. Thomas Latta was the first to administer intravenously water and salt solutions to dying patients with no favorable results [29,35]. The use of different mixes of water and salt in conjunction with the poor hygienic practices of those days ,did not prove safe, thus   the fluid therapy did not   initially gain acceptance [15].

In 1880, Sidney Ringer observed the different protoplasmic activity of sodium, calcium, potassium and chloride salts, and introduced Ringer’s solution [15]. George Crile in 1899, using a hemorrhagic shock animal model, studied the solution and recommended it in warm form. War injuries during the First World War were treated with primitive saline and colloid solutions. The Gum arabic, a natural colloid from the Acacia senegalis tree used by W.C. Cannon, was a high point colloid [15].

In 1924, the intravenous "drip" was introduced  by Rudolph  Matas. In 1930, Hartmann  and  Senna in order to avoid the hyperchloremic acidosis resulting from the use of Ringer´s solutions [36], added sodium lactate, allowing sodium to be linked to the excessive chloride.  This  facilitated  the  lactate  metabolism, thus, giving rise to the Hartman solution or Ringer- lactate. The work of Ringer, Hartmann,  and others emphasized the importance  of the composition  of IV fluids and laid the foundations for the balanced solutions in use today [37].

During the Second World War, blood and plasma were massively administered, even in the battlefield, aimed at prolonging the lives of injured soldiers. In general, FT has been changing continuously, depending on current trends [15]. As the  metabolic response to  injury  was increasingly investigated in the 1940s and 1950s, the cause of post- operative oliguria was widely debated by Moore and Shires, the most  prominent  surgeons [37]. During Vietnam War, the preservation of renal function, became a therapeutic objective, thus, allowing the use of large amounts of crystalloid for the hemorrhagic shock in the Da Nang Army Hospital. At this time pulmonary  complications called Da Nang lung or wet lung syndrome, previous denominations  of the present Adult Respiratory Distress Syndrome (ARDS), was observed which derived from prescribing large volumes of fluids.

These differences in opinion, coupled with reports on survival of injured soldiers from the Korean war who   received large IV fluid infusions, dictated the surgical practice of liberal IV fluid administration until  very recently  [37].  Recent  research  in  fluid therapy has explored the concept of fluid restriction. Shoemaker and colleagues also pioneered the concept of fluid administration  to achieve supranormal indices of cardiorespiratory function, which has led to  the  advent  of  goal-directed fluid therapy  [37]. Alongside the development of balanced solutions, the renewed focus on perioperative fluid therapy has led to IVF administration being guided by physiological principles with a new consideration  of the lessons gleaned from history [37]. With military medicine advances, fluid therapy has attracted particular attention, and become increasingly important  when it is combined with hemotherapy. But, with no clear ideas about the proper volume. Currently, fluid therapy stills with large controversies.

Fluid Physiology

The fundamentals governing fluid and electrolyte management in patients date to the 19th century [38]. In the first half of the 20th century work by Gamble and  Darrow and  colleagues defined the electrolyte content of extracellular, intracellular and interstitial fluid compartments [38]. The body of adult human contains 60% water, of which two-third is intracellular and the remaining one- third is in the extracellular space, which in turn is divided between intravascular and extravascular or interstitial compartments [39,40].

The  interstitial  compartment  is actually a  matrix, a collagen/gel substance that  allows the interstitium to  provide  structural  rigidity  which  resists against gravity and can maintain  structural  integrity during extracellular volume depletion. The collagen/gel interstitial space, especially in skin and connective tissue, is an important reservoir of extracellular fluid [38].

The  total  intravascular  volume, also  referred  to as blood volume, is approximately 5 liters of which 2 liters (40%) form intracellular structures such as red and white cells and platelets and 3 liters(60%) constitute extracellular component (plasma) [39,40]. Extracellular compartment is important for oxygen and nutrients transport, and the elimination of carbonic anhydride and other products from cellular metabolism. Another  compartment  is transcellular that includes fluids not equilibrated with the other fluids, and constitutes synovial, cerebrospinal fluids, gastrointestinal  secretions, etc. This  compartment through the lymphatic system returns the fluids to the intravascular space [40].

The cells and the intravascular space have membranes that preserve their structural integrity and allow the molecule and fluid interchange between different compartments.

The  main  function  of  membranes is to  preserve osmolarity and  the electronegativity in of each  compartment. The cell wall separates the intracellular space from the extracellular compartment. The capillary endothelium  and  the walls of arteries and veins divide the  extracellular compartment into the intravascular and the interstitial areas such as tissue or extravascular compartments. Water moves freely through cell and vessel walls and enters all these compartments. The energy-dependent Na/K adenosine triphosphatase in cell walls extrudes Na+  and Cl- and maintains a sodium gradient across the  cell  membrane  with  Na+    as  an  extracellular ion. The capillary endothelium  is freely permeable to small ions such as Na+  and Cl-, but is relatively impermeable  to  larger molecules such as albumin and   the   semisynthetic  colloids like  gelatin  and starch, which normally remain in the intravascular space. Plasma is a solution in water of inorganic ions including predominantly sodium chloride, simple molecules such as urea, and larger organic molecules like albumin and the globulins. Plasma and interstitial fluid are highly interchangeable. Fluid exchange through a capillary is regulated by Starling’s law, which mathematically summarizes the forces governing the flow of fluid out of blood vessels into surrounding tissues, and expressed as Qf = Kf [(Pc − Pi) − R(πc − πi)].

Where Qf is the total fluid flux out of capillaries (not the quantity, but just the speed of water movement [16]) and Kf is the filtration coefficient (the product of the membrane conductance  and the membrane surface area), Pc is intravascular hydrostatic pressure, Pi is interstitial hydrostatic pressure, πc is colloid osmotic pressure within the vasculature, πi is interstitial colloid osmotic pressure gradient across the vessel wall, and R is the oncotic reflection coefficient, the tendency of a membrane to impede the passage of oncotically active particles (Starling, 1896) [16,41]. R of 0 indicates a membrane that is totally permeable to protein while R of 1 represents a membrane that completely prevents protein diffusion. Distribution  terminates  when the  balance of the hydrostatic pressure and the osmotic pressure cancel each other out. Because the interstitium consists not only of free space but also of absorbent gel, captured water in the gel does not contribute to lowering the osmotic pressure in the interstitium [16]. Therefore, the osmotic pressure does not easily change until the gel is saturated by flow of water. This is a mechanism of edema formation. Thus, Starling’s law does not determine the distribution ratio between plasma and interstitium, it just explains the movement of water through the capillary wall [16].

The original interpretation of an equilibrium including fluid reabsorption at the venous end of the microcirculation is now known to be incorrect through actual measurement of the pressures involved. Rather, a steady state is involved, with a level of permeability to plasma proteins in the microvascular walls. Net fluid movement occurs in the vessels from the intravascular to the perivascular space [42]. The fluid transfer is mediated by the endothelial glycocalyx layer (EGL), a physiological entity discovered and studied over the past 30 years [43]. A model for fluid transfer across the EGL [44] accounts for the discrepancies observed in fluid transfer as predicated by Starling's original equation, and proposes a modified hypothesis based on pressures involving the generation of fluid through the glycocalyx rather than the interstitial space [45] modifying the Starling equation to: Qf = (Pc – Pi)-R (πc – πg).

Where Pi and πg are the hydrostatic and osmotic pressures respectively, exerted by the formation  of ultrafiltrate across the glycocalyx [46]. While the EGL is the conduit for water passage from the intravascular to the extravascular space, plasma proteins cross the endothelial barrier through  a separate pathway, the large pore system [46].

This model is perturbed  by a number  of factors during anesthesia and surgery. Patients scheduled for surgery are presented with a variety of conditions that result in altered fluid distribution. Many anesthetic drugs  like IV  induction  medications  and volatile anesthetics cause vasodilation, leading to a reduction in the ratio between the circulating volume and the capacity of  the  intravascular  space, or  myocardial impairment, causing a reduction in flow through the circulation.

Fluid shifts between compartments may also reduce the circulating volume representing third-space losses and loss of intravascular fluid into the interstitium because of altered endothelial permeability in sepsis and inflammatory states .[39]

The following section is devoted to a resume of the main characteristics of different kind of solutions.

Crystalloids

 A crystalloid fluid  is a  solution  of  small water- soluble  molecules  that   can  diffuse easily  across semi-permeable membranes. The properties of these solutions  are  largely determined  by their  tonicity (osmolality  relative to  plasma)  and  their sodium content (affecting their distribution  within body compartments)  [47]. They redistribute  throughout the extracellular fluid (ECF) compartment, of which 75% is interstitial fluid. This suggests that 4 litres of crystalloid are required to replace a blood loss of 1 litre. [48] Studies have shown that the volume kinetics of infused crystalloid solutions differ between normovolaemic and hypovolaemic patients. [49]

IV infusions of isotonic saline solution only expand the intravascular space by a maximum of one-third of the volume used in normal  subjects, with only 16% left after 30 minutes. The volume of crystalloid required  to  replace  an  acute  blood  loss  remains 3-4-fold because of  redistribution  into,  and  rapid elimination from, the ECF [48].

Isotonic Solutions

Isotonic or iso-osmolar solutions, with an osmolality ≈300  mOsm/L,  such  as sodium   chloride  0.9% (normal  saline), Ringer’s solution or plasmalyte, do not  change plasma osmolality and do not  increase brain water content [50]. They also contain sodium at physiological plasma concentrations. These fluids distribute freely within the ECF compartment causing little change in sodium concentration and osmolality. As a result, this limits the movement of water out of the  extracellular fluid (ECF) into  the  intracellular fluid (ICF) compartment and vice versa. Commercial lactated Ringer’s solution  is not  truly  iso-osmolar with respect to plasma. Its measured osmolality is ⋲254 mOsmol/kg, which explains why administration of large volumes can reduce plasma osmolality and increase brain water content and intracranial pressure (ICP). [50]

Hypotonic Solutions

 Large amounts of hypo-osmolar or hypotonic fluids reduce  plasma  osmolality, drive water  across  the blood brain barrier (BBB), and increase cerebral water content and ICP. A solution of 5% dextrose (D5W) is essentially water   since the sugar is metabolized very quickly and provides free water which disperses throughout the intracellular and extracellular compartments with little use as a resuscitative fluid [50]. Therefore , hypo-osmolar crystalloids (0.45% NaCl or D5W) should be avoided in neurosurgical patients [50].

Hypertonic crystalloids: Mannitol and hypertonic saline

Osmotherapy   agents  such  as  hypertonic  saline (HTS) are currently used in the treatment of patients with post-traumatic  cerebral edema and raised ICP resulting from  TBI [51].It  is believed   to  have a particularly useful role in the treatment of ICP whilst administering small volume fluid resuscitation [52]. HTS  solutions  typically improve  cardiovascular output as well as cerebral oxygenation whilst reducing cerebral oedema. Hypertonicity seems to affect some innate immune-cell functions, specifically neutrophil burst activity in preclinical studies, probably providing beneficial impact on modulation of the inflammatory response to trauma [53-58]. Clinical studies however do not provide compelling evidence to support  the use of HTS either for TBI or  for  haemorrhagic  shock. A  small  randomized clinical trial (RCT) reported a significant reduction in mortality when comparing a 250 ml bolus of HTS/ dextran  with  isotonic  saline in  222 patients  with haemorrhagic  shock  [59].  But  many  other  RCTs have not  demonstrated  reduction    in mortality  in this group of patients [60-63], including the most recent and largest RCT comparing HTS, HTS/ dextran and normal saline. This trial recruited two separate cohorts, one with TBI (n = 1087) [64] and the other with haemorrhagic shock (n = 853) [65]; with primary endpoints of neurological outcome at 6 months after TBI and 28 day survival respectively. The TBI study was terminated  early due to futility, as  interim   analysis was  unable   to   demonstrate an  improvement  in  neurological status  or  indeed mortality at 6 months. Another  study demonstrated no significant difference in mortality at 28 days, and was terminated early for concerns of potential (albeit statistically non-significant)  increase  in  mortality observed with a subgroup of patients receiving HTS but no blood transfusions within the first 24 h [65]. Wade et al., [66] undertook  a cohort  analysis of individual patient data from a previous prospective randomized    double-blinded    trial    to    evaluate improvements in survival at 24 hours and discharge after  initial  treatment  with  HSD in  patients  who had TBI (head region Abbreviated Injury Score ≥4) and  hypotension  (systolic blood pressure ≤90 mm Hg). They found that treatment  with HSD resulted in  survival until  discharge of  37.9% (39  of  103) compared  with  26.9% (32  of  119) with  standard care (p=0.080). Using logistic regression, adjusting for  trial  and  potential  confounding  variables, the treatment effect can be summarized by the odds ratio of 2.12 (p=0.048) for survival until discharge. They concluded that patients with traumatic brain injuries in the presence of hypotension and receiving HSD are about twice as likely to survive as those who receive standard of care.

Rockswold et al., [67] examined the effect of hypertonic saline on ICP, cerebral perfusion pressure (CPP), and brain tissue oxygen tension (PbtO2), and found that hypertonic saline as a single osmotic agent decreased ICP while improving CPP and PbtO2 in patients with severe traumatic brain injury. Patients with higher baseline ICP and lower CPP levels responded to hypertonic saline more significantly.

Colloids

 Colloids are fluids with larger, more insoluble molecules that do not readily cross semi-permeable membranes, across which they exert oncotic pressure. Water is drawn in from the interstitial and ICF by osmosis. Their movement  out  of the intravascular space and their duration of action is  dependent on their molecular weight, shape, ionic charge and the capillary permeability  [47].  Apart  from  albumin, all colloids are polymers and contain particles with different  molecular weights [48]. They may increase plasma volume by more  than  the volume infused, because of their higher osmolality; hence the term plasma expanders [48]. Studies suggest they can cause significant impairment of clot formation activity [68,69].

Albumin Albumin is a multifunctional, non-glycosylated, negatively charged plasma protein, with a molecular weight of 69 kD. It is a biological therapeutic, manufactured  from an inherently variable material source  using  a  variety of  purification  techniques. Albumin is an  effective volume expander, has not been  associated  with  allergic-type  reactions,  and has no intrinsic effects on clotting [50]. Its use as reanimation fluid has not been linked to better survival compared with the synthetic colloids, a fact that together with costs, discredits its use in critically ill patients [70]. There are in different concentrations: iso-oncotic (4-5% albumin) and hyper-oncotic (20% albumin). The later has adverse renal events.

Synthetic colloids

 Gelatins

 Gelatin products are semi-synthetic colloids derived from bovine collagen and prepared as polydispersive solutions by multiple chemical modifications [48,71]. Gelatins for volume therapy have been withdrawn from the US market due to the high rate of anaphylactic reactions  [71].  Conventional  gelatin  preparations have a mean molecular weight of 30-35kDa and a low molecular mass range. Their intravascular persistence is short (2-3 hours), particularly for the urea-linked gelatins, with rapid renal excretion (80% molecules < 20 kDa). Since the cross-linked gelatin molecules contain negative charges, chloride concentrations of the solvent solution are reduced in contrast to other types of colloid. Since the latter fact results in slight hyposmolality, infusion of large amounts  of gelatin solutions may reduce plasma osmolality and ultimately foster the genesis of intracellular edema [71]. The rapid urinary excretion of gelatin is associated with increased  diuresis  and  has  to  be  substituted  by adequate crystalloid infusion to prevent dehydration. Gelatin  infusion  may  furthermore  increase blood viscosity and  facilitate red  blood  cell aggregation without influencing the results of crossmatching. Severe anaphylactoid  reactions  are  low  (though more likely with gelatins than with other colloids), and usually occur only with rapid infusions (1/13000 for   succinylated  gelatin;  1/2000  for   urea-linked gelatin), although much less with newer formulations. Reactions are usually mild (incidence < 0.4%).(48) Clinically, they have little effect on coagulation [48].

Dextranes

 These are neutral, high-molecular-weight glucopolysaccharides based on  glucose monomers. Dextranes are derived from the action of the bacterium Leuconostoc mesenteroides on a sucrose medium via the dextran sucrose enzyme. This produces a group of branched polysaccharides of 200,000 glucose units. Subsequent  partial  hydrolysis produces  molecules of mean MW 40, 60, 70 and 110 kDa, with half-lives ranging from 15 minutes to several days. They are mainly excreted via the kidneys (70%), with the rest metabolized  by  endogenous  dextranase. They  are relatively cheap (£4-5 per 500 ml).

Dextran 40 is hyperoncotic and initially acts as a plasma expander before its rapid elimination by the kidney. Its main use is in promoting peripheral blood flow in cases of prophylaxis for deep vein thrombosis and  arterial insufficiency. Dextran  70 and  110 are mainly used for plasma expansion; 6% dextran 110 is no longer available clinically. Blood flow improvement results from a reduction in blood viscosity, possibly by coating  the  vascular endothelium  and  cellular elements of blood, thus reducing their interaction. Dextrans inhibit  platelet adhesiveness, enhance fibrinolysis and may reduce factor VIII activity. Doses above 1.5 g/kg cause bleeding tendency. Initial use should be limited to 500–1000 ml with a restriction of 10–20 ml/kg/day thereafter.

Modern solutions do not interfere with blood cross- matching or cause roleaux formation, which was a feature of the early, very high MW dextrans. They can  impair  renal  function  by tubular  obstruction from dextran casts. This is usually seen with dextran 40 combined  with  hypovolaemia and  pre-existing renal dysfunction. They can also cause an osmotic diuresis. Severe anaphylactic reactions like immune complex type III can occur resulting from prior cross- immunization against bacterial antigens forming dextran reactive antibodies. The incidence of 1/4500 is reduced  with monovalent  hapten  pre-treatment (injection of 3 g dextran 1) to 1/84000. This blocks the antigen-binding  sites of circulating antidextran antibodies, preventing formation of immune complexes with subsequent infusions of dextran 40 or 70. Dextran 1 (MW 1000 Da) is not available in the UK [48].

Hydroxyethyl starch(HES)

 HES is a semi-synthetic colloid, related to glycogen and was used extensively to treat wounded soldiers during  the  Vietnam  War  (1959-1975) [71].  It  is prepared   from   amylopectin,  a   highly  branched polymer  of  glucose,  derived  from   either  waxy- maize or  potato  starch  [71], which are  etherified with hydroxyethyl groups into the D-glucose units. HES have a much  lower viscosity than  dextran  or gelatin, but do not reach the low viscosity of albumin. The mean  molecular weight of the  different HES preparations ranges from 70 and 670 kDa. Following infusion of HES, small molecules <60kDa are filtrated into the urine, whereas larger molecules are degraded by plasma amylase.

The kinetics of this degradation are mainly determined  by the molar substitution  and the C2/ C6 ratio representing the quotient  of the numbers of glucose residues hydroxyethylated at positions 2 and 6, respectively) [71].  A high molar substitution and a high C2/C6 ratio make the HES molecule less susceptible to  plasma  amylase, and  thus  increase its intravascular half-life.  Part of the HES is stored within  the reticulo-endothelial  system and  slowly degraded to CO2 and water [71].

However, massive infusion of old, high-molecular-weight preparations with a high degree of substitution, particularly heta- and hexastarch, may be  associated  with excessive tissue  storage. With modern  preparations  such as 6% HES 130/0.4, no plasma accumulation and greatly reduced tissue storage have been  reported  in  the literature  [71]. The reduction in viscosity of HES solutions results from  the  globular structure   associated  with  the high degree of branching  [71]. They are classified as shown in Table 1. Different preparations of HES are hydrolysed to smaller molecules by amylase and renal elimination is rapid for polymers over 50 kDa. The action of amylase is suppressed by higher degrees of substitution and with greater etherification at the C2 versus C6 position. Intravascular half-life is thus maximized especially when the initial MW is high. In addition to the persistent plasma expansion, HES plug capillary leaks induced by sepsis and major trauma and restore macrophage function after hemorrhagic shock. Compared with 20% albumin in these patients, 10% HES significantly improves hemodynamic parameters in the systemic and microcirculation (splanchnic perfusion) [48].

Table 1.

Classification of hydroxyethyl starch preparations

MWw (kDa)	High (450–480)Medium (130–200) Low (40–70)
Degree of substitution	High (0.6–0.7) Low (0.4–0.5)
C2:C6 ratio	High > 8 Low < 8
Concentration	High 10% Low 6%

Fluid Therapy and Traumatic Brain Injury

 Clinically acceptable fluid restriction has little effect on edema formation. The first human study on fluid therapy demonstrated  that  reduction of 50% in the standard’ maintenance volume in neurosurgical patients (2.000 mL/day of 0.45 normal saline in 5% dextrose)  increases  serum  osmolality  over  about a week [72]. Thus the old concept of benefit from fluid  restriction  was simply a  consequence  of  an increased osmotic gradient over time [50]. The available data indicate that volume replacement and expansion  will have no  effect on  cerebral  edema as long as normal  serum osmolality is maintained, and cerebral hydrostatic pressures are not markedly increased due to true volume overload and elevated right heart pressures. Whether this is achieved with crystalloid  or  colloid  seems  uncertain,  although the  osmolality of  the  selected fluid  is crucial. As previously mentioned,  lactated Ringer’s solution  is not  strictly iso-osmotic (measured osmolality 252- 255 mOsmol/kg), particularly when administered to patients whose baseline osmolality has been increased by hyperosmolar fluids (mannitol, HS) [50].

In TBI, a blunt or penetrating injury incites mechanical and autodigestive destruction of the normally tightly intact endothelium of the blood brain barrier [73]. This allows uncontrolled movement of fluid and serum proteins into the brain parenchyma, eventually leading to vasogenic cerebral edema and increased ICP. It has been shown that in critically ill patients, there is increased leakage of albumin across the capillary wall [74]. In the brain, this increased extravasation of albumin  could lead to heightened interstitial oncotic pressure and exacerbate cerebral edema.

Pre-hospital

 In nine randomised controlled trials and one cohort study of pre-hospital fluid treatment in patients with TBI [75], hypertonic crystalloids and colloid solutions were not  more  effective than  isotonic  saline [76]. In a combined  polytrauma  model of uncontrolled haemorrhage  and  TBI in  swine, Teranishi  et  al., [6]  investigated  if  pre-hospital  administration   of the  haemoglobin  based oxygen carrier HBOC-201 will improve tissue oxygenation and physiologic parameters  compared  to  LR solution. They found that mean TBI force (2.4±0.1 atm; means ± standard error  of the mean)  and  blood loss (22.5±1.7 mL/ kg) were similar between groups. Survival at the 6h endpoint was similar in all groups (≈50%). Cerebral perfusion  pressure (CPP)  and  brain  tissue oxygen tension were significantly greater in HBOC-201 as compared with LR animals (p<0.005). Mean arterial pressure (MAP) and mean pulmonary artery pressure (MPAP)  were  not  significantly different  amongst groups. Blood transfusion requirements were delayed in HBOC-201 animals. Animals treated with HBOC-201  or  LR  showed  no  immunohistopathological differences in glial fibrillary acidic protein  (GFAP) and microtubule-associated protein 2 (MAP-2). Severity of subarachnoid  and intraparenchymal haemorrhages were similar for HBOC and LR groups. They concluded from their polytrauma swine model of uncontrolled haemorrhage and TBI with a 30-min delay to  hospital  arrival, pre-hospital  resuscitation of  patients  by one  bolus of  HBOC-201 indicated short term benefits in systemic and cerebrovascular physiological  parameters.   True   clinical   benefits of this strategy need to be confirmed on TBI and haemorrhagic shock patients.

In-hospital

 A defined strategy for volume replacement and fluid balance that includes maintenance of normovolemia and  colloid osmotic pressure in combination  with a neutral  to  a slightly negative fluid balance is a cornerstone of the intracranial pressure (ICP)- targeted therapy for severe TBI [77]. In contrast to hemorrhage  and  hemorrhagic  shock,  possibilities for  life-saving  interventions   are  very  limited  in CNS injury. The significant contribution  of HS to brain injury mortality further illustrates the role of hemorrhage control in reducing mortality in trauma patients [78].

Crystalloid resuscitation should be targeting a corridor  of safety, avoiding both  extremes of overt hypovolaemia and  fluid overload. While avoidance of edema formation is a prime objective and concern in visceral surgery, efforts to restrict fluids, such as ‘forced hypovolaemia’, are associated with oliguria and  occasionally renal shutdown, and  may impair nutritional  microvascular blood flow in other vascular  beds  such  as  the  splanchnic  circulation. Fluid  excess, on  the  other  hand,  is presumably  a cause  of  perioperative  morbidity   and   mortality [79]. Sequelae of volume overload are particularly well known,  and  the  pathophysiological  cascades of events have been worked out best for the patient with aggressive crystalloid resuscitation after major trauma: Manifestations of crystalloid overload might include ARDS and brain edema in the patient with concomitant head injury [80-84].

Wahlström  et al., [77] analyzed the occurrence of organ failure and mortality in patients with severe TBI treated by a protocol that includes defined strategies for fluid therapy including albumin administration to  maintain  normal  colloid osmotic  pressure  and advocating a neutral to slightly negative fluid balance. Studies conducted on 93 patients with severe TBI and Glasgow Coma  Scale (GCS) ≤8 during  1998-2001 retrieved the medical records of patients in the first 10 days. Organ dysfunction was evaluated with the Sequential Organ Failure Assessment (SOFA) score. Mortality was assessed after 10 and 28 days, 6 and 18 months. They found that the total fluid balance was positive on days 1-3, and negative on days 4-10, and  the crystalloid balance was negative from  day 2. The mean serum albumin was 38±6 g/L. Colloids constituted 40-60% of the total fluids given per day. Furosemide was administered to 94% of all patients. Severe organ failure defined as SOFA≥3 was evident only for respiratory failure, which was observed in 29% with none developing renal failure. After 28 days, mortality was 11% and, after 18 months, it was 14%. Thus, a protocol including albumin administration combined with a neutral to a slightly negative fluid balance was associated with low mortality in patients with severe TBI despite a relatively high frequency (29%) of respiratory failure, assessed by the SOFA. Acute lung injury (ALI) and  ARDS are reported commonly   after  TBI  and   their   appearance   is associated with fluid management. ALI and ARDS are considered as independent factors for mortality [85- 88].

A single equimolar  infusion of 7.45% hypertonic saline solution  is as effective as 20% mannitol  in decreasing ICP in patients  with brain  injury [79]. In the Taiwan guidelines for TBI management, with needed massive fluid transfusion, it is recommended that  normal  saline is better  than  lactated Ringer's solution (grade D). Fresh frozen plasma is only indicated for coagulopathy and not used as a regular volume expander (grade C). Hypertonic saline may be useful in patients with complication of severe TBI and systemic shock (grade D) [89]. The Saline versus Albumin Fluid Evaluation (SAFE) study  was an  international  trial  that  randomized critically  ill  patients   to   either   4%  albumin   or normal  saline fluid resuscitation for 28 days [89]. Although there was no overall difference in 28-day mortality between the 2 groups, there was a trend toward increased mortality in patients with trauma randomized to albumin resuscitation. This increased mortality  appeared  to  be driven  by patients  with trauma with TBI compared with those with trauma without  TBI. A post hoc analysis of patients  with TBI randomized during the SAFE study confirmed that resuscitation with albumin was associated with increased mortality  at  24 months  compared  with normal saline [90-92]. This increased risk was entirely driven by patients with severe TBI, defined as GCS ≤8. Sekhon et al., [91] in their study on 171 patients attemted  to  determine  if there  was an  association between synthetic colloids and mortality in patients with severe TBI,and   found  that  patients receiving pentastarch had higher acute physiology and chronic health II scores (23.9 vs 21.6, p<0.01), frequency of craniotomy (42.5% vs 21.6%, p=0.02), longer duration of intensive care unit stay (12 vs 4 days, p<0.01), and mechanical ventilation (10 vs 3 days, p<0.01). On unadjusted  Cox regression, patients  in the  highest quintile  of  cumulative  pentastarch  administration had a higher rate of mortality compared with those receiving no colloid (hazard ratio, 3.8; 95% confidence interval, 1.2-12.4; p=0.03). However, this relationship did not persist in the final multivariable model (hazard ratio 1.0; 95% confidence interval, 0.25-4.1; p= 0.98). They concluded that there was no association between cumulative exposure to pentastarch and mortality in patients with severe TBI. Elliot et al., [51] in a study focused on the hypothesis that hypertonic saline-induced improvements in histological outcome are time dependent and may be associated with alterations in astrocyte hypertrophy after cortical contusion injury. They examined histopathological changes at 7 days after controlled cortical  impact  (CCI)  injury  in  a  rat  model  and found   that   hypertonic  saline  treatment   reduced tissue loss. This correlated with attenuated astrocyte hypertrophy characterized by reductions in astrocyte immunoreactivity  without  changes in  the  number of astrocytes after CCI injury. Delayed treatment  of hypertonic saline resulted in the greatest reduction in tissue loss compared to all other treatments [0.9% normal  saline (NS; n=12),  7.5% hypertonic  saline (HS; n=15), delayed NS (n=3), delayed HS (n=4), or no treatment  (CCI control; n=18)]  indicating that there was a therapeutic window for hypertonic saline use after TBI.

Hypertonic/hyperoncotic solutions

 Recently, attention has been directed at hypertonic/ hyperoncotic solutions typical of hypertonic hetastarch  or  dextran  solutions.  Because of  the haemodynamic stabilizing properties of these fluids in hypovolaemic shock, their administration  in patients with trauma and TBI might be particularly advantageous for the prevention of secondary ischaemic brain damage. Small volumes of such solutions can rapidly restore normovolaemia without increasing ICP. They have been successfully used to treat intracranial hypertension in TBI patients [66], and in other neurosurgical acute emergencies [SAH [92] and stroke [93]].

Fluid Therapy for Decompressive Craniectomy

Some  general  principles  of  enhanced   recovery associated     with          fluid     management     and recommendations    for    the    enhanced    recovery partnership are as follows [94]:

Pre-operative:

• Maintain good pre-operative hydration.

• Give carbohydrate drinks.

• Avoid bowel preparation.

Peri-operative

• Use fluid management  technologies to  deliver individualized goal directed fluid therapy.

• Avoid crystalloid excess (salt and water overload). Maintenance’ fluid, if utilized, should be limited to  less than  2  ml/kg/hr  including  any  drug infusions. The use of isotonic balanced electrolyte such as Hartmann’s solution will minimize hyperchloraemic acidosis.

Post-operative:

• Avoid post-operative i.v. fluids when it is possible.

• Always ask the  question; ‘what are we giving fluids for? Maintenance fluid? -Push early drinking and eating;

• Replacement fluid? ; Considering oral before i.v. and prescribing oral fluids Resuscitation fluids? ; Using goal directed fluid therapy

Physiological responses during the perioperative phase

In the critically ill, effects of surgery per se and its associated changes in the hormonal milieu interne are exaggerated by a systemic inflammatory response with development of capillary leak. This leads to difficult- to-balance losses into the interstitium and frequently visible oedema  formation.  Resulting abnormalities of fluid and electrolyte balance in the critically ill are purposefully or involuntarily influenced, in addition, by nutritional support and measures that affect acid– base homeostasis [79].

Surgery alters fluid balance [39], generates a systemic inflammatory response which increases oxygen consumption, and is associated with increase in cardiac output and oxygen delivery. Failure to meet the metabolic demands of recovery from surgery is associated with increased morbidity  and  mortality [95].  The  stress response  to  surgery  and  trauma involves a number of different physiological reactions. Importantly, the renine-angiotensine-aldosterone system is stimulated, leading to increased sodium and fluid retention, decreased urinary output and altered fluid balance. In addition, the activated inflammatory response causes vasodilatation and increased permeability of capillary wall [47]. This affects the intravascular duration of fluids administration, with increased capillary leak of fluids into the interstitial tissues. As a result, the perioperative period is a time when the body’s management of fluids is dramatically altered and needs to be considered carefully when prescribing fluids [47]. In conclusion, perioperative fluid therapy continues to be an exercise in empiricism, with nagging questions about its efficacy and complications. There are  no  evidence-based guidelines or  standards  of care for the management of fluid therapy in patients undergoing decompressive craniectomy. Knowledge of the properties of the various available IV fluids, and the awareness of the pathophysiology of endothelial, parenchymal and endocrine alterations emerging in TBI should guide i.v. fluid administration, to reach a good medium that favors better neurological, morbidity and mortality outcomes.

Conflict of Interest: None declared.