Intensive Care Med (2013) 39:14131422 DOI 10.

1007/s00134-013-2978-9

ORIGINAL

Carole Ichai Jean-Franc ois Payen Jean-Christophe Orban Quintard Herve Hubert Roth Robin Legrand Gilles Francony Xavier M. Leverve

Half-molar sodium lactate infusion to prevent intracranial hypertensive episodes in severe traumatic brain injured patients: a randomized controlled trial

Received: 20 January 2013 Accepted: 22 May 2013 Published online: 8 June 2013 Springer-Verlag Berlin Heidelberg and ESICM 2013 X. M. Leverve is deceased. Electronic supplementary material The online version of this article (doi:10.1007/s00134-013-2978-9) contains supplementary material, which is available to authorized users.

H. Roth X. M. Leverve INSERM U1055, Laboratoire de Biologie e, Po le Fondamentale et Applique Recherche, Albert Michallon Hospital, Joseph Fourier University, 38042 Grenoble, France

C. Ichai ()) J.-C. Orban H. Quintard Medicosurgical Intensive Care Unit, Saint Roch Hospital, Nice Sophia-Antipolis University, 06000 Nice, France e-mail: ichai@unice.fr Tel.: ?33-4-92033627 Fax: ?33-4-92033558 J.-F. Payen R. Legrand G. Francony Department of Anesthesiology and Intensive Care, Albert Michallon Hospital, 38043 Grenoble, France J.-F. Payen G. Francony INSERM U836, Grenoble Institut des Neurosciences, Joseph Fourier University, 38042 Grenoble, France

Abstract Purpose: Preventive treatments of traumatic intracranial hypertension are not yet established. We aimed to compare the efciency of half-molar sodium lactate (SL) versus saline serum solutions in preventing episodes of raised intracranial pressure (ICP) in patients with severe traumatic brain injury (TBI). Methods: This was a double-blind, randomized controlled trial including 60 patients with severe TBI requiring ICP monitoring. Patients were randomly allocated to receive a 48-h continuous infusion at 0.5 ml/kg/h of either SL (SL group) or isotonic saline solution (control group) within the rst 12 h post-trauma. Serial measurements of ICP, as well as uid, sodium, and chloride balance were performed over the 48-h study period. The primary outcome was the number

of raised ICP (C20 mmHg) requiring a specic treatment. Results: Raised ICP episodes were reduced in the SL group as compared to the control group within the 48-h study period: 23 versus 53 episodes, respectively (p \ 0.05). The proportion of patients presenting raised ICP episodes was smaller in the SL group than in the saline group: 11 (36 %) versus 20 patients (66 %) (p \ 0.05). Cumulative 48-h uid and chloride balances were reduced in the SL group compared to the control group (both p \ 0.01). Conclusion: A 48-h infusion of SL decreased the occurrence of raised ICP episodes in patients with severe TBI, while reducing uid and chloride balances. These ndings suggest that SL solution could be considered as an alternative treatment to prevent raised ICP following severe TBI. Keywords Intracranial pressure Intracranial hypertension Lactate solution Osmotherapy Traumatic brain injury

Introduction
More than 2,000 head injuries per one million inhabitants occur annually in the USA and Europe, and when severely traumatic, may result in up to 50 % of patients experiencing unfavorable outcomes [1]. Prolonged or repetitive brief episodes of raised intracranial pressure

(ICP) have been shown to play a major role in the worsening of neurological outcome, and thus most efforts to treat these injuries focus on controlling the ICP and maintaining optimal cerebral perfusion pressure (CPP) [2, 3]. Among therapies to treat episodes of elevated ICP, mannitol and hypertonic saline solutions are used to create an osmotic gradient across plasma and brain

1414

compartments to facilitate water extrusion from brain cells [4, 5]. However, each solution has a time-limited efcacy and side effects [6, 7], and their preventive action against the development of subsequent elevated ICP episodes has not been established. In our search for alternatives to mannitol and hypertonic saline solutions, we previously found that bolus infusion of half-molar sodium lactate (SL) was more effective at reducing elevated ICP than an equimolar bolus of mannitol in patients with severe traumatic brain injury (TBI) [8]. The rationale for using such a solution in an injured brain was based on accumulating evidence of the use of lactate as one major oxidative substrate by injured brain cells [911]. Exogenous lactate anions can enter into injured brain cells to be utilized as an energy source [1214]. Infusion of hyperosmotic sodium lactate for 24 h signicantly lowered ICP and improved cerebral blood ow in a model of focal TBI [15], in line with the role of lactate to induce cerebral vasodilatation when oxygen availability is reduced [16]. In addition, the administration of SL could restore osmotic and charge balance across the cell membrane because exogenous, non-metabolized sodium remains in the extracellular compartment, leading to the extrusion from cells of a nonmetabolized anion, i.e., chloride, accompanied by the extrusion of water from cells and, consequentially, a reduced cell volume. Whatever the mechanisms by which it may impact ICP, whether SL can exert a preventive action against the subsequent development of raised ICP after severe TBI is unknown. Our aim was therefore to investigate the effectiveness of 48-h infusion of SL at decreasing the number of raised ICP episodes in severe brain-injured patients with the concomitant determination of uid, sodium and chloride balance.

measurement of ICP as part of their management within the rst 12 h following injury. Whatever the conditions at the scene or in the ambulance, patients had to be in stable or stabilized conditions to be considered eligible for the study. Patients were included if they fullled the following criteria: mechanical ventilation under continuous sedation with initial PaO2 over 80 mmHg and PaCO2 between 35 and 50 mmHg, CPP over 60 mmHg, measured serum osmolality ranging between 280 and 320 mOsm/kg, and body temperature between 36 and 38 C. Patients were not included if they met any of the following criteria: pregnant, cardiac arrest at initial presentation, GCS of three with bilateral xed dilated pupils, imminent cranial or extracranial surgery, decompressive craniectomy prior to enrolment, a leak or drainage of cerebral spinal uid, evidence of severe respiratory, renal, liver or cardiac dysfunction, hemoglobin content below 10 g/dL, persistent hemodynamic or respiratory instability despite treatments, i.e., mean arterial pressure \70 mmHg, PaO2 \80 mmHg, PaCO2 \30 mmHg or [50 mmHg, measured serum osmolality over 320 mOsm/kg, serum sodium concentration over 155 mmol/ L, or use in the 6 h prior to enrolment of mannitol, hypertonic saline solution, or barbiturates. Study protocol Patients were randomly assigned in a 1:1 ratio to receive either half-molar sodium lactate (SL group) (TotilacTM, Innogene Kalbiotech, Singapore, Malaysia) or 0.9 % saline solution (control group). Treatment allocation was stratied by center and randomization was assured using a computerized random-number generator list provided by a statistician not involved in patient recruitment or outcome assessment. ICU physicians and nurses were blinded to the assigned treatments, indiscernible in appearance and undisclosed in verbal or written reports. SL solution contained Na 504 mmol/L, K 4 mmol/L, Ca 1.36 mmol/L, Cl 6.74 mmol/L, and L-lactate 504.1 mmol/L. SL and saline solutions were intravenously administered at a rate of 0.5 ml/kg/h during 48 h. Management of patients with severe TBI was consistent with the guidelines of the Brain Trauma Foundation [3]. Both ICUs had similar patterns of management and were run by intensivists 24/7. Treatments during the study period were provided to prevent systemic secondary insults (ESM content 1). Collected variables included the patients demographic and baseline characteristics. Each patient had a CT scan on admission to classify brain lesions according to the Marshalls classication [17]. CT scan was systematically repeated on day 1 after injury and later again if needed. The ICP was continuously monitored with an intraparenchymal device (Codman Microsensors ICP transducers, Codman France, Johnson & Johnson,

Materials and methods

Patient population This prospective, randomized, controlled, double-blinded (caregiver, investigator) trial was conducted between January 2010 and July 2011 at the University Hospitals of Nice and Grenoble, and included two intensive care units (ICU). The study design was approved by the local diInstitutional Ethics Committee (08-050 CPP Sud-Me e V). Written informed consent was obtained terranne from the patients relatives before inclusion wherever possible, or from a legally authorized representative and subsequently from the patients in accordance with French Legal Ethics. Patients aged between 18 and 65 years were prospectively enrolled in the study if they were admitted following a severe non-penetrating TBI with an initial Glasgow Coma Scale (GCS) score of \9, and required

1415

The study population size was calculated based on the knowledge that half of patients with severe TBI have at least one episode of raised ICP during the initial 3 days post-injury [20]. From our previous results with SL bolus used to treat elevated ICP [8], we hypothesized a 50 % reduction in raised ICP episodes requiring specic treatment in the SL group during the 48 h following the initiation of infusion compared with the control group. For a statistical power of 80 % and a two-sided type error rate of 0.05, 24 patients per group were required. A sample size of 30 patients per group allowed for withdrawal and loss to follow-up. Descriptive statistics included frequencies and percentages for categorical variables, and means (SD) or medians (25th:75th interquartiles) for continuous variables. Comparisons between groups were performed using the Chi square test, Students t test or MannWhitney test where appropriate (Statview 5.0, SAS Institute, Cary, NC). A two-way ANOVA for repeated measures was used to analyze the interaction between time and group, followed by Scheffe F tests (intra- and intergroup comparisons) as Endpoints post hoc analysis. Statistical signicance was declared The primary study outcome was the number of raised ICP when P \ 0.05. The study is registered with ClinicalTrials.gov, numepisodes within the 48-h study period. A raised ICP episode was dened as an elevation of ICP of more than 20 mmHg ber NCT 00995683. for more than 10 min, that was unrelated to procedural pain and required specic treatment [5, 18]. Treatment for raised ICP episodes included rst-tier therapies, i.e., mannitol, cerebrospinal uid drainage, deep sedation, increasing Results CPP, mild hyperventilation (to PaCO2 30 mmHg), and muscle relaxants, and second-tier therapies included the Of the 101 eligible patients, 60 were consecutively and use of barbiturates, moderate hypothermia (to 33 C), or randomly assigned for treatment and analyzed (30 decompressive craniectomy. The choice of treatment patients per group). Reasons for exclusion before ranwithin rst- and second-tier therapies was left to the dis- domization are given in Fig. 1. All included patients cretion of the physician in charge. Physicians were completed the entire period of the protocol. None were permitted to prescribe simultaneous therapeutic interven- excluded for loss during the 6-month neurological GOS tions for one ICP episode if needed. Decompressive follow-up. The two groups were comparable at baseline craniectomy was indicated if medical therapies failed. (Table 1). There were six patients in the SL group and Secondary study outcomes included the AUC-ICP and ve patients in the control group who developed diabetes AUC-CPP, as well as uid, sodium and chloride balance insipidus during the study period. All of these had serum during the 48-h study period. Effective strong ion dif- sodium and serum osmolality measurements that were ference (SID) was calculated as the sum of serum kept within the safety margins during the study period. bicarbonates, albuminate and phosphate (ESM content 2) [19]. Other secondary endpoints were the number of raised ICP episodes, the nature of treatment administered Brain parameters throughout ICP monitoring, and neurological outcome at 6 months. Neurological outcome was assessed via phone Raised ICP episodes were signicantly reduced in the SL interview by a physician blinded to treatment group using group compared to the control during the 48-h study the Glasgow Outcome Scale (GOS) score at 6 months period: 23 versus 53 episodes, corresponding to 0 (0:1) after trauma. Patients were dichotomized for analysis into versus 1 (0:3) episodes per patient, respectively those with good outcome (good recovery ? moderate (P \ 0.05) (Fig. 2 and ESM Table 1). The number of

Issy-les-Moulineaux, France). Hourly recordings of ICP and CPP during the whole period of ICP monitoring allowed the calculation of the area under the curve (AUC) of both variables (AUC-ICP, AUC-CPP) using the trapezoidal rule. Blood ow velocity (FV) in each of the two middle cerebral arteries was measured on admission and, if possible, during the study period via transcranial doppler (TCD) (General Electric Vivid S5, Velizy, France; Looki 1-TC, Atys Medical, Soucieu en Jarrest, France). Systolic, diastolic and mean blood ow velocities in the middle cerebral arteries (respectively FVs, FVd, FVm) were measured to calculate the pulsatility index (PI), using PI = (FVs - FVd)/FVm. TCD recordings were taken from the most injured hemisphere. Other variables included arterial serum and urinary electrolyte concentrations and measured osmolalities. Variables were recorded before randomization and at 24 and 48 h after randomization. Fluid, sodium and chloride balance was determined at 48 h by calculating for each individual the difference between the cumulative amounts of uids and electrolytes administered throughout the 48 h-study period and the cumulative amounts of uids and electrolytes collected in the urine.

disability) and poor outcome (severe disability ? vegetative state ? death). Statistical analysis

1416

Fig. 1 Patient ow diagram showing the number of identied, excluded and analyzed patients

Eligible patients : traumatic brain injury (TBI) with a glasgow coma scale < 9 n = 101 Non included TBI n = 41
3 consent refusals 2 minors 7 osmotherapy prior hospital admission 6 neurosurgical intervention at admission 3 spontaneous CSF evacuation 4 delayed admission 7 GCS 3 with bilateral fixed dilated pupils 9 polytrauma with systemic secondary insults

Randomized TBI n = 60

Randomly assigned to NaCl 0.9% n = 30

Randomly assigned to half-molar sodium lactate n = 30

patients presenting at least one raised ICP episode was smaller in the SL group: 11 (36 %) versus 20 patients (66 %) (P \ 0.05). Although there was no signicant difference in the nature of intervention to treat elevated ICP episodes, the number of interventions was signicantly smaller in the SL group: 31 versus 53 interventions during the 48-h study period (P \ 0.05). It should be noted that two or three simultaneous therapeutic interventions were required to treat eight ICP episodes for four patients in the SL group (ESM Table 1). The groups showed no difference in the highest value of ICP recorded during the 48-h study period: 25 (19:32) (SL group) versus 28 mmHg (23:41) (control group). The median values of AUC-ICP and AUC-CPP during the study period were comparable between the two groups (ESM Fig. 1a and 1b). The number of raised ICP episodes throughout the entire period of ICP monitoring was also signicantly higher in the SL group 27 versus 80 episodes (P \ 0.05). Nine and 11 patients in the control and SL groups, respectively, had TCD measurements during the study period. No major differences were found between the two groups (ESM content 3). Other data obtained during the 48-h infusion period are shown in ESM content 3. No rebound ICP was found in either of the two groups after the cessation of protocol infusion. Six and ve patients in the control and SL groups, respectively, were given barbiturates. All had more ICP episodes, higher AUC-ICP during the study period, and required more boluses of mannitol and decompressive

craniectomy than the 49 other patients (all P \ 0.01). Their neurological outcome at 6 months was less favorable (P \ 0.05). No effect from the type of study infusion was found in this subgroup of patients. Biological parameters Biological parameters were comparable at baseline between the two groups (Table 2). There was a signicant interaction between group and serial measurements of serum sodium concentration (F test = 6.9; P \ 0.01). This was due to a higher value at H48 in the SL group compared to control (Table 2). Both treatment conditions were responsible for a signicant increase in serum chloride concentration, with no differences between the two groups. Serum glucose, albumin and haemoglobin concentrations decreased over time and comparably between the two groups. Of note were the unchanged measurements of serum and urine osmolalities throughout the study period in both groups. Despite lactate infusion in the SL group, both groups showed a gradual decrease in serum lactate concentration during the 48-h study period. Arterial pH increased over time in both groups, however, this effect was more pronounced in the SL group (F test = 10.4; P \ 0.01) and was associated with a more pronounced increase in serum bicarbonate concentration (F test = 16.3; P \ 0.01).

1417

Table 1 Demographic and baseline characteristics from the 60 patients before treatment initiation
6

p<005 Number of treated raised intracrnial pressure episodess (48 first hours) Control group SL group 1 3 2 3 3 3 20
20 10 0

Characteristics

Control group (n = 30) 33 15 26/4 72 12 6 (57) 16 (1132) 65 (5474) 1 (3) 8 (27) 21 (70) 59 (4573) 36 (2944) 26 (1929) 1.2 0.3 88 13 90 20 36.5 1 140 81 323 93 4.9 2.4 15.7 8.5 6.5 3.7

SL group (n = 30) 40 18 28/2 78 13 6 (48) 15 (1224) 66 (5173) 1 (3) 10 (34) 19 (63) 65 (5083) 39 (3146) 23 (1931) 1.3 0.5 83 10 85 19 36.5 0.9 170 115 376 233

Fig. 2 Raised intracranial pressure episodes requiring either no 5.8 5.3 intervention or one intervention or more within the rst 48 h after 17.2 11.6 severe traumatic brain injury. The total number of raised intracra6.2 0.9 nial pressure episodes in the half-molar sodium lactate group was signicantly lower than in the control group (23 versus 53 episodes, P \ 005). Control = 09 % saline solution. SL = half-molar Data are mean SD, median [2575th interquartile range] or sodium lactate number (percentage). There were no signicant differences between the two groups at baseline GCS Glasgow Coma Scale, ICP intracranial pressure, CPP cerebral perfusion pressure, TCDB Trauma Coma Data Bank, FVs, FVm and group during the study period: 1,332 (1,096:1,466) versus FVd time-averaged systolic, mean and diastolic values of blood 712 mmol (539:904) (P \ 0.01), and 711 (577:916) verow velocities, respectively, PI pulsatility index, ICU intensive sus 332 mmol (105:457), respectively (P \ 0.01). This care unit did, however, lead to a comparable sodium balance at a 19 missing values (nine in control group, ten in SL group)

Age (years) Sex (M/F) Weight (kg) Median initial GCS score Initial ICP (mmHg) Initial CPP (mmHg) TCDB classication (n %): Diffuse I Diffuse II Diffuse IIIIV Transcranial Doppler on admissiona FVs (cm/s) FVm (cm/s) FVd (cm/s) PI Mean arterial pressure on admission (mmHg) Heart rate on admission (min-1) Temperature on admission (C) Injury to admission period (min) Admission to ICP monitoring period (min) Duration of ICP monitoring (days) Length of stay in ICU (days) Duration of mechanical ventilation (days)

4 3 6 11
10

Number of patients

Cumulative uid, sodium and chloride balance The same volumes of crystalloids were administered during the study period between the two groups: 24 mL/ kg (24:25). However, the SL group had lower requirements for additional saline solution than the control group to maintain CPP within normal limits: 13 (6:21) versus 26 mL/kg (18:33), respectively (P \ 0.05). Comparable volumes of colloids were given to the two groups during the study period: 6 (0:15) (SL group) versus 13 mL/kg (5:24) (control group). The hourly urine output over the 48-h study period was comparable between the two groups: 1.1 (0.9:1.4) (SL group) versus 1.0 (0.7:1.3) mL/ kg/h (control group). Therefore, the cumulative uid balance at 48 h was signicantly smaller in the SL group: 5 (-5:19) versus 27 mL/kg (11:43) (P \ 0.01) (Fig. 3a). Both cumulative sodium intake and output were signicantly higher in the SL compared to in the control

48 h between the two groups of patients (ESM Fig 2). On the other hand, cumulative chloride intake was signicantly smaller in the SL than in the saline solution at 48 h: 295 (143:421) versus 700 mmol (519:845) (P \ 0.01). Since chloride urine output was comparable between the two groups [295 (210:446) versus. 390 mmol (239:553)] (P = 0.28), cumulative chloride balance was markedly reduced in the SL group compared to the control group (P \ 0.01) (Fig. 3b).

Neurological outcome Neurological outcome at 6 months showed no signicant difference between the two groups: 12 patients (SL group) versus 15 patients (control group) had a poor outcome at 6 months. However, more survivors from the control group had poorer neurological outcome compared to the SL group: nine versus four patients, although this difference did not reach signicance (P = 0.14).

1418

Table 2 Course of biological Variables parameters during the 48-h study protocol according to Serum sodium (mmol/L) treatment group: patients Control group receiving saline solution SL group (control) versus those receiving half-molar sodium lactate (SL) Serum potassium (mmol/L) Control group SL group Serum chloride (mmol/L) Control group SL group Serum glucose (mmol/L) Control group SL group Serum albumin (g/L) Control group SL group Serum creatinine (mg/dL) Control group SL group Haemoglobin (g/dL) Control group SL group Serum lactate (mmol/L) Control group SL group Serum Osmolality (mOsm/kg) Control group SL group Arterial pH Control group SL group PaCO2 (mmHg) Control group SL group PaO2 (mmHg) Control group SL group HCO3- (mmol/L) Control group SL group SID (mEq/L) Control group SL group Urine sodium (mmol/L) Control group SL group Urine chloride (mmol/L) Control group SL group Urine osmolality (mOsm/kg) Control group SL group

Baseline 141 4 142 3 3.8 0.5 3.8 0.5 107 2 107 4 7.4 2.2 8.3 2.1 40 5 39 5 0.7 0.2 0.7 0.1 12.7 1.7 13 1.8 1.9 1.2 2.3 1.3 301 10 301 10 7.36 0.07 7.36 0.07 40 7 41 6 204 64 193 71 22 2 22 2 35 2 35 2 95 48 94 52 112 48 104 65 630 192 630 251

24 h 144 5** 149 6** 4.2 0.1 3.5 0.1 112 5** 110 5* 6.5 1.4** 7.2 2.2** 32 5** 32 4** 0.7 0.1 0.7 0.1 11.4 1.6** 12 1.7** 1.5 1.2 1.9 0.7 299 10 305 11 7.39 0.05 7.48 0.07**, 36 5* 36 7** 137 38** 130 44** 21 2 27 4**, 32 3** 38 4**,

48 h 143 4# 147 5**, 4 0.4 3.5 0.1 111 5** 111 3** 6.6 1.3** 6.6 1.6** 29 3**, 30 4**, 0.6 0.2 0.7 0.1 10.5 1.6**, 11.2 1.8**, 1 0.5**, 1.5 0.5**, 298 11 304 10 7.42 0.03**, 7.49 0.05**, 35 3** 34 6** 120 36**, 117 37** 23 2## 27 4**, 33 2** 38 4**,
# ## ## ## ## ##, ## ##

80 41 149 58**, 109 61 82 44 660 242 641 169

96 60* 213 67**, 139 76 80 41 741 192 676 183

##,

Data are mean SD. SID, strong ion difference (see text for calculation) P \ 0.05 and P \ 0.01 versus Control group (intergroup analysis); * P \ 0.05 and ** P \ 0.01 versus baseline (intragroup analysis); # P \ 0.05 and ## P \ 0.01 48 h versus 24 h (intragroup analysis)

Discussion
A 48-h SL infusion decreased by more than 50 % the occurrence of raised ICP episodes requiring treatment in patients with severe TBI. This effect was associated with reduced uid and chloride balance, and no change in plasma osmolality. Our results support previous

experimental and clinical ndings in favor of the use of SL in an injured brain, and suggest that SL might be considered during the early phase of resuscitation of patients with severe TBI. Current guidelines recommend preventing secondary brain damage through the control of ICP in order to minimize cerebral ischemia after severe TBI [2, 3].

1419

Cumulative chloride balance over the 48 hrs of the study period (mmol)

Cumulative fluid balance over the 48 hrs of the study period (mL/kg)

Fig. 3 Cumulative uid (a), and chloride (b) balances (median, 25th and 75th interquartiles) over the 48-h study period. Control = 09 % saline solution. SL = halfmolar sodium lactate. p \ 001 SL versus Control

A
50 40 30

700 600 500 400 300 200 100 0 -100

20 10 0 -10 -20

Control group (n = 30)

SL group (n = 30)

Control group (n = 30)

SL group (n = 30)

Prevention of subsequent elevation of ICP can be achieved by various actions such as deep sedation, maintenance of euvolemia and CPP, optimization of ventilation and electrolytes disturbances, and therapeutic hypothermia. External ventricular drainage is also effective in measuring and treating elevated ICP, however, it is sometimes impossible to insert a ventricular drain in TBI patients with compression of the ventricular system. The preventive action of osmotic agents against subsequent elevation of ICP has not been established in patients with severe TBI. Retrospectively, continuous infusion of hypertonic saline solution for several days allowed the decrease in ICP in patients with refractory intracranial hypertension, but also led to hypernatremia and hyperchloremia [21, 22]. In addition to glucose, there are other substrates such as lactate, acetate and ketone bodies to consider as alternatives in conditions of inadequate glucose availability [23]. Exogenous lactate has successfully been used for years as an alternative fuel for the brain in situations of increased energy demand. In rodent models of TBI, lactate infusion was associated with a lower reduction of cerebral ATP after the trauma [24], the preservation of extracellular cerebral glucose levels, and improved mitochondrial oxidative respiration [25] and cognitive dysfunction several days post-injury [24, 26]. More recently, the intracerebroventricular administration of lactate after ischemia was shown to decrease brain lesion size and improve neurological outcome in mice [27]. In humans, lactate infusion was associated with a signicant correction of brain dysfunction during severe hypoglycaemia [12, 28]. Moreover, the brain cells of head injured patients can utilize exogenous lactate as an energy source via the tricarboxylic acid cycle [10]. Collectively, these ndings indicate that lactate may appropriately fuel aerobic brain energy metabolism, especially during periods

of increased brain energy requirements, and may play a central role in the metabolic interaction between astrocytes and neurons [29, 30]. In one recent study, however, the ability of exogenous lactate to maintain cerebral ATP content was not conrmed following diffuse TBI in rats [31]. Because we did not measure carotidjugular differences in blood lactate or cerebral microdialysis lactate, we cannot conrm that the exogenous lactate provided by the SL solution was indeed preferentially utilized in the injured brains of our patients. In the present study, the interventional group of patients (SL group) was given hypertonic sodium in addition to lactate. In a previous study, SL solution was infused safely into patients following elective cardiac surgery and shown to decrease oxygen extraction fraction from tissues [32]. It was later shown to increase cardiac index in patients who had undergone coronary artery bypass grafting by comparison with Ringers lactate, in association with a negative uid balance [33]. In a previous randomized trial, we also found that boluses of SL were more effective in treating elevated ICP than equiosmolar doses of mannitol with no concomitant changes in plasma osmolality [8]. In line with these results, our present study has shown that continuous infusion of SL was more effective than saline solution in preventing subsequent raised ICP episodes, an effect that was concomitant with a neutral uid balance. One could argue here that SL infusion prevented any subsequent rise in ICP due to its hyperosmolar hypernatremic effects. However, despite having a similar hyperosmolality, SL solution is less hypertonic than mannitol 20 % or 3 % hypertonic saline solutions due to the transport of lactate across the cell membrane [34]. In the present study, despite the transitory increased serum sodium concentration, the lack of change in measured plasma osmolality in the SL group indicates that the effects on ICP were

1420

unlikely to have been mediated by the creation of an osmotic gradient. Due to its high sodium concentration, SL infusion increases extracellular positive charges. While lactate is rapidly metabolized in the cell, serum sodium accumulation creates an imbalance between the positive (nonorganic cations) and the negative (non-organic anions) charges. Thus, a net efux of negative charges from the cell is obligatory to restore the extracellular electrical neutrality. Chloride might be involved in this efux because it is the major intracellular inorganic anion. There is growing evidence showing that chloride is largely involved in the cell volume regulation via chloride cations cotransporters and channels [35, 36]. The efux of chloride might be associated with a cell efux of water to maintain the osmotic equilibrium. Our ndings support this assumption: the amount of urinary chloride excretion was comparable in both groups regarding to the low chloride exogenous load in the SL group. Thus, an endogenous chloride load is necessary and we may hypothesize a chloride extrusion from cells. Interestingly, the increased serum chloride in the SL group may support this hypothesis of redistribution between intra- and extracellular chloride compartments. The positive sodium balance associated with the neutral chloride balance following SL infusion supports the obligatory chloride urine excretion coupled to the high sodium urine excretion. In addition, the 48-h uid balance was markedly lowered in the SL group compared to the control group, an event likely due to reduced uid requirements during SL administration. Therefore, one possible hypothesis to explain the effects of SL on ICP could be mediated through favoring cell extrusion of chloride and water, thereby limiting cellular edema. However, in the absence of brain diffusionweighted MRI to document cell edema, this assumption remains speculative. The response to ICP during SL administration might have been mediated through other mechanisms such as improvement in brain haemodynamics [15], or an attenuation of metabolic crisis after TBI, as reected by hyperglycolysis, deprivation in brain glucose and high lactate/pyruvate ratios [11, 37, 38]. The dose of SL for infusion was determined on the basis of previous studies [8, 32, 33], and the amount of exogenous lactate infused was below the maximal metabolizing capacity of our patients. In our study, SL infusion was responsible for a mild metabolic alkalosis (see Table 2), according to the Stewart concept [19]. As mentioned above, the entering of infused lactate into the cell increases the strong ion difference between cell and plasma compartments. The plasma bicarbonate (weak anions) level in turn increases to restore electrical neutrality, which leads to metabolic alkalosis (see Table 2). While vasoconstriction of brain vessels during respiratory

alkalosis is well-known, there is a lack of data on the impact of mild metabolic alkalosis on cerebral blood ow [39, 40]. We found no aggravation of TCD parameters in the SL group. Therefore, a deleterious impact of mild metabolic alkalosis on cerebral blood ow seems unlikely in this group. Moderate hypokalemia was also noted and should be prevented by appropriate supplementation. There are several limitations with this study. First, we chose to infuse the SL solution over a 48-h period and at a predened dose. Whether a longer period or higher dose would lead to different results, in particular on neurological outcome, remains to be evaluated. Second, we did not use a brain tissue oxygen tension probe, microdialysis monitoring or cerebral blood ow measurements. Nevertheless, estimation of cerebral blood ow using TCD showed no major changes in TCD parameters during the study period. Multimodal monitoring regarding brain circulation and oxygenation during SL administration should be further investigated. Third, further studies comparing equiosmolar SL and hypertonic saline solutions are needed to evaluate the individual roles of sodium and lactate in the SL solution. Fourth, reducing the number of episodes of intracranial hypertension has previously been associated with better neurological outcome [2]. We found no signicant benets of SL solution in terms of neurological outcome among survivors at 6 months. However, this study was primarily designed to evaluate the preventive action of SL solution in patients at risk of suffering subsequent episodes of raised ICP. In conclusion, our study shows that a 48-h SL infusion allows a reduction in the number of raised ICP episodes in patients with severe TBI. This effect might be mediated through the urine extrusion of chloride and water, although other mechanisms such as changes in brain metabolism or circulation cannot be excluded.
Acknowledgments This paper is dedicated to the memory of Professor Xavier Leverve, who provided the intellectual framework for this clinical study. Regrettably, he died on November 8th, 2010. The research team behind this work is honored to have worked under his superb guidance and vitality. We thank Innogene Kalbiotech, Pte. Ltd (24 Rafes Place 27 - 06 Clifford Centre, Singapore 048621, Malaysia) for their gift of NaCl and half-molar sodium lactate solutions. The half-molar sodium lactate-containing solution is patented (WO 2004/096204 -11/11/ 04, Gazette 2004/46) and registered (TotilacTM). The trial was investigator driven. Half-molar sodium lactate solution was kindly provided by Innogene Kalbiotech, Singapore, Malaysia. The company had no role in the study design, data collection, data analysis, data interpretation or writing of the report. Conicts of interest All authors had full access to all the study data, and all agreed to submit for publication. Professor Xavier Leverve was a member of the Innogene International Scientic Board of Advisors. The other authors declare no conict of interests.

1421

References
1. Tagliaferri F, Compagnone C, Korsic M, Servadei F, Kraus J (2006) A systematic review of brain injury epidemiology in Europe. Acta Neurochir (Wien) 148:255268 2. Stein DM, Hu PF, Brenner M, Sheth KN, Liu KH, Xiong W, Aarabi B, Scalea TM (2011) Brief episodes of intracranial hypertension and cerebral hypoperfusion are associated with poor functional outcome after severe traumatic brain injury. J Trauma 71:364373 3. Bratton SL, Chestnut RM, Ghajar J, McConnell Hammond FF, Harris OA, Hartl R, Manley GT, Nemecek A, Newell DW, Rosenthal G, Schouten J, Shutter L, Timmons SD, Ullman JS, Videtta W, Wilberger JE, Wright DW (2007) Guidelines for the management of severe traumatic brain injury IX. Cerebral perfusion thresholds. J Neurotrauma 24(Suppl 1):S59S64 4. Bratton SL, Chestnut RM, Ghajar J, McConnell Hammond FF, Harris OA, Hartl R, Manley GT, Nemecek A, Newell DW, Rosenthal G, Schouten J, Shutter L, Timmons SD, Ullman JS, Videtta W, Wilberger JE, Wright DW (2007) Guidelines for the management of severe traumatic brain injury II. Hyperosmolar therapy. J Neurotrauma 24(Suppl 1):S14S20 5. Francony G, Fauvage B, Falcon D, Canet C, Dilou H, Lavagne P, Jacquot C, Payen JF (2008) Equimolar doses of mannitol and hypertonic saline in the treatment of increased intracranial pressure. Crit Care Med 36:795800 6. Paczynski RP (1997) Osmotherapy. Basic concepts and controversies. Crit Care Clin 13:105129 7. Kamel H, Navi BB, Nakagawa K, Hemphill JC 3rd, Ko NU (2011) Hypertonic saline versus mannitol for the treatment of elevated intracranial pressure: a meta-analysis of randomized clinical trials. Crit Care Med 39:554559 8. Ichai C, Armando G, Orban JC, Berthier F, Rami L, Samat-Long C, Grimaud D, Leverve X (2009) Sodium lactate versus mannitol in the treatment of intracranial hypertensive episodes in severe traumatic brain-injured patients. Intensive Care Med 35:471479 9. Schurr A, West CA, Rigor BM (1988) Lactate-supported synaptic function in the rat hippocampal slice preparation. Science 240:13261328 10. Gallagher CN, Carpenter KL, Grice P, Howe DJ, Mason A, Timofeev I, Menon DK, Kirkpatrick PJ, Pickard JD, Sutherland GR, Hutchinson PJ (2009) The human brain utilizes lactate via the tricarboxylic acid cycle: a 13C-labelled microdialysis and high-resolution nuclear magnetic resonance study. Brain 132:28392849 11. Oddo M, Levine JM, Frangos S, Maloney-Wilensky E, Carrera E, Daniel RT, Levivier M, Magistretti PJ, Leroux PD (2012) Brain lactate metabolism in humans with subarachnoid hemorrhage. Stroke 43:14181421 12. Maran A, Cranston I, Lomas J, Macdonald I, Amiel SA (1994) Protection by lactate of cerebral function during hypoglycaemia. Lancet 343:1620 13. Chen T, Qian YZ, Rice A, Zhu JP, Di X, Bullock R (2000) Brain lactate uptake increases at the site of impact after traumatic brain injury. Brain Res 861:281287 14. Meierhans R, Brandi G, Fasshauer M, Sommerfeld J, Schupbach R, Bechir M, Stover J (2012) Arterial lactate above 2 mM is associated with increased brain lactate and decreased brain glucose in patients with severe traumatic brain injury. Minerva Anestesiol 78:185193 15. Shackford SR, Zhuang J, Schmoker J (1992) Intravenous uid tonicity: effect on intracranial pressure, cerebral blood ow, and cerebral oxygen delivery in focal brain injury. J Neurosurg 76:9198 16. Gordon GR, Choi HB, Rungta RL, Ellis-Davies GC, MacVicar BA (2008) Brain metabolism dictates the polarity of astrocyte control over arterioles. Nature 456:745749 17. Marshall LF, Marshall SB, Klauber MR, van Clark Berkum M, Eisenberg HM, Jane JA, Luerssen TG, Marmarou A, Foulkes MA (1991) A new classication of head injury based on computerized tomography. J Neurosurg 75:S14S20 18. Bratton SL, Chestnut RM, Ghajar J, McConnell Hammond FF, Harris OA, Hartl R, Manley GT, Nemecek A, Newell DW, Rosenthal G, Schouten J, Shutter L, Timmons SD, Ullman JS, Videtta W, Wilberger JE, Wright DW (2007) Guidelines for the management of severe traumatic brain injury VIII. Intracranial pressure thresholds. J Neurotrauma 24(Suppl 1):S55S58 19. Stewart PA (1978) Independent and dependent variables of acid-base control. Respir Physiol 33:926 20. Stocchetti N, Colombo A, Ortolano F, Videtta W, Marchesi R, Longhi L, Zanier ER (2007) Time course of intracranial hypertension after traumatic brain injury. J Neurotrauma 24:13391346 21. Roquilly A, Mahe PJ, Latte DD, Loutrel O, Champin P, Di Falco C, Courbe A, Buffenoir K, Hamel O, Lejus C, Sebille V, Asehnoune K (2011) Continuous controlled-infusion of hypertonic saline solution in traumatic brain-injured patients: a 9-year retrospective study. Crit Care 15:R260 22. Froelich M, Ni Q, Wess C, Ougorets I, Hartl R (2009) Continuous hypertonic saline therapy and the occurrence of complications in neurocritically ill patients. Crit Care Med 37:14331441 23. Prins ML (2008) Cerebral metabolic adaptation and ketone metabolism after brain injury. J Cereb Blood Flow Metab 28:116 24. Holloway R, Zhou Z, Harvey HB, Levasseur JE, Rice AC, Sun D, Hamm RJ, Bullock MR (2007) Effect of lactate therapy upon cognitive decits after traumatic brain injury in the rat. Acta Neurochir (Wien) 149:919927 25. Levasseur JE, Alessandri B, Reinert M, Clausen T, Zhou Z, Altememi N, Bullock MR (2006) Lactate, not glucose, up-regulates mitochondrial oxygen consumption both in sham and lateral uid percussed rat brains. Neurosurgery 59:11221130 26. Rice AC, Zsoldos R, Chen T, Wilson MS, Alessandri B, Hamm RJ, Bullock MR (2002) Lactate administration attenuates cognitive decits following traumatic brain injury. Brain Res 928:156159 27. Berthet C, Lei H, Thevenet J, Gruetter R, Magistretti PJ, Hirt L (2009) Neuroprotective role of lactate after cerebral ischemia. J Cereb Blood Flow Metab 29:17801789 28. King P, Kong MF, Parkin H, MacDonald IA, Barber C, Tattersall RB (1998) Intravenous lactate prevents cerebral dysfunction during hypoglycaemia in insulin-dependent diabetes mellitus. Clin Sci (Lond) 94:157163 29. Tsacopoulos M, Magistretti PJ (1996) Metabolic coupling between glia and neurons. J Neurosci 16:877885 30. Belanger M, Allaman I, Magistretti PJ (2011) Brain energy metabolism: focus on astrocyte-neuron metabolic cooperation. Cell Metab 14:724738

1422

38. Vespa P, Bergsneider M, Hattori N, Wu 31. Prieto R, Tavazzi B, Taya K, Barrios L, 35. Kahle KT, Staley KJ, Nahed BV, HM, Huang SC, Martin NA, Glenn TC, Gamba G, Hebert SC, Lifton RP, Mount Amorini AM, Di Pietro V, Pascual JM, McArthur DL, Hovda DA (2005) DB (2008) Roles of the cation-chloride Marmarou A, Marmarou CR (2011) Metabolic crisis without brain ischemia cotransporters in neurological disease. Brain energy depletion in a rodent is common after traumatic brain injury: Nat Clin Pract Neurol 4:490503 model of diffuse traumatic brain injury a combined microdialysis and positron is not prevented with administration of 36. Jourdain P, Pavillon N, Moratal C, Boss emission tomography study. J Cereb D, Rappaz B, Depeursinge C, Marquet sodium lactate. Brain Res 1404:3949 Blood Flow Metab 25:763774 P, Magistretti PJ (2011) Determination 32. Mustafa I, Leverve XM (2002) 39. Pannier JL, Demeester G, Leusen I of transmembrane water uxes in Metabolic and hemodynamic effects of (1974) The inuence of nonrespiratory neurons elicited by glutamate hypertonic solutions: sodium-lactate alkalosis on cerebral blood ow in cats. ionotropic receptors and by the versus sodium chloride infusion in Stroke 5:324329 cotransporters KCC2 and NKCC1: a postoperative patients. Shock 40. Yoon S, Zuccarello M, Rapoport RM digital holographic microscopy study. 18:306310 (2012) PCO(2) and pH regulation of J Neurosci 31:1184611854 33. Leverve XM, Boon C, Hakim T, Anwar cerebral blood ow. Front Physiol 37. Bergsneider M, Hovda DA, Shalmon E, M, Siregar E, Mustafa I (2008) Half3:365. doi:10.3389/fphys.2012.00365 Kelly DF, Vespa PM, Martin NA, molar sodium-lactate solution has a Phelps ME, McArthur DL, Caron MJ, benecial effect in patients after Kraus JF, Becker DP (1997) Cerebral coronary artery bypass grafting. hyperglycolysis following severe Intensive Care Med 34:17961803 traumatic brain injury in humans: a 34. Halestrap AP, Wilson MC (2012) The positron emission tomography study. monocarboxylate transporter family J Neurosurg 86:241251 role and regulation. IUBMB Life 64:109119