The potential adverse effects of ketamine in neurosurgical anesthesia have been well established and involve increased intracranial pressure (ICP) and cerebral blood flow. However, reexamination of ketamine is warranted because data regarding the effects of ketamine on cerebral hemodynamics are conflicting.
Eight patients with traumatic brain injury were studied. In all patients, ICP monitoring was instituted before the study. Control of ICP (less than 25 mmHg), hemodynamic values, and blood gas tension (partial pressure of carbon dioxide in arterial blood between 35-38 mmHg) was obtained with propofol infusion (3 mg x kg(-1) x h(-1)) and mechanical ventilation. The effects of three doses of ketamine, 1.5, 3, and 5 mg/kg, respectively, on ICP, cerebral perfusion pressure, jugular vein bulb oxygen saturation, middle cerebral artery blood flow velocity, and electric activity of the brain (EEG) were measured. The three doses were administered intravenously at 6-h intervals over 30 s through a central venous line. Systemic and cerebral hemodynamics and end-tidal carbon dioxide were continuously monitored and recorded at 1-min intervals throughout the 30-min study periods.
Ketamine, in all three doses studied (1.5, 3, and 5 mg/kg) was associated with a significant decrease in ICP (mean +/- SD: 2 +/- 0.5 mmHg [P < 0.05], 4 +/- 1 mmHg [P < 0.05], and 5 +/- 2 mmHg [P < 0.05]) among the study patients regardless of the ketamine dose used. There were no significant differences in cerebral perfusion pressure, jugular vein bulb oxygen saturation, and middle cerebral artery blood flow velocity. Ketamine induced a low-amplitude fast-activity electroencephalogram, with marked depression, such as burst suppression.
These results suggest that ketamine may not adversely alter cerebral hemodynamics of mechanically ventilated head-trauma patients sedated with propofol. These encouraging results should be confirmed in larger groups of similar patients.
An important objective in the medical treatment of patients with severe head injury is maintenance of an adequate cerebral perfusion pressure (CPP). This can be done by controlling intracranial pressure (ICP) and maintaining adequate mean arterial blood pressure (MAP), blood gases, body temperature, serum glucose concentration, electrolytes, and osmolarity. Sedation is one of the major treatments of intracranial hypertension. Ketamine is an anesthetic drug that is well suited to hypovolemic or normovolemic patients because it stimulates the cardiovascular system and furthermore maintains hemodynamic status. Unfortunately, potential adverse effects of ketamine in neurosurgical anesthesia have been well established,  and the drug is usually contraindicated in neurosurgical patients who have intracranial hypertension because of its reported effect on ICP and cerebral blood flow (CBF). [2,3] However, ketamine has been shown to decrease CBF  and cerebral spinal fluid pressure  in animals and does not to affect ICP in anesthetized humans.  Thus anesthetics and the pressure of carbon dioxide in arterial blood (PaCO2) appear to influence ketamine effects on the cerebral vasculature. The purpose of this study was to assess the effects of ketamine bolus on cerebral hemodynamics and electroencephalographic (EEG) activity, in patients with traumatic brain injury under controlled conditions.
Materials and Methods
After approval by the ethics committee of our institution, we obtained informed consent from members of each patient's family.
We studied eight male patients admitted to the intensive care unit with severe head injury (Glasgow Coma Scale score <or= to mg). Table 1shows intracranial disease, age, initial diagnosis, initial Glasgow Coma Scale score, associated injuries, and outcome concerning these patients. The hemodynamic status of the patients was considered stable: ICP < 25 mmHg for 3 h and CPP > 65 mmHg for 6 h or more. Sedation was achieved using a continuous infusion of propofol (3 mg [center dot] kg sup -1 [center dot] h sup -1), and neuromuscular blockade was performed using a continuous infusion of vecuronium bromide (8 mg/h). The PaCO2was maintained between 35–38 mmHg.
Heart rate, arterial hemoglobin oxygen saturation, end-tidal CO2, and invasive arterial blood pressure were continuously monitored with a component monitoring system (model 66; Hewlett-Packard, Waltham, MA). The ICP was continuously monitored with a Camino Catheter System (OLM Intracranial Pressure Monitoring Kit; Camino Laboratories, San Diego, CA), which uses a sub-arachnoid bolt and a sterile miniature ICP transducer. A jugular bulb catheter (Edslab double-lumen O2Sat II 94–040 -4F Baxter Healthcare Corp., Irvine, CA) was inserted in the ipsilateral jugular vein of the ICP monitoring system.
The level of sedation was deepened with increasing doses of ketamine of 1.5, 3, and 5 mg/kg. The three doses were administered intravenously at 6-h intervals, in 30 s through a special intravenous line. Heart rate, MAP, ICP, CPP (MAP-PIC), oxygen saturation, jugular bulb oxygen saturation, and end-tidal CO2were continuously measured and recorded at 1-min intervals throughout each 30-min study period. Arterial and jugular vein bulb blood samples were obtained at baseline and at 5 and 20 min after administration of each dose of ketamine. The cerebral arteriovenous oxygen content difference (AVDO2) was calculated according to the following formula: AVDO2=(Hgb x (SaO2, - SvjO2) x 1.39)+([PaO2- PjvO2,]) x 0.003), where Hgb = hemoglobin concentration, SaO2= arterial hemoglobin oxygen saturation, SvjO2, = jugular bulb oxygen saturation, PaO2= arterial oxygen partial pressure, and PjvO2= jugular vein bulb oxygen partial pressure. Arteriojug-ular lactate difference (AVDL) was arterial lactate concentration minus jugular bulb lactate concentration expressed in mM. Lactate-oxygen index (LOI =-AVDL/AVDO2) was considered an ischemia index, and a LOI of 0.08 or more accurately predicts increased cerebral lactate production. A 2-MHz pulsed Doppler ultrasound device (Angiodine 2; DMS, Montpellier, France) was used to measure erythrocyte velocity. After identification of the right anterior cerebral artery and middle cerebral artery (MCA), the depth was adjusted by 2-mm increments to obtain signals from the proximal (M1) segment of the MCA. The intonation depth was between 45–50 mm. The mean blood flow velocity (VMCA) was calculated from the formula: VMCA=[(systolic flow velocity-diastolic flow velocity)/3]+ diastolic flow velocity. Electroencephalogram was continuously monitored with a 10-channel polygraph (ECEM Medelec, Paris, France) at a speed of 15 mm/s. Two (C4–02, C3-O1) or six (Fp2-C4, C4-T4, T4-O2, Fpl-C3, C3-T3, and T3-O1) scalp-to-scalp bipolar linkages were recorded for 15 min before the ketamine bolus and during the 30 min of the study period. Each EEG was visually analyzed by one of us (M.R.) without knowledge of ICP and CPP variations. Increased ICP deemed clinically dangerous could be treated by the critical care nurses in accordance with standard therapy in the critical care unit.
Results are presented as their mean +/- SD. Baseline values represent an average of five measurements obtained during 5 min before drug administration. Analysis of variance for repeated measurements and Dunnett's test were used to determine the effects of the three doses within each group and between groups. A probability value < 0.05 was considered significant.
Changes in heart rate, MAP, ICP, CPP, and jugular bulb oxygen saturation are presented in Table 2and in Figure 1for individual ICP values. There were no differences in baseline values for the studied parameters. In the three groups, there were no statistically significant variations in heart rate, MAP, and CPP. However, a small but significant (P < 0.05) decrease in ICP was observed in the three doses. Maximum decreases in ICP were 18% after 1.5 mg/ kg, 30% after 3 mg/kg, and 30% after 5 mg/kg and were seen after 3, 2, and 3 min, respectively. In the 1.5-mg/kg group, ICP increased slightly (27%) at 10 min, and in the 5-mg/kg group, it increased at 30 min (23%); these changes were statistically significant. During the study period pH, PaO2, paC sub O2, AVDO2, and LOI (Table 3) remained unchanged. No change was observed in VMCA(Table 2). In all patients, EEG before ketamine injection was characterized by a regular, continuous, medium voltage slow-wave activity associated in four patients with some superimposed low-amplitude fast activity. The basal EEG tracings were similar before each injection. Ketamine induced a low-amplitude fast EEG activity with electrogenic depression such as burst suppression (Figure 2and Table 4).
Our data show that ketamine administration in propofol-sedated patients with traumatic brain injury under controlled conditions was not associated with clinically significant changes in cerebral hemodynamics. Depending on the experimental model used, ketamine has been shown to increase, [2,3] decrease, [6–8] or have no effect on ICP. [4,9] This lack of consistency may be, in part, a result of differences in experimental designs among studies, such as absence or presence of other medications or use of background anesthetics, and control of PaCO2. For instance, Pfenninger et al.  showed that ketamine induces a marked respiratory depression in the presence of intracranial hypertension, and they suggest that increased ICP is caused by ensuing hypercapnia. Further, as an N-methyl-D-aspartate receptor antagonist, ketamine has been shown experimentally to have neuroprotective properties during transient ischemia [10,11] and experimental head trauma in rats.  However, ketamine is considered contraindicated in patients with increased ICP. Because there are conflicting results regarding ketamine effects on cerebral hemodynamics, and because ketamine is the only N-methyl-D-aspartate receptor antagonist currently approved for clinical use as an anesthetic, we believe that the use of ketamine in patients with traumatic brain injury warrants reevaluation.
In the present study, we found that in sedated and ventilated patients with traumatic brain injury, ketamine decreases ICP, does not decrease MAP, and AVDO2and VMCA, do not change. Electroencephalographic activity is decreased with EEG depression, such as burst suppressions. Electroencephalographic activity data suggest that ketamine may depress cerebral metabolism oxygen consumption (CMR sub O2). Because VMCA and AVDO2did not change in the patients we studied, we may hypothesize that flow remained coupled to metabolism and that CBF and CMRO2may have decreased in the same way. Other studies in animals and patients without head trauma have reported similar effects on ICP, [8–10] cerebral hemodynamics, [6,14] and metabolism. [7,15,16] In a study of anesthetized patients undergoing craniotomy, Mayberg et al.  found that after ketamine (1 mg/kg), MAP, CPP, and AVDO2were unchanged. In addition, ICP decreased significantly, as did VMCA. As stated by the authors themselves, their study does not address any cerebral protective effect of ketamine, but rather their results show that ketamine can be given to ventilated, anesthetized patients without adversely altering cerebral hemodynamics.  In the present study, we also found that ketamine may not be contraindicated in all patients at risk for intracranial hypertension. Indeed, in patients with headinjury who are sedated with propofol, doses of ketamine as high as 5 mg/kg do not induce adverse effects on cerebral and systemic hemodynamics.
We did not directly measure either CBF or CMRO2. The transcranial Doppler was used because it allows continuous estimation of CBF in a noninvasive manner. Because VMCAis determined by both CBF and the diameter of the MCA, which may vary among the population, absolute values for VMCAand CBF do not correlate well. However, cerebral arterial diameters have been shown to remain fairly constant with changes in blood pressure and PaCO2,  and the percentage change in VMCAhas been shown to correlate well with the percentage change in CBF.  In the current study, VMCAwas not modified. Although CMRO2was not directly measured, our data suggest that CMRO2was at least not increased and may have been decreased because EEG activity decreased with burst suppressions, whereas AVDO2remained unchanged.
To reconcile our present findings with previously reported effects of ketamine, it is important to discern the contributing roles of modes of ventilation and background anesthesia. Ketamine is usually avoided in the anesthetic management of patients at risk for intracranial hypertension because early studies suggested associated increases in CMR sub O2, CBF, and ICP. [2,3] The reported increases in CBF, however, may be partly mediated by an increase in blood pressure and partly by a concomitant increase in PaCO2in spontaneously breathing patients. More recent studies report no increase in ICP when ventilation is controlled [9,10] or when diazepam  or thiopenthal  is given concurrently. When administered in a background of isoflurane-N2O anesthesia, no increase in flow velocity is seen.  A more plausible explanation for our findings is that when ketamine is added to a background anesthetic, its property of central nervous “excitation” is blunted and it increases the depth of anesthesia. The observation of burst suppressions in EEG activity supports the contention that ketamine acts to increase the depth of anesthesia. The explanation that the excitatory properties of ketamine are blunted by propofol is supported by the fact that Hui et al.  showed that the sedative effects of the combination of ketamine and propofol were additive at hypnotic and anesthesia end points. Another possible explanation is that the peripheral vascular stimulation often seen with ketamine was blunted in the present study, as demonstrated by a lack of increase in MAP, as it has been reported by other authors. [14,20] Finally, to what extent the results of the present study were affected by the cumulative effects of ketamine administered during the 12-h study period cannot be evaluated given the design of the study protocol. However, we must note that the effects on ICP and EEG activity were observed a few minutes after each ketamine injection.
In conclusion, the data of the present study show that, in ventilated patients with traumatic brain injury who are sedated with propofol, ketamine does not increase MAP and mildly decreases ICP. The lack of change in AVDO2and VMCAsuggests no impairment in the balance between CBF and oxygen metabolism. The burst suppressions observed in EEG activity may account for a decrease in CMRO2. The present study does not address any cerebral protective effect of ketamine but indicates a clinical advantage because it maintains CPP.