| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Department of Anesthesia and Intensive Care, Royal Adelaide Hospital/University of Adelaide, North Terrace, Australia; and Department of Neuroanesthesia, Århus University Hospital, Denmark
Address correspondence to Mads Rasmussen, MD, PhD, Department of Neuroanesthesia, Århus University Hospital, Nørrebrogade 44, 8000 Århus C., Denmark. Address e-mail to maras{at}as.aaa.dk.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In distinction to standard treatment strategies, such as hyperventilation and osmotic therapy, IV administration of indomethacin rapidly reduces cerebral blood flow (CBF) and ICP along with increases in mean arterial blood pressure (MABP) and cerebral perfusion pressure (CPP) in isoflurane-anesthetized patients undergoing craniotomy (3). In contrast to these findings, a recent randomized study demonstrated that a bolus dose of indomethacin followed by continuous infusion did not reduce ICP or the degree of brain swelling in propofol- and fentanyl-anesthetized tumor patients (4). The different effects on cerebral hemodynamics of volatile anesthetics compared with propofol may explain the lack of effect under these conditions. Whereas some studies have demonstrated that the reduction in CBF with propofol anesthesia was proportional to the decrease in cerebral metabolic rate of oxygen (5), others have demonstrated that the percentage reduction of CBF was larger than the reduction of cerebral metabolic rate of oxygen (6,7). This suggests that propofol may have a direct cerebral vasoconstricting effect in some circumstances. In contrast, during isoflurane anesthesia, ICP and CBF are either unchanged compared with values obtained during fentanyl/nitrous oxide (8) or seem to increase with the larger anesthetic concentrations (9).
Any interaction between different anesthetics and the effect of indomethacin on ICP requires exploration because the therapeutic effect of indomethacin may be dependent on the choice of anesthetic regimen. Thus, the aim of this study was to investigate the effects of a bolus dose of indomethacin on ICP during isoflurane or propofol anesthesia in a sheep model of intracranial hypertension. Furthermore, we studied whether CBF and arterio-venous oxygen difference (AVDO2) differed between groups.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
) in ICP is 6 mm Hg, standard deviation (sd) of 3 mm Hg, power of 80%, and a significance level of 5%. A crossover design was used in which every sheep was subjected to either isoflurane or propofol anesthesia in random order to remove the between-animal variation from the treatment differences. Thus, six sheep were included. Sheep were housed in floor pens between experiments and had free access to food and water. Female Merino sheep with a mean weight of 46 kg (range, 38–53 kg) were instrumented under thiopental and halothane anesthesia, as described previously (10). In brief, catheters were chronically implanted in the pulmonary artery (for drug administration), the superior sagittal sinus (for sampling of cerebral venous blood), and the femoral artery (for measurement of MABP and sampling of arterial blood). An ultrasonic, range-gated Doppler probe (Tritonics Medical Instruments, Iowa City, IA) was positioned on the dorsal sagittal sinus for measurement of changes in CBF (11,12). For determination of ICP, a strain-gauge tipped ICP monitor (Microsensor ICP transducer, Codman, Randolph, MA) was placed in the epidural space and calibrated according to instructions from the manufacturer. To increase ICP, an 8F Foley catheter with a 3-mL balloon was placed in the epidural space opposite to the ICP sensor. Silver electroencephalographic (EEG) and reference electrodes were placed on the skull and under the skin. The anesthetic drug effect was quantitated from the EEG using entropy, which is a valid indicator of the hypnotic effect of propofol and volatile anesthetics (13,14). The entropy indices, response entropy (RE), and state entropy (SE) were derived from the EEG and electromyographic signal and calculated with the Datex-Ohmeda S/5 Entropy Module (Datex-Ohmeda Division, Instrumetarium Corp, Helsinki, Finland) every second (15). The Doppler transducer, ICP monitor, sagittal sinus catheter, Foley catheter, and electrodes for entropy measurement were then secured, and the sheep were allowed to recover from anesthesia. A period of 2–4 days elapsed between animal preparation and the first study. Two days elapsed between the studies in each sheep.
On the day of study, sheep were placed in a metabolic crate and supported by a sling. Measurements consisted of ICP, MABP, sagittal sinus Doppler frequencies, SE, RE, and blood gases for determination of Paco2, Pao2, and AVDO2. Measurements were performed in the awake sheep before the anesthetic induction (time point 0). Anesthesia was induced with propofol (5 mg/kg IV), and the sheep was endotracheally intubated. Controlled ventilation was applied (fraction of inspired oxygen [Fio2] = 50% with oxygen/air) to achieve Paco2 and Pao2 levels in the range of 35–40 mm Hg and >100 mm Hg, respectively. Anesthesia was maintained with 2% end-tidal of isoflurane or with an infusion of 0.66–0.47 mg · kg–1 · min–1 of propofol to achieve equal anesthetic depth with SE values between 30 and 40 and RE values between 40 and 50. Fluid therapy consisted of IV isotonic saline 5 mL · kg–1 · h–1throughout the experiment.
After 30 min, to allow the induction drug to clear (16) and to obtain steady-state conditions, "anesthesia alone" measurements were performed (time point 1). ICP was then increased to approximately 15–20 mm Hg over a period of 15 min by gradually filling of the epidural balloon. After an additional 5 min, pre-indomethacin (pre-indo) measurements were performed (time point 2). An IV bolus injection of indomethacin (Confortid®) 0.2 mg/kg (see Appendix) was administered over 1 min, and measurements were repeated 3, 5, 10, 20, 30, and 40 min after the start of the indomethacin injection (time points 3–8). The total duration of the experiment was 95 min.
Sagittal sinus Doppler frequencies and pressure data were continuously sampled at 1 Hz to a personal computer (Microbits IBM 486-based IBM compatible) and stored in a database. Changes in CBF were inferred from changes in Doppler frequencies, which reflect the blood velocity in sagittal sinus. Quantifying CBF with a Doppler ultrasound venous outflow method has been validated in sheep as a measure of global CBF against angiographic, retrograde dye, and timed venous outflow studies (12). This method represents 75% of total CBF and has been demonstrated to be in agreement with measurements made using the Kety-Schmidt nitrous oxide method in sheep (11). Thus, we will use the term CBF to describe changes in sagittal sinus blood flow velocity. Blood flow velocity in the sagittal sinus is only representative of CBF if the diameter of the sagittal sinus is constant. Because the inflation of the epidural balloon may have compressed the sinus, it was not considered appropriate to directly compare the velocities recorded before and after the inflation of the balloon. Thus, it was not possible to determine the effect of increasing the ICP on CBF. However, the following comparisons could be made unaffected by this concern: (a) the effect of anesthesia on CBF by comparing time points 1 and 0 and (b) the effect of indomethacin on CBF by comparing time points 3–8 with time point 2.
Entropy indices were recorded using a second personal computer (IBM compatible, Pentium). Blood gases were measured using a standard blood gas analyzer (ABL 625, Radiometer, Copenhagen, Denmark). At each time point, ICP, CBF, and MABP data were averaged during a period of 20 s using data averaging software. Entropy data were averaged for 1 min.
We used repeated-measures analysis of variance, across time points 2–8, to analyze ICP, MABP, Paco2, Pao2, and AVDO2 data and across time points 3–8 to analyze CBF data. Unless otherwise stated, the analysis of variance term of interest was the treatment group x time interaction. This tests whether the profile of the outcome variable over time was affected by treatment group. The Greenhouse-Geisser adjustment was used to correct for multi-sample asphericity (17). Uncertainty surrounding means is expressed as the within-subject sd (18).
Paired t-tests adjusted for multiple comparisons with step-down Holm-Bonferroni corrections were used for analysis of entropy data, awake ICP, and MABP data and whether the order of the studies affected awake ICP. A paired t-test (not adjusted for multiple comparisons) was used for analysis of fluid volume required to increase ICP. Entropy data, fluid volume, awake ICP, and MABP data are expressed as mean ± sd. SYSTAT version 10 (SPSS Inc., Chicago, IL) was used to analyze the data. Two-sided P 
0.05 was regarded as statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
During anesthesia, SE values (isoflurane = 30 ± 7; propofol = 38 ± 5; P = 0.71) and RE values (isoflurane = 40 ± 9; propofol = 46 ± 2; P = 0.71) were comparable. No statistically significant differences in the values of Paco2, Pao2, and AVDO2 between the two anesthetic groups were demonstrated (Table 1). The "anesthesia alone" MABP was higher with propofol than for isoflurane (P = 0.04; Fig. 1) without differences in the profile of the means (P = 0.31; Fig. 1). The two anesthetics had different effects on CBF. For isoflurane, the CBF was reduced by 11% compared with the awake value, whereas the reduction was 34% for propofol.
|
|
Adding a fixed volume to the intracranial balloon produced an immediate increase in ICP. Typically, the ICP then decreased from its peak value and reached a plateau that thereafter decreased slowly. By adding volume in decreasing increments over 15 min, it was possible to reach a relatively stable level of increased ICP that did not differ between the two anesthetic groups (Fig. 2; P = 0.79). There was no significant difference between the anesthetic groups in the mean volume required to increase the ICP to the target value (isoflurane = 3.0 ± 2 mL; propofol = 2.7 ± 2 mL; P = 0.44).
|
For both anesthetic groups, the administration of indomethacin was characterized by an increase in MABP and decreases in CBF and ICP. The increase in MABP (albeit from different anesthesia-alone values) was similar between anesthetics (Fig. 1), with no statistical difference in the profile of mean values (P = 0.31). The effects of indomethacin on ICP are shown in Figure 2. There was a significant difference in profile of means between the two groups because of a more pronounced decrease in ICP with isoflurane (P = 0.009).
However, there were important quantitative differences between the drugs with respect to the cerebral vasculature. By comparing time points 3–8 with time point 2 after the balloon was inflated, it is possible to infer the effect of indomethacin on CBF (Fig. 3). Indomethacin caused an additional 17% and 14% (maximum) decrease in CBF for isoflurane and propofol, respectively. There was no difference in the profile of mean values (P = 0.43). The between-groups analysis showed that the decrease in CBF with indomethacin was significantly more pronounced for the propofol group (P = 0.02). In summary, whereas indomethacin produced a slightly larger reduction in CBF under isoflurane anesthesia, this reduction was from a higher pre-indomethacin value because of the intrinsic differences in the effect of isoflurane and propofol on CBF. Thus, the combination of propofol and indomethacin resulted in a lower CBF than the combination of isoflurane and indomethacin (Fig. 3).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The rapid ICP-reducing effect of indomethacin is similar to that obtained with barbiturates. Thiopental, however, is accompanied by a decrease in CPP (22). In contrast, the effect of hyperventilation and mannitol is only maximal after 10–15 min and 30–60 min, respectively (23,24). Thus, compared with other treatments of high ICP, indomethacin is unique in affecting an immediate decrease in ICP associated with an increase in CPP.
Our results may have clinical significance. We previously demonstrated that a subdural ICP >13 mm Hg is associated with a 95% risk of brain swelling in patients undergoing brain tumor surgery (2). In the current study, we managed to reduce ICP to less than this threshold for periods of approximately 10 minutes and 15 minutes during propofol and isoflurane anesthesia, respectively. Although we were unable to measure the degree of brain swelling, the temporary reduction in ICP may be sufficient to reduce the incidence of brain swelling during craniotomy. This is supported by a previous study in tumor patients anesthetized with isoflurane in which a bolus dose of indomethacin significantly reduced ICP, the degree of brain tightness, and the need for further treatment to control ICP compared with a placebo group (3). Thus, the temporary reduction in ICP provided by indomethacin may enable the neurosurgeon to open the Dura and start further decompression without the hampering disturbance of brain swelling.
The finding that a bolus dose of indomethacin reduces ICP during propofol anesthesia is in contrast to a recent study where an indomethacin infusion did not influence ICP in patients with cerebral tumors (4). Several factors may explain this difference. First, in the study by Rasmussen et al. (4), the indomethacin infusion was administered before the anesthetic induction and terminated after opening of the Dura mater. The mean time interval between indomethacin administration and the first ICP measurement was 117 minutes. Baseline recordings of ICP were not performed, and there is the possibility that indomethacin initially decreased ICP but failed to sustain the effect during the long infusion time. The authors suggested that indomethacin-induced cerebral vasoconstriction shows adaptation; however, this has not been previously demonstrated (4). Second, Rasmussen et al. (4) demonstrated a 50% decrease in CBF after the induction with propofol compared to a 35% reduction in CBF observed in this study. Thus, propofol-induced cerebral vasoconstriction was possibly near maximal in the study by Rasmussen et al. (4), with limited room for further vasoconstriction by indomethacin. Consequently, the difference in the reduction of CBF and degree of vasoconstriction observed in the two studies may explain why indomethacin in the current study managed to further constrict the resistance vessels and reduce ICP. Third, differences regarding species, dose-response relationships, and methods of measuring ICP (subdural versus epidural) may have influenced the results.
The difference in the ICP-reducing effect of indomethacin during isoflurane and propofol could be explained by an intergroup difference in anesthetic depth. We found similar awake and anesthesia-alone entropy indices, indicating that the depth of anesthesia was similar in the two groups. It may be argued that entropy indices may not provide the same indication of anesthetic drugs in sheep compared with humans. However, the concept of entropy quantifies the degree of complexity and irregularity of the EEG signal and may thus be species-independent. This is supported by a recent study in which entropy values in sheep measured during the induction of anesthesia were similar to entropy values predicted from a theoretical human model (25). Furthermore, SE and RE values during propofol anesthesia were similar to values obtained in anesthetized humans (26).
In both groups, the indomethacin-induced decrease in ICP was accompanied by a near similar decrease in CBF. This effect could be explained by a direct vasoconstrictor effect of indomethacin on the cerebral resistance vessels. However, the finding that CBF did not recover from the decrease produced by indomethacin is not consistent with the time course of the ICP recovery, and this observation, therefore, suggests an alternative mechanism of indomethacin in decreasing CBF. An autoregulatory response to the simultaneous increase in MABP could contribute to the decrease in ICP. This is supported by the observation that the change in ICP is consistent with the time course and the magnitude of the MABP change. In this study, preserved autoregulation was substantiated by the observation that after the initial decrease in CBF caused by indomethacin, CBF remained unchanged despite alterations in CPP because of an increase in ICP. This is further supported by previous studies on the same model in which autoregulation was shown to be preserved during similar doses of isoflurane and propofol (27,28). Thus, our data suggest that the effect of indomethacin on the cerebral resistance vessels is partly mediated by a myogenic autoregulatory response.
It may be argued that the observed 3% difference in the relative reduction of CBF after indomethacin administration is insufficient to explain the significant difference in the reduction of ICP. Although the decrease in CBF with indomethacin was nearly similar, CBF and, hence, cerebral blood volume (CBV), were supposed to be larger in the isoflurane group. This implies that the pressure-volume compensatory reserve during high ICP was reduced to a larger extent during isoflurane anesthesia compared with propofol. Under these conditions, the administration of indomethacin would cause a larger reduction in ICP during isoflurane, as observed. However, the relative reduction in CBF and possibly CBV would theoretically be the same in both groups as demonstrated. Furthermore, previous studies suggest that changes in ICP are determined largely by changes in CBV and not CBF (29).
To apply the results from this study to the clinical setting, it is appropriate to discuss whether the CBF effects of isoflurane and propofol in sheep are similar to humans. Compared with the awake state, we observed that isoflurane and propofol reduced CBF by 11% and 34%, respectively. The reduction in CBF during isoflurane is in agreement with previous studies in humans in which isoflurane caused a 22% and 27% reduction in CBF compared with awake measurements (30,31). Conversely, other clinical studies with isoflurane have demonstrated that CBF either remains unchanged compared with values obtained during fentanyl/nitrous oxide (8) or increases with increasing anesthetic concentrations (9). The decrease in CBF observed during propofol anesthesia is consistent with previous studies in humans in which propofol caused a 40%–60% reduction in CBF compared with awake measurements (5,32).
There are some potential limitations of the present study. ICP was increased by inflation of an epidural balloon over a short time period. This method mimics a rapidly expanding extradural hematoma and may not resemble the effect on intracranial dynamics of a slow growing tumor possibly surrounded by edema. This difference may influence brain compliance and hence the effect of indomethacin in reducing ICP. However, it was not the intention of this study to mimic the dynamics of a cerebral tumor. Instead, our primary intention was to produce a controllable increase in ICP that was associated with a risk of brain swelling and suitable for the evaluation of therapeutic strategies for increased ICP.
The changes in CBF were estimated from changes in Doppler frequencies measured in the superior sagittal sinus. This can be compared with the concurrent oxygen extraction measurements. Although there was no statistical significance between the AVDO2 values in the 2 groups (P = 0.13), there was a tendency to consistently lower AVDO2 values during propofol anesthesia, which is in accordance with the CBF measurements. Furthermore, the increase in AVDO2 values five minutes after the administration of indomethacin (relative to predrug values) was comparable (42% and 47%), which is in agreement with the comparable decrease in CBF. However, the changes in AVDO2 suggest that the absolute reduction in CBF in both groups may have been more pronounced than demonstrated.
In contrast to humans, sheep have a carotid rete mirabile—a small network of arteries that could influence the CBF response to indomethacin. However, studies have demonstrated that the role of the carotid rete mirabile in regulation of resistance to blood flow is negligible (33). Thus, the cerebrovascular response to indomethacin in sheep seems relevant to that observed in humans.
In conclusion, we demonstrated that a bolus dose of indomethacin reduces ICP and increases CPP during both propofol and isoflurane anesthesia, with the effect on ICP being more pronounced during isoflurane. We suggest that this effect is partly mediated by an autoregulatory mechanism.
We thank Dr. John Ludbrook, Director, Biomedical Statistical Consulting Service, for assistance with statistical analysis.
|
Footnotes |
|---|
Accepted for publication December 15, 2005.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
