Performance of Anesthetic Depth Indexes in Rabbits under Propofol Prediction Probabilities and Concentration-effect Relations

Aura Silva, M.Sc.,* So´ nia Campos, M.Sc.,† Joaquim Monteiro, Ph.D.,‡ Carlos Venaˆ ncio, M.Sc.,* Bertinho Costa, Ph.D.,§ Paula Guedes de Pinho, Ph.D.,ʈ Luis Antunes, Ph.D.# Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021 ABSTRACT What We Already Know about This Topic • A number of electroencephalogram-based measures have Background: The permutation entropy, the approximate been studied to determine their usefulness for intraoperative entropy, and the index of consciousness are some of the most of anesthetic depth • recently studied electroencephalogram-derived indexes. In In rats anesthetized with a potent volatile anesthetic, burst suppression-corrected permutation entropy had the best cor- this work, a thorough comparison of these indexes was per- relation with anesthetic dose formed using propofol anesthesia in a rabbit model. Methods: Six rabbits were anesthetized with three propofol Ϫ Ϫ infusion rates: 70, 100, and 130 mg ⅐ kg 1 ⅐ h 1, each main- tained for 30 min, in a random order for each animal. What This Article Tells Us That Is New Data recording was performed in the awake animals 20, • In rabbits receiving propofol by infusion, burst suppression- 25, and 30 min after each infusion rate was begun in the corrected permutation entropy and burst suppression-cor- rected composite multiscale permutation entropy had the recovered animals and consisted of electroencephalogram best propofol concentration-effect relationships and predic- recordings, evaluation of depth of anesthesia according to a clinical tions of clinical signs scale, and arterial blood samples for plasma propofol determination. Median and spectral edge frequencies were analyzed for single-scale permutation entropy and composite multiscale permutation en- better prediction probabilities than did the other electroen- tropy, approximate entropy, index of consciousness, and the spec- cephalogram-derived parameters but not better than the tral parameters. The spectral parameters and single-scale and mul- electromyographic activity. tiscale permutation entropies were corrected for the presence of Conclusion: Single-scale and multiscale permutation entro- burst suppression. Performance of the indexes was compared by pies may be promising measures of propofol anesthetic depth prediction probability and pharmacodynamic analysis. when corrected for burst suppression. Additional studies should Results: The single-scale and composite multiscale permu- investigate the information measured by electromyography al- tation entropies with a burst suppression correction showed gorithms from commercial monitors of anesthetic depth. The rabbit may be a promising animal model for electroencephalo- * Doctoral Student, † Master’s Degree Student, Veterinary Sciences graphic studies because it provides a good-quality signal. Department, Universidade de Tra´s-os-Montes e Alto Douro, Via Real, Portugal. ‡ Researcher, ʈ Professor, REQUIMTE, Toxicology and Pharmacology Departments, Faculdade de Farma´cia da Universi- N recent years, the electroencephalogram has been ac- dade do Porto, Porto, Portugal. § Assistant Professor, Instituto Su- I tively studied for intraoperative monitoring of anesthetic perior Te´cnico, Universidade Te´cnica de Lisboa, Lisbon, Portugal. depth, resulting in a multitude of measures extracted from # Professor, Veterinary Sciences Department, Universidade de Trás-os- Montes e Alto Douro, and Researcher, Institute for Molecular and Cell this signal to translate it into a simple number. Biology, Porto, Portugal. Most recent parameters are based in nonlinear time series Received from the Veterinary Sciences Department, Universidade analysis, attempting to overcome limitations of older mea- de Tra´s-os-Montes e Alto Douro, Vila Real, Portugal. Submitted for sures based on the power spectrum.1,2 However, there is no publication August 3, 2010. Accepted for publication April 14, 2011. Supported by the Foundation for Science and Technology, Lisbon, consensus on the choice of the best parameter for clinical use. Portugal, under projects FCT PTDC/EEA-ACR/69288/2006, FCT PTDC/ Recently, permutation entropy (PE) was introduced and has CVT/101999/2008/COMPETE:FCOMP-01-0124-FEDER-009525, and FCT shown great potential because of ability to predict anesthetic PTDC/CVT/099022/2008/COMPETE:FCOMP-01-0124-FEDER-009497. concentration, high resistance to artifacts, and fast computa- Address correspondence to Dr. Silva: Quinta de Prados, Apartado 2 1013 5001-801 Vila Real, Portugal. [email protected]. Information on tion. It has been compared with approximate entropy (AE) purchasing reprints may be found at www..org or on in patients receiving sevoflurane and propofol anesthesia and the masthead page at the beginning of this issue. ANESTHESIOLOGY’s has shown better results.1,2 PE has been further explored in a articles are made freely accessible to all readers, for personal use only, 6 months from the cover date of the issue. multiscale approach in an attempt to better understand Copyright © 2011, the American Society of Anesthesiologists, Inc. Lippincott the dynamic characteristics of the electroencephalogram, Williams & Wilkins. Anesthesiology 2011; 115:303–14 resulting in composite multiscale permutation entropy

Anesthesiology, V 115 • No 2 303 August 2011 Anesthetic Depth Indexes in Propofol Anesthetized Rabbits

(CMSPE).3 Another recently studied measure is the index of Electroencephalographic Recording consciousness (IoC), the newest commercial monitor in this The electroencephalographic recordings were performed using context4,5; it is based in symbolic dynamics but, unlike the the IoC-View monitor (Aircraft Medical, Barcelona, Spain). PE, CMSPE, and AE, its exact calculation method is not Before induction of anesthesia, all of the animals’ heads totally revealed. Some limitations of PE, CMSPE, and AE were shaved, cleaned, and surface layers removed with fine have been discussed previously: the higher sensitivity to arti- sandpaper and acetone. Gel-coated, silver-silver chloride facts of AE and the difficulty of PE and CMSPE to detect electrodes (Swaromed, Innsbruck, Austria) were applied to burst suppression patterns in humans receiving sevoflurane record the electroencephalogram. Two electrodes (one for and propofol anesthesia.2,3,6 In a previous laboratory study, each eye) were placed 1 cm caudal to the lateral eye cantus; a central electrode was placed on the midline on the frontal the classic burst suppression correction could be applied suc- Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021 bone 3 cm from each previously applied electrode. This lo- cessfully to the PE in rats anesthetized with isoflurane, with ௡ better results than AE, IoC, and power spectrum-derived calization was based on previous works for the BIS monitor 10 parameters, median edge frequency (MEF) and spectral edge (Aspect Medical Systems, Newton, MA) in rabbits, which frequency 95 (SEF).7 has been determined to give the best quality electroenceph- Propofol is one of the most used hypnotics in clinical alographic signal after testing of different positions in pilot practice,8 and it is important to know its effects on these studies with the IoC-View Monitor (Morpheus-Medical, parameters, particularly to compare their performance at dif- Barcelona, Spain). ferent anesthetic depths achieved with the drug. This com- Impedance was automatically checked by the monitor and parison may benefit from the use of animal models because maintained at less than 15,000 ohms with a digitizing rate of 1,024 Hz. The electrodes were connected to the IoC-View such studies allow the comparison of parameter performance monitor, which was connected by Bluetooth® (Bluetooth SIG, in controlled conditions with reduced variability.9 In this Inc., Kirkland, WA) to a personal computer with the IoC-View context, the rabbit model shows great potential for transla- Graph software version 1.4 (Morpheus-Medical), a storage soft- tional research between animals and humans because of its ware provided by the manufacturer. small muscle layer between the skull and skin. We hypoth- esize that a rabbit model may allow the recording of an electroencephalographic signal with small electromyo- Anesthesia and Monitoring graphic contamination, which is important to guarantee After electroencephalogram baseline recording in the fully that brain effects, and not muscle tone, are being mea- awake animals for a period of 5 min, the fur on the ears was sured and reflected in the indexes. This is essential for clipped, the skin cleaned with alcohol, and a local analgesic future applications of the indexes in the electromyogra- cream was applied to the ear skin (EMLA; Nycomed US Inc, phy-rich human frontal montages and may contribute a Melville, NY). Thirty minutes later, two 22-gauge catheters rapid clarification on the potential use of different elec- were placed, one in the marginal ear vein and the other in the troencephalogram-derived indexes. central ear artery for arterial pressure monitoring. Both au- In the current study, the rabbit was used as a potential ricular catheter systems were flushed with heparinized saline translational research model for the comparison of clinical and fixed to the skin. The animals were then oxygenated with indexes because of its anatomic characteristics. The perfor- a facial mask at 5 l/min for 5 min. mance of different electroencephalogram-derived parameters Anesthesia was induced with intravenous propofol (20 was compared: IoC, PE, CMSPE, AE, MEF and SEF and the mg/kg) using a syringe pump (Asena GH; Alaris Medical burst suppression-corrected MEF, SEF, PE and CMSPE Systems, San Diego, CA) controlled by the Rugloop II soft- (BSMEF, BSSEF, BSPE, and BSCMSPE). The ability of ware (developed by Tom DeSmet, DEMED Engineering, Gent, Belgium) at an infusion rate of 200 ml/h. After blind these parameters to distinguish between different depths of orotracheal intubation with an endotracheal tube with an propofol anesthesia was assessed using prediction probability internal diameter of 2.5 mm, propofol administration began analysis and pharmacodynamic modeling. according to an infusion scheme in which three infusion rates were performed in every animal: each infusion (70, 100, and ⅐ Ϫ1 ⅐ Ϫ1 Materials and Methods 130 mg kg h ) was maintained for 30 min. The order of the administration rates for each animal was randomized Animals using the RAND function in Microsoft Excel 2007 (Mi- All procedures were carried out under personal and project crosoft Corporation, Redmond, WA). licenses approved by the national regulatory office (Direcc¸a˜o The rabbits were placed in ventral recumbence above a Geral de Veterina´ria, Lisbon, Portugal). Six healthy male heating blanket, and rectal temperature was continuously New Zealand white rabbits, average weight 2.79 Ϯ 0.25 kg, monitored and maintained between 37° and 38°C. Anes- were anesthetized for this study. Animals were group housed thetic monitoring included cardiorespiratory monitoring in floor pens with shelters, and behavior was assessed daily. provided by a Datex S/5 Anesthetic station (Datex Ohmeda,

Anesthesiology 2011; 115:303–14 304 Silva et al. PERIOPERATIVE MEDICINE

Table 1. Definition of the Anesthetic States and Attributed Numerical Scale

Numerical Scale Anesthetic State Clinical Signs Observed of Anesthesia Awake Fully awake and alert 0 Sedated Relaxed but still responsive to stimulation and with righting reflex 1 present Shallow anesthesia Lost its righting reflex but still responds to any stimulation 2 Medium anesthesia Responds only to painful stimulation (ear pinch and pedal withdrawal 3 reflexes) Surgical anesthesia Does not respond to painful stimulation (ear pinch and pedal withdrawal 4 reflexes) but still has corneal reflex Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021 Close to death Without corneal reflex, mean arterial blood pressure below 40 mmHg; 5 apneic

Definition of six anesthetic states based on clinical signs. The main clinical signs observed in each state are described.

Helsinki, Finland), which included pulse oxymetry and pulse Plasma Propofol Sampling rate monitoring with the probe placed in the ear, invasive Arterial blood samples were collected at exactly the same time arterial pressure, and inspired and end-tidal concentrations points that clinical evaluation of depth of anesthesia was of oxygen and carbon dioxide. These data were stored using performed, except at the extubation moment: in the awake the Rugloop II software. Animals were mechanically venti- animals, 20, 25, and 30 min after the beginning of each lated with 100% oxygen, with ventilation parameters set to infusion rate and in the totally recovered animal (fig. 1). This infusion scheme was designed to achieve a steady-state in the maintain the EtCO2 between 35 and 45 mmHg. At the end of the infusion scheme, the fresh gas flow rate last 10 min of each infusion rate based on pharmacokinetic 13 was increased to 5 l/min of 100% oxygen until the rabbits data of clearance from Cockshott et al. regained swallowing reflexes, and at this point extubation After the blood was collected, the serum was separated was performed. Animals were considered recovered from an- through centrifugation at 3,000 rpm for 15 min and imme- Ϫ esthesia when they exhibited an alert stance and had regained diately placed at 77°C and stored until analysis. ambulation and limb coordination. Continuous infusion of Propofol plasma concentrations were determined by gas Ϫ1 Ϫ1 chromatography mass spectrometry according Guitton et physiologic saline at a rate of 10 ml ⅐ kg ⅐ h was main- al.14 with some modifications. Briefly, 50 ␮l internal stan- tained during anesthesia. dard thymol solution (0.01 mg/ml) and 1 ml water were A clinical evaluation of depth of anesthesia was performed added to 0.5-ml aliquot serum or propofol calibration stan- according to a subjective numerical scale of anesthesia from 0 dards. To this solution, 0.5 ml borate buffer (pH 9) was (awake) to 5 (excessive depth of anesthesia) (table 1) based on added and mixed by inversion. Then, 300 ␮l chloroform: the evaluation of the postural reflex, muscular tone, palpebral ethylacetate (70:30, v:v) was added and mixed for 20 min at reflex, corneal reflex, laryngeal reflex, ear pinch, and digital 50 rpm, after which 1 ␮l organic phase solution was injected (pedal) reflexes. This scale was elaborated according to liter- into the gas chromatography-mass spectrometry injector in ature descriptions of clinical anesthetic depth evaluation in splitless mode at 250°C. The quantification of propofol was 10–12 rabbits. These evaluations were performed by the same performed in a Varian CP-3800 gas chromatograph (Varian, investigator, who was blinded to the electroencephalogram, Walnut Creek, CA) equipped with a ion trap mass detector and recorded at specific time points of the study (fig. 1): in (Varian Saturn 4000). The chromatographic column was a the awake animals, 20, 25, and 30 min after the beginning of Varian Factor Four ms (30 m ϫ 0.25 mm ϫ 0.25 ␮m). The each infusion rate, at extubation, and in the totally recovered column temperature was programmed to 100°C (1 min), animal. Animals were considered totally recovered when they 15°C/min until 300°C (10 min). The detection of propofol exhibited an alert stance and had regained ambulation and and thymol was conducted in Full Scan mode, and the quan- limb coordination. tification performed by monitoring the characteristic mass-to-

Fig. 1. Experimental design relative to the propofol infusion and data collection scheme. The electroencephalogram and clinical signs were recorded and analyzed at the moment of extubation (Rext), although no blood sampling was performed. Arrows with ϭ ϭ R moments of recording of electroencephalogram, clinical evaluation of anesthetic depth and arterial blood samples; R0 ϭ ϭ at baseline; R1,R2,R3 during each infusion 20, 25, and 30 min after the beginning of the respective infusion rate; Rrec in the totally recovered animal.

Anesthesiology 2011; 115:303–14 305 Silva et al. Anesthetic Depth Indexes in Propofol Anesthetized Rabbits

charge ratio (m/z) fragments of each molecule: for propofol, the applied to the PE. According to results by Li et al.,3 we used m/z used were 178 and 163, and for thymol, the m/z used were scale values of 1, 2, and 3 for the calculation of the CMSPE. 150 and 135. The retention times for each compound were, With the scale value of 1, the scalar PE is equal to the original respectively, 5.6 and 5.0 min. PE, which we designate as single scale PE. The composite index consists of the average of the PE values calculated with Indexes Derivation scales 1, 2, and 3. More details on the calculation of the CMSPE can be found in the literature.3 The IoC-View Graph 1.4 software stored the electroenceph- The IoC is a proprietary index that was developed to alographic data as binary files, which were further converted match the anesthetic depths observed in patients anesthe- to the MATLAB format to be processed offline using the ௡ tized with a variety of anesthetics. For its calculation, a fast MATLAB software (MathWorks, Natick, MA). Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021 Fourier transform is carried out after applying a hamming The IoC and the electroencephalogram suppression ratio, 4 function to a 3-s window of the electroencephalogram. The equivalent to the burst suppression ratio (BS), and the elec- tromyographic activity were automatically derived by the calculation of the IoC is based on the energies in different frequency bands (1–6, 6–12, 10–20, and 30–45 Hz) com- IoC-View Monitor. The monitor includes an electromyo- ␤ graphic activity filter that eliminates most of the potential bined with fuzzy logic rules. It also integrates the ratio interfering electromyographic activity before the derivation (frequency range between 11 and 42 Hz) during superficial of the index of consciousness and calculates the electromyo- anesthesia and the amount of suppression of the electroen- graphic activity, which is given as a percentage and shows the cephalogram (equivalent to the BS). In humans, decreasing values of IoC correspond to a gradual loss of consciousness energy of the electromyographic activity level in the fre- 4,5,7 quency band of 30–45 Hz.4 and a deepening of the depth of anesthesia. The monitor also displayed the signal quality index every second. All of the other parameters (MEF, SEF, AE, PE, and Data Analysis the traditional BS) were derived offline. The signal was first Prediction Probability Analysis. The capacity of the studied decreased eight times, resulting in a sampling frequency of indexes (MEF, SEF, AE, PE, CMSPE, IoC, electromyo- 128 Hz. The indexes were derived in epochs of 8 s, and only graphic activity, BSMEF, BSSEF, BSPE, and CMSPE) to epochs that showed a signal quality index value of 100 were detect the different anesthetic states, as reflected in the nu- considered. Visual inspection of every analyzed electroen- merical scale of anesthesia (table 1), was evaluated using pre- cephalogram fragment was also performed to ensure that no diction probability (Pk) statistics by correlating the parame- visible artifacts were included in the analysis. A Butterworth ter values during 64 s (8 measurements) at each of the defined bandpass filter of eighth order with cutoff frequencies of 0.5 study periods (fig. 1) with the numerical scale. Pk was calcu- 19 and 30 Hz was used, followed by removal of the mean value lated using a custom spreadsheet macro, PKMACRO. A of the signal to get out any threshold. Pk of 1 means that the parameter always decreases as the The spectral parameters were corrected for the presence of subject reaches deeper anesthetic states. Alternatively, a Pk burst suppression pattern, using the value of BS, according to value of 0.5 or less would mean that the indicator is useless the correction factor proposed by Rampil for human electro- for predicting the depth of anesthesia. Pk was calculated encephalogram studies15 and applied in rats.7,16 The same using pooled data from all animals. The Pk value and the correction factor was applied to PE, resulting in BSPE, cal- Jackknife SE are shown in Results. culated as follows: BSPE ϭ PE ϫ (1 Ϫ BS/100), where BS is Pharmacodynamic Modeling. The relation between the defined as the epoch length in which the electroencephalo- propofol plasma concentrations and the electroencephalo- gram voltage did not exceed 5 ␮V. The analysis was also graphic effect reflected in each studied index was modeled 20 performed with a higher limit of 10 ␮V, and the results of with a inhibitory sigmoid Emax model : each threshold applied were compared. c␥ ϭ Ϫ The AE and PE were computed according to published E E0 Emax ␥ ␥ (1) c ϩ EC algorithms.1,2 Briefly, the calculation of the AE depends on 50 three factors: the embedding dimension (m), the number of where E is the pharmacodynamic effect (i.e., the studied samples considered for each calculation (N), and the noise indexes of anesthetic depth), EC50 is the drug concentration ϭ ϭ threshold (r). In the current study, N 1,024, m 2, and that produces half-maximum effect, E0 is the effect at zero ϭ r 0.2 were selected for AE calculation, based on previous concentration, Emax is the maximum effect, C is the drug studies.4 For the PE calculation, the length of subvectors (m) plasma concentration effect, and the Hill exponent ␥ is a and the analyzed signal interval (length N) are main factors. measure of curve steepness. In the current study, we used m ϭ 3 and N ϭ 1,024. A more The pharmacodynamic parameters were estimated by detailed description of the AE and PE calculation can be population analysis using WinNonlin (WinNonlin Profes- found in published works.1,2,6,17 sional 5.0.1; Pharsight Corporation, Mountain View, CA). The CMSPE was recently introduced.3 It is based in the The goodness-of-fit for each index was compared using multiscale entropy previously proposed for sample entropy18 the values obtained for the coefficient of determination (R2).

Anesthesiology 2011; 115:303–14 306 Silva et al. PERIOPERATIVE MEDICINE

The indexes that showed the best goodness-of-fit accord- ing to these parameters were analyzed regarding the esti- ␥ mated pharmacodynamic parameters: Emax,E0,EC50, , and Ϫ Emax E0. Statistical Analysis. All of the indexes derived from the data resultant from the electroencephalographic process- ing were exported from MATLAB to GraphPad Prism (Version 5; GraphPad Software Inc, San Diego, CA) for statistical analysis. The Kolmogorov-Smirnov test was used to determine whether data sets were normally distributed. All tests were Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021 detailed with a statistical significance defined as P Ͻ 0.05. The Spearman rank correlation coefficient was calcu- lated between the propofol plasma concentration and the other studied parameters: clinical scale of anesthesia, MEF, SEF, AE, PE, CMSPE, BSMEF, BSSEF, BSPE, BSCMSPE, IoC, electromyographic activity, mean arte- rial pressure, and heart rate. The Wilcoxon matched pairs signed-rank test was used to compare the parameter values corrected with a BS limit cal- culated with 5 or 10 ␮V thresholds. Data are presented as mean Ϯ SD in Results.

Results Ϯ In the current study, six rabbits weighing 2.79 0.25 Kg Fig. 2. Electroencephalographic patterns with increasing were anesthetized with propofol in an infusion scheme that depths of anesthesia (DoA) and increasing propofol plasma included three consecutive infusion rates (70, 100, and 130 concentrations (Cp; ␮g/ml). Depth of anesthesia was evalu- Ϫ Ϫ mg ⅐ kg 1 ⅐ h 1). Every animal could be successfully anes- ated using clinical signs according to a scale from 0 (awake) thetized with the three rates, each maintained for 30 min. to 5 (very deep). Epochs with 8-s length are shown. The total duration of anesthesia (from induction to the stop- ping of the last infusion) was 109.5 Ϯ 6.8 min. All of the awake state to superficial depths of anesthesia, the electroenceph- animals recovered well from anesthesia, taking 21 Ϯ 6.8 min alogram shifted from a low-amplitude, high-frequency pattern to from the end of infusion to extubation and 28 Ϯ 11.2 min to higher amplitudes and lower frequencies. However, during me- recover limb ambulation. dium and profound depths with plasma concentrations of propofol Because of the randomization of the infusion schemes, the higher than 18 ␮g/ml, the electroencephalogram amplitude de- same infusions and duration did not produce the same creased progressively to values between Ϫ5 and 5 ␮V. In every Ϫ1 Ϫ1 propofol plasma concentrations. Different anesthetic depths animal, at the higher propofol infusion rate (130 mg ⅐ kg ⅐ h ) were also seen with the same infusions when performed in a the electroencephalogram was below these amplitude limits. different order. The electroencephalogram was recorded along with the The duration of the infusions was stipulated, based on IoC, electromyographic activity, and the electroencephalo- pharmacokinetic data of propofol in rabbits from Cockshott gram suppression ratio all derived automatically by the IoC- et al.,13 to produce a 10-min steady state in the end of each View Monitor. These parameters were included in the over- infusion rate. However, this steady state was achieved only in all analysis of the indexes of anesthetic depth. The other Ϫ Ϫ the lower infusion rate (70 mg ⅐ kg 1 ⅐ h 1). Thus, the indexes were derived offline from the raw electroencephalo- analysis of clinical and electroencephalographic data were gram and consisted of the MEF, SEF, AE, PE, CMSPE, BS, performed based on the plasma concentrations determined BSMEF, BSSEF, BSPE and BSCMSPE. Visual inspection of and not on the infusion rates administered. the raw electroencephalogram and analysis of the signal qual- With regard to the clinical depth of anesthesia, it was ity index given by the IoC-View Monitor were performed necessary to attribute an intermediate grade (4.5) between before derivation of the indexes of anesthetic depth to discard depths 4 and 5 for animals that did not have a perceptible poor quality electroencephalogram epochs. Four epochs corneal reflex but responded to the pedal withdrawal reflex were discarded in the extubation measurements of one rabbit (deep nociceptive reflex). and four epochs on the recovery measurements of another. With regard to the raw electroencephalographic findings, The values of the studied indexes at increasing propofol the electroencephalogram was gradually suppressed with in- plasma concentrations are shown in figure 3. To simplify this creasing plasma concentrations of propofol (fig. 2). From the representation, the propofol plasma concentrations were or-

Anesthesiology 2011; 115:303–14 307 Silva et al. Anesthetic Depth Indexes in Propofol Anesthetized Rabbits Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021

Fig. 3. Studied parameters at different ranges of propofol concentration (at 0, 1–10, 10–20, 20–30, 30–40, and 40–50 ␮g/ml plasma propofol) in the six rabbits. Clinical scale of anesthesia (A). Median edge frequency [MEF (Hz); red line] and burst suppression corrected MEF [BSMEF (Hz); blue line](B). Spectral edge frequency 95% [SEF (Hz); red line] and burst suppression corrected SEF [BSSEF (Hz); blue line](C). Approximate entropy (AE) (D). Permutation entropy (PE; red line) and burst suppression corrected PE (BSPE; blue line)(E). Composite multiscale permutation entropy (CMSPE; red line) and burst suppression corrected CMSPE (BSCMSPE; blue line)(F). Index of consciousness (IoC) (G). Electromyographic activity [EMG (%)] (H). The mean and upper SD (vertical bars) are shown. ganized in theoretical ranges: 0, 1–10, 10–20, 20–30, 30– graphic activity, BSSEF, and IoC than did the other param- 40, and 40–50 ␮g/ml. The real concentrations for each of eters, as shown in table 2. these ranges were, respectively, 0, 4.87 (1.98–7.22), 14.7 No significant differences between the Pk values of (10.70–18.99), 25.9 (20.15–29.00), 34.8 (30.48–38.42) the corrected indexes BSMEF, BSSEF, BSPE, and and 45.6 (41.70–48.85) ␮g/ml. BSCMSPE with a BS correction calculated with 5 or 10 The capacity of all these indexes to adequately assess the ␮V thresholds. clinical anesthetic depth was evaluated by Pk analysis, which The correlation of the parameters with the plasma con- revealed higher Pk values for BSPE, BSCMSPE, electromyo- centration of propofol is shown in table 3.

Anesthesiology 2011; 115:303–14 308 Silva et al. PERIOPERATIVE MEDICINE

Table 2. Prediction Probabilities Calculated Between Table 3. Correlation Coefficients (Spearman R) between the Subjective Anesthetic Depth Scale and the Studied the Studied Parameters and the Plasma Concentration Parameters of Propofol

Parameter Pk SE Parameter Spearman R with Cp AE 0.59 0.05 AE Ϫ0.04 BSCMSPE 0.83 0.02 BSCMSPE Ϫ0.89* BSMEF (Hz) 0.73 0.04 BSMEF (Hz) Ϫ0.62* BSPE 0.82 0.02 BSPE Ϫ0.87* BSSEF (Hz) 0.77 0.04 BSSEF (Hz) Ϫ0.71* CMSPE 0.53 0.06 CMSPE 0.00 EMG (%) 0.84 0.03 DoA 0.84* Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021 HR (beats/min) 0.71 0.05 EMG (%) Ϫ0.84* IoC 0.75 0.04 HR (beats/min) Ϫ0.53* MAP (mmHg) 0.74 0.03 IoC Ϫ0.50* MEF (Hz) 0.48 0.05 MAP (mmHg) Ϫ0.73* PE 0.52 0.06 MEF (Hz) 0.15 SEF (Hz) 0.54 0.05 PE 0.02 SEF (Hz) 0.01 Prediction probability (Pk) values calculated with pooled data from all animals (n ϭ 6). The standard error (SE) is also shown. Spearman rank correlation coefficient from pooled data of all AE ϭ approximate entropy; BS ϭ burst suppression; animals (n ϭ 6) is shown. BSCMSPE ϭ BS-corrected composite multiscale permutation * Significant at the 0.001 level. ϭ entropy; BSMEF (Hz) BS-corrected median edge frequency; AE ϭ approximate entropy; BS ϭ burst suppression; ϭ ϭ BSPE BS-corrected permutation entropy; BSSEF (Hz) BSCMSPE ϭ BS-corrected composite multiscale permutation ϭ BS-corrected spectral edge frequency; CMSPE composite multi- entropy; BSMEF (Hz) ϭ BS-corrected median edge frequency; ϭ ϭ scale PE; EMG (%) electromyographic activity; HR (beats/min) BSPE ϭ BS-corrected permutation entropy; BSSEF (Hz) ϭ ϭ ϭ heart rate; IoC index of consciousness; MAP (mmHg) mean BS-corrected spectral edge frequency; CMSPE ϭ composite multi- ϭ ϭ arterial pressure; MEF (Hz) median edge frequency; PE per- scale PE; Cp ϭ plasma concentration of propofol; DoA ϭ depth of ϭ mutation entropy; SEF (Hz) spectral edge frequency 95%. anesthesia; EMG (%) ϭ electromyographic activity; HR (beats/ min) ϭ heart rate; IoC ϭ index of consciousness; MAP (mmHg) ϭ mean arterial pressure; MEF (Hz) ϭ median edge frequency; PE ϭ The clinical scale of anesthesia increased monotoni- permutation entropy; SEF (Hz) ϭ spectral edge frequency 95%. cally with increasing propofol plasma concentrations until the range of 20–30 ␮g/ml (fig. 3A), revealing an increas- activity with regard to the estimated pharmacodynamic param- ing deepening of anesthesia. However, at higher concen- eters EC50 and E0 and, with regard to all of the pharmacody- trations, the depth of anesthesia stabilized. Mean arterial namic parameters, than did BSSEF (table 5). Figure 6 shows the pressure achieved low values at propofol plasma concen- pharmacodynamic fitting curves obtained for the studied pa- trations above 10 ␮g/ml (fig. 4) and showed a negative rameters that showed the higher coefficients of determination. correlation with propofol plasma concentrations (Ϫ0.73; Animals exhibited normal behavior at daily monitoring P Ͼ 0.001; table 3). Heart rate also decreased with in- for 2 weeks after the procedure. creasing propofol plasma concentrations, but the response was less consistent than that of mean arterial pressure (fig. Discussion 4), with a lower correlation coefficient (Ϫ0.53; P Ͼ In the current study, propofol anesthesia in a rabbit model 0.001; table 3). was used to compare the performance of different electro- The IoC-View Monitor was incapable of detecting the low voltage of the electroencephalogram during the higher concentrations in two rabbits. With the electroencephalo- gram completely suppressed (voltage between Ϫ5 and 5 ␮V) and propofol plasma concentrations of 35.3 Ϯ 4.8 ␮g/ml, the IoC showed values of 70.6 Ϯ 8.5 and an electroenceph- alogram suppression ratio (provided automatically by the monitor) of 0. In these animals, offline analysis with the BS algorithm revealed high values of BS (95.1 Ϯ 3.4%). The representation of two animal index values at the different study periods are shown in figure 5. With regard to pharmacodynamic modeling, the highest co- efficient of determination was shown by the electromyographic Fig. 4. Mean and SD of mean arterial pressure [MAP (mmHg)] 2 ϭ 2 ϭ activity (R 0.92), followed by BSSEF (R 0.82), and BSPE (A) and heart rate [HR (bpm)] (B) at different ranges of propo- 2 and BSCMSPE (both R ϭ 0.79) (table 4). However, BSPE fol concentration (at 0, 1–10, 10–20, 20–30, 30–40, and and BSCMSPE had less variability than did electromyographic 40–50 ␮g/ml plasma propofol) in the six rabbits.

Anesthesiology 2011; 115:303–14 309 Silva et al. Anesthetic Depth Indexes in Propofol Anesthetized Rabbits Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021

Fig. 5. Representation of the studied parameters in two rabbits: rabbit anesthetized with infusion rates order of 70, 100, 130 mg ⅐ kgϪ1 ⅐ hϪ1, respectively (A), and rabbit anesthetized with infusion rates order 130, 70, 100 mg ⅐ kgϪ1 ⅐ hϪ1 (B) at different study periods. The parameter mean values of eight epochs (64 s) are shown. AE ϭ approximate entropy; Awake ϭ in the awake state; (BS)CMSPE ϭ burst suppression corrected CMSPE; (BS)MEF ϭ burst suppression corrected MEF; (BS)PE ϭ burst suppression corrected PE; (BS)SEF ϭ burst suppression corrected SEF; CMSPE ϭ composite multiscale permutation entropy; Cp ϭ plasma propofol concentration; EMG ϭ electromyographic activity; IoC ϭ index of consciousness; MEF ϭ median edge frequency; PE ϭ permutation entropy; R1S1, R1S2, R1S3 ϭ samples 1, 2, and 3 of the first rate, each 20, 25, and 30 min after the beginning of that infusion rate; R2S1, R2S2, R2S3 ϭ samples 1, 2, and 3 of the second rate, each 20, 25, and 30 min after the beginn- ing of that infusion rate; R3S1, R3S2, R3S3 ϭ samples 1, 2, and 3 of the third rate, each 20, 25, and 30 min after the beginning of that infusion rate; Rec ϭ in the recovered animal; SEF ϭ spectral edge frequency 95%. encephalogram-derived parameters: MEF, SEF, AE, PE, than for BSPE, which also showed less variability with regard CMSPE, and IoC. The ability of these parameters to dis- to the estimated pharmacodynamic parameters. tinguish different depths of propofol anesthesia was as- Previous results with volatile anesthesia in rats showed a sessed using prediction probability analysis and pharma- better correlation of BSPE than AE, IoC, and spectral param- codynamic modeling. eters with the anesthetic dose.7 However, the signal analyzed The classic BS correction was applied to the MEF, SEF, PE, in that study was recorded intracranially, and thus is less and CMSPE, and the performance of the corrected parameters likely to be contaminated with movement artifacts and noise (BSMEF, BSSEF, BSPE, and BSCMSPE) was also compared. than are extracranial signals. The current results also show a Both BSPE and BSCMSPE showed better correlation co- superior capacity of BSPE and BSCMSPE to be used in efficients with propofol plasma concentrations than did the extracranial signals during propofol anesthesia, with better other studied indexes. They also showed better Pk than did concentration-effect relations and greater capacity to predict the other electroencephalogram-derived indexes. However, clinical signs than the other electroencephalogram-derived the electromyographic activity, as detected by the IoC-View indexes. Monitor, showed a higher Pk value than did BSPE and Despite the thin muscle layer between the skull and the BSCMSPE. The Pk value and the correlation with propofol skin on the rabbit head, electromyographic activity was cap- plasma concentrations were slightly higher for BSCMSPE tured by the IoC-View Monitor, showing better ability to

Anesthesiology 2011; 115:303–14 310 Silva et al. PERIOPERATIVE MEDICINE

Table 4. Goodness-of-fit for the Pharmacodynamic Modeling of the Studied Indexes

Index R2 MIN–MAX AE 0.59 (0.29–0.79) BSCMSPE 0.79 (0.52–0.91) BSMEF 0.75 (0.65–0.95) BSPE 0.79 (0.42–0.91) BSSEF 0.82 (0.76–0.94) CMSPE 0.57 (0.49–0.73) EMG 0.92 (0.79–0.99) IoC 0.68 (0.33–0.98) Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021 MEF 0.50 (0.10–0.96) PE 0.60 (0.42–0.96) SEF 0.61 (0.55–0.90)

The coefficient of determination (R2) values are shown. The min- imum and maximum values for the R2 obtained after analysis of each animal’s data (n ϭ 6) are shown as MIN–MAX. The best models are in italic. AE ϭ approximate entropy; BS ϭ burst suppression; BSCMSPE ϭ BS-corrected composite multiscale permutation entropy; BSMEF (Hz) ϭ BS-corrected median edge frequency; BSPE ϭ BS-corrected permutation entropy; BSSEF (Hz) ϭ BS- corrected spectral edge frequency; CMSPE ϭ composite multi- scale PE; EMG (%) ϭ electromyographic activity; IoC ϭ index of consciousness; MEF (Hz) ϭ median edge frequency; PE ϭ per- Fig. 6. Representation of effect (vertical axis) versus propofol mutation entropy; SEF (Hz) ϭ spectral edge frequency 95%. plasma concentration (horizontal axis) of the parameters that showed the greatest coefficient of determination R2: burst detect clinical signs of anesthesia than the electroencephalo- suppression-corrected median edge frequency (BSMEF) (A); gram-derived indexes. The cortical depression reflected in burst suppression-corrected spectral edge frequency 95% the electroencephalogram would be expected to have better (BSSEF) (B); burst suppression-corrected permutation en- prediction probability with propofol clinical effects.21 We tropy (BSPE) (C); burst suppression-corrected composite multiscale permutation entropy (BSCMSPE) (D); index of con- point out some possible explanations for the ability of the sciousness (IoC) (E); electromyographic activity [EMG 0%)] electromyographic activity to predict clinical signs: First, (F). Dots represent pooled data from all animals (n ϭ 6) there are some important limitations related to the use of observed and the black line represents the predicted effect 22 clinical scales of anesthetic depth. Although applied in the values using the inhibitory sigmoid Emax model. daily clinical practice, their correlation with real cerebral hypnotic drug effect is not well established because surrogate cortical depression mirrored by the electroencephalogram.26 observations, such as reflex responses and eyeball position, In addition, the band frequencies considered by the IoC- result from independent neurologic pathways and are not View Monitor as electromyographic activity are between 30 directly involved in the neurologic process of conscious- and 45 Hz.4 The electroencephalogram also shows activity in ness.23 In addition, propofol myorelaxant properties have these frequencies, called the ␥ band. One possibility is that been demonstrated,24,25 and a plausible explanation for the the electromyographic activity recorded by the monitor in good correlation of electromyographic activity with the clin- such a thin skull as the rabbit’s was incorporating informa- ical scale of anesthesia could be that the mechanism by which tion on the ␥ band, which decreases from the awake to the propofol induces myorelaxation could be closer to the mech- anesthetized state.27 This possibility was not assessed in the anism by which it produces the loss of autonomic reflexes current study and should be ruled out in future investigations evaluated by the clinical scale of anesthesia, rather than the using neuromuscular blocking agents to abolish muscle ac-

Table 5. Results from the Pharmacodynamic Modeling for the Indexes that Showed the Best Goodness-of-fit ␥ Emax EC50 E0 BSCMSPE 0.94 Ϯ 0.04 23.06 Ϯ 3.86 0.21 Ϯ 0.14 3.97 Ϯ 1.71 BSPE 0.87 Ϯ 0.04 22.76 Ϯ 3.91 0.19 Ϯ 0.13 3.85 Ϯ 1.66 BSSEF 30.98 Ϯ 1.94 10.10 Ϯ 9.42 0.25 Ϯ 10.8 0.82 Ϯ 0.41 EMG 99.02 Ϯ 5.30 6.58 Ϯ 0.74 Ϫ0.25 Ϯ 5.15 1.95 Ϯ 0.38

Data are estimate Ϯ standard deviation. BSCMSPE ϭ burst suppression-corrected composite multiscale permutation entropy; BSPE ϭ burst-suppression corrected permu- ϭ ϭ ϭ tation entropy; BSSEF burst suppression-corrected spectral edge frequency; EMG electromyographic activity; Emax maximum ϭ ϭ ␥ ϭ effect; EC50 concentration at 50% maximum effect; E0 baseline effect; Hill exponent.

Anesthesiology 2011; 115:303–14 311 Silva et al. Anesthetic Depth Indexes in Propofol Anesthetized Rabbits

tivity, and the ␥ band input in indexes of anesthetic depth in the clinical scale of anesthesia and different propofol should be investigated. plasma concentrations. This may be related to the appear- Another interesting finding of the current study was the ance of BS, which has been recognized as a difficulty with smaller correlation between propofol plasma concentration PE2 and CMSPE,3 which increases paradoxically during this and the clinical scale of anesthesia than between propofol pattern and has been attributed to the presence of noise dur- plasma concentration and the BSCMSPE and BSPE. This ing the suppression periods.6 The spectral parameters (MEF may be related to the limitations of clinical assessment re- and SEF) also lacked correlation with propofol plasma con- ferred to above, which is mainly based on movement re- centration and showed very low prediction probabilities with sponses that are merely spinal cord reflexes,28–30 and thus clinical scale of anesthesia. This is in accordance with the may not be related to the cortical effects of anesthetics.31 literature15 and was expected, particularly with the current Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021 These results highlight the need for the development of ob- experimental protocol, in which high levels of hypnosis were jective alternatives to the evaluation of clinical signs for as- achieved with increased BS values. When a BS correction was sessment of depth of anesthesia. However, these are the stan- applied to MEF, SEF, PE, and CMSPE, the parameters be- dard methods for evaluating depth of anesthesia in rabbits. came able to detect clinical changes and correlate with propo- Our results also confirm that knowledge of the drug concen- fol plasma concentrations, albeit with inferior performances tration may be an important complement for a more com- of BSMEF and BSSEF. Consequently, a BS component plete anesthetic depth monitoring. There are several factors seems indispensable when high anesthetic depths are ex- influencing the anesthetic depth in addition to anesthetic pected. The PE and CMSPE capacities for detecting nonlin- concentration, such as patient and surgery characteristics, ear properties of the electroencephalogram and their high existence and magnitude of the noxious stimulation, and resistance to artifacts2,6,17 may be in the origin of their better concomitant use of other drugs,32,33 that limit the use of performance when corrected for BS. anesthetic concentration as a unique means to assess anes- In raw electroencephalographic changes, propofol caused thetic depth in clinical practice. Nevertheless, the use of a gradual suppression on the electroencephalogram. The ef- propofol plasma concentrations has the advantage of being fects of low doses of propofol (from 4 to 15 ␮g/ml) were the an objective measure and providing a continuous range of characteristic low-frequency, high-amplitude patterns pres- concentration values, which is important when assessing the ent in shallow anesthetic depths or deep sedation.8 From this performance of electroencephalogram-derived indexes be- pattern, with increasing plasma concentrations of propofol, cause the electroencephalogram is also continuous.23 At this the electroencephalogram amplitude decreased until burst level, the use of laboratory animal models, such as the rabbit, suppression and complete isoelectricity. The burst suppres- may in the future allow comparing the real effect site con- sion pattern observed in this study was not the typical pattern centrations with the electroencephalographic responses by seen in rabbits anesthetized with volatiles.35 As described using more invasive techniques.34 previously,36 the bursts produced by propofol in rabbits were With regard to the results of concentration-effect rela- less distinct, consisting of spikes on slow waves with a tions, there was a clear better fit of the inhibitory sigmoid smoother wave form than the abrupt change in direct current

Emax model for electromyographic activity, BSSEF, BSPE, level seen in isoflurane burst suppression, without the ap- and BSCMSPE. However, BSPE and BSCMSPE showed pearance of the typical spindles seen with propofol anesthesia less variability on the estimated parameters and better model in humans. In general, high concentrations of the drug pro- adjustment than did electromyographic activity and BSSEF. duced a complete suppression, rather than a burst suppres- The incorporation of different scales that resulted in the sion pattern. In patients, burst suppression may appear with CMSPE seems to be slightly better than the single-scale PE. propofol plasma concentrations of approximately 6 ␮g/ml.37 The CMSPE was proposed recently.3 Multiscale entropy In this study, the higher concentrations achieved were much analysis has been used previously to describe the complexity higher (almost 10 times). Therefore, although the metabo- of time series over multiple scales, and the plots of entropy lization of propofol seems to be faster in rabbits than hu- values versus scales can be used to characterize the underlying mans,13 these are high plasma concentrations, which make physiologic process.18 The CMSPE is a unitary index that the appearance of an almost isoelectric signal not surprising. aims to reflect the anesthetic drug effect by integrating the As observed in previous studies in rabbits,38 the achievement information contained in the PE values derived at multiple of electroencephalographic silence was accompanied by con- scales. Compared with PE, it has the advantage of exploring siderable cardiovascular depression, with the occurrence of the dynamic characteristics inherent in brain electrical activ- hypotension at propofol plasma concentrations higher than ities on multiple scales.3 In a recent study, it showed better 10 ␮g/ml. Apart from the propofol effects on the brain elec- performance than did the single-scale PE in patients anesthe- trical activity, the prolonged systemic hypotension also could tized with sevoflurane.3 In this study, it was studied in a contribute to the electroencephalographic silence, as ob- different species with a different anesthetic drug, and the served previously in humans.39,40 Although no signs of brain performance of this index was close to that of the single-scale ischemic damage were found after anesthesia (in that all an- PE, with both showing a general incapacity to detect changes imals recovered well and showed a normal behavior 2 weeks

Anesthesiology 2011; 115:303–14 312 Silva et al. PERIOPERATIVE MEDICINE

after the procedure), the presence of hypotension during the infusion rate orders in different animals. Although the order recordings is an important limitation of the current study be- effects on pharmacokinetics are not an important interference cause the extent to which cardiovascular depression influenced because real plasma concentrations of propofol were used for electroencephalographic patterns and consequently its effects on statistical analysis, the possibility that the order interfered with concentration-effect relations cannot be established. the observed effects cannot be excluded. The commercial index IoC showed high variability in the In conclusion, using the most recent indexes of anesthetic prediction probability results, which may be attributable to depth, promising results were found with single-scale permu- its difficulty in detecting BS in two animals, resulting in a tation entropy and composite multiscale permutation en- paradoxical increase of the index in very high depths. This tropy during propofol anesthesia when a burst suppression situation has been reported in patients with other commer- correction was included in their calculation. Future re- Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021 cial monitors of anesthetic depth.41,42 The monitors seem to search should investigate the information present in elec- confound the characteristic low voltage of the completely tromyographic activity seems to be a good correlate with suppressed electroencephalogram with the awake state. Be- anesthetic depth. Burst suppression algorithms incorpo- cause in the current study this did not happen in every animal rated into monitors for intraoperative use should be accu- that achieved electroencephalogram isoelectricity, we may rately calibrated, to avoid dangerous paradoxical increases speculate that it is a situation dependent on the presence of during deep anesthesia. residual noise on the almost isoelectric signal. The IoC is a proprietary monitor, and its exact calculation algorithm is The authors gratefully acknowledge the contributions of Pedro 4 Oliveira, Ph.D. (Professor, Instituto de Cieˆncias Biome´dicas de Abel not published, but it is known that it has a BS component. Salazar, Universidade do Porto, Portugal), Nadja Bressan, M.Sc. In human patients, the depths of anesthesia achieved are not as (Ph.D. Student, Faculdade de Engenharia da Universidade do Porto, deep as those achieved in the current study. Thus, the paradox- Porto), and Ana Castro, B.Sc. (Ph.D. Student, Faculdade de Engen- haria da Universidade do Porto), who helped in the data analysis. ical increase of the indexes in case of isoelectricity may not be seen commonly in practice. However, the need for an accurate method for burst suppression detection should be emphasized References because paradoxical increases in this situation could result in 1. Bruhn J, Ro¨pcke H, Hoeft A: Approximate entropy as an dangerous overdosage. From another point of view, this IoC electroencephalographic measure of anesthetic drug effect during desflurane anesthesia. ANESTHESIOLOGY 2000; 92: difficulty could be explained by the fact that it is calibrated for 715–26 human electroencephalographic characteristics. Although there 2. Li X, Cui S, Voss LJ: Using permutation entropy to measure are differences in electroencephalogram amplitude between spe- the electroencephalographic effects of sevoflurane. ANESTHE- SIOLOGY 2008; 109:448–56 cies,43 the burst suppression limit capable of detecting the rab- 3. Li D, Li X, Liang Z, Voss LJ, Sleigh JW: Multiscale permutation bits’ electroencephalographic silence was similar to that used in entropy analysis of EEG recordings during sevoflurane anes- human patients. thesia. J Neural Eng 2010; 7:046010 With regard to translational research between animals 4. Jensen EW, Jospin M, Gambus PL, Vallverdu M, Caminal P: Validation of the Index of Consciousness (IoC) during seda- and humans, it was expected that the rabbit would provide a tion/analgesia for ultrasonographic endoscopy. Conf Proc high quality electroencephalographic signal because of the IEEE Eng Med Biol Soc 2008; 2008:5552–5 small muscle layer between the skull and the head skin and 5. Revuelta M, Paniagua P, Campos JM, Ferna´ndez JA, Martínez little electromyographic activity. However, the electromyo- A, Jospin M, Litvan H: Validation of the index of conscious- ness during sevoflurane and remifentanil anaesthesia: A com- graphic activity as recorded by the IoC-View Monitor parison with the and the cerebral state showed a very good dose-response relation, which may indi- index. Br J Anaesth 2008; 101:653–8 cate that electromyographic activity recorded in the rabbit 6. Olofsen E, Sleigh JW, Dahan A: Permutation entropy of the electroencephalogram: A measure of anaesthetic drug effect. skull during the awake state and in lighter anesthetic depth. Br J Anaesth 2008; 101:810–21 Nevertheless, this rabbit model provided good quality re- 7. Silva A, Cardoso-Cruz H, Silva F, Galhardo V, Antunes L: cordings with little artifact and low electromyographic activ- Comparison of anesthetic depth indexes based on thalamo- ity during deeper anesthesia and may be an interesting model cortical local field potentials in rats. ANESTHESIOLOGY 2010; 112:355–63 for future research in this area. 8. San-juan D, Chiappa KH, Cole AJ: Propofol and the electro- In the current study, plasma concentrations were used for encephalogram. Clin Neurophysiol 2010; 121:998–1006 the pharmacodynamic analysis instead of effect-site concen- 9. Vahle-Hinz C, Detsch O: What can in vivo electrophysiology trations because the plasma effect-site equilibration constant in animal models tell us about mechanisms of anaesthesia? Br J Anaesth 2002; 89:123–42 (ke0) was not determined. A first-order, effect-site model 44 10. Martín-Cancho MF, Lima JR, Luis L, Criso´stomo V, Lo´pez MA, could be used to find the ke0. However, the fact that the Ezquerra LJ, Carrasco-Jime´nez MS, Uso´n-Gargallo J: Bispec- expected (according to previous pharmacokinetic data13) tral index, spectral edge frequency 95% and median fre- steady state was not observed at the higher infusion rates may quency recorded at varying desflurane concentrations in pigs. Res Vet Sci 2006; 81:373–81 suggest nonlinear pharmacokinetics in this species, and applying 11. Flecknell P: Laboratory Animal Anaesthesia, 2nd edition. San the first-order, effect-site model could lead to inaccurate results. Diego, Academic Press, 1996, pp 76–86 Another limitation of the current study was the use of different 12. Thurmon J, Tranquilli W, Benson G, Lumb W, Jones E: Lumb

Anesthesiology 2011; 115:303–14 313 Silva et al. Anesthetic Depth Indexes in Propofol Anesthetized Rabbits

and Jones’ Veterinary Anesthesia, 3rd edition. Baltimore, the gamma frequency band and electromyogram signal. Br J Maryland, Williams & Wilkins, 1996, pp 721–7 Anaesth 2001; 86:50–8 13. Cockshott ID, Douglas EJ, Plummer GF, Simons PJ: The 28. Antognini JF: Hypothermia eliminates isoflurane require- pharmacokinetics of propofol in laboratory animals. Xenobi- ments at 20 degrees C. ANESTHESIOLOGY 1993; 78:1152–6 otica 1992; 22:369–75 29. Rampil IJ: Anesthetic potency is not altered after hypother- 14. Guitton J, Desage M, Lepape A, Degoute CS, Manchon M, mic spinal cord transection in rats. ANESTHESIOLOGY 1994; Brazier JL: Quantitation of propofol in whole blood by gas 80:606–10 chromatography-mass spectrometry. J Chromatogr B Biomed 30. Rampil IJ, Mason P, Singh H: Anesthetic potency (MAC) is Appl 1995; 669:358–65 independent of forebrain structures in the rat. ANESTHESIOL- 15. Rampil IJ: A primer for EEG signal processing in anesthesia. OGY 1993; 78:707–12 ANESTHESIOLOGY 1998; 89:980–1002 31. Sandin RH, Enlund G, Samuelsson P, Lennmarken C: Aware- 16. Ihmsen H, Schywalsky M, Plettke R, Priller M, Walz F, ness during anaesthesia: A prospective case study. Lancet Schwilden H: Concentration-effect relations, prediction 2000; 355:707–11 Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/115/2/303/255162/0000542-201108000-00017.pdf by guest on 30 September 2021 probabilities (Pk), and signal-to-noise ratios of different elec- 32. Bruhn J, Myles PS, Sneyd R, Struys MM: Depth of anaesthesia troencephalographic parameters during administration of monitoring: What’s available, what’s validated and what’s desflurane, isoflurane, and sevoflurane in rats. ANESTHESIOL- next? Br J Anaesth 2006; 97:85–94 OGY 2008; 108:276–85 33. Palanca BJ, Mashour GA, Avidan MS: Processed electroen- 17. Jordan D, Stockmanns G, Kochs EF, Pilge S, Schneider G: cephalogram in depth of anesthesia monitoring. Curr Opin Electroencephalographic order pattern analysis for the sep- Anaesthesiol 2009; 22:553–9 aration of consciousness and unconsciousness: An analysis of 34. Ludbrook GL, Upton RN, Grant C, Martinez A: Prolonged approximate entropy, permutation entropy, recurrence rate, dysequilibrium between blood and brain concentrations of and phase coupling of order recurrence plots. ANESTHESIOL- propofol during infusions in sheep. Acta Anaesthesiol Scand OGY 2008; 109:1014–22 1999; 43:206–11 18. Costa M, Goldberger AL, Peng CK: Multiscale entropy anal- 35. Venaˆncio C, Silva A, Souza A, Ferreira D, Antunes L: Com- ysis of complex physiologic time series. Phys Rev Lett 2002; parison of EEG derived anesthetic depth indexes in rabbits 89:068102 under fentanyl-isoflurane anesthesia. Paper presented at: An- nual Meeting of the American Society of Anesthesiologists, 19. Smith WD, Dutton RC, Smith NT: Measuring the perfor- October 18, 2009, New Orleans, Louisiana mance of anesthetic depth indicators. ANESTHESIOLOGY 1996; 84:38–51 36. Hartikainen K, Rorarius M, Ma¨kela¨ K, Yli-Hankala A, Ja¨ntti V: Propofol and isoflurane induced EEG burst suppression pat- 20. Holford NH, Sheiner LB: Kinetics of pharmacologic re- terns in rabbits. Acta Anaesthesiol Scand 1995; 39:814–8 sponse. Pharmacol Ther 1982; 16:143–66 37. Bruhn J, Bouillon TW, Shafer SL: Onset of propofol-induced 21. Jensen EW, Litvan H, Struys M, Martinez Vazquez P: Pitfalls burst suppression may be correctly detected as deepening of and challenges when assessing the depth of hypnosis during anaesthesia by approximate entropy but not by bispectral by clinical signs and electronic indices. index. Br J Anaesth 2001; 87:505–7 Acta Anaesthesiol Scand 2004; 48:1260–7 38. Wang M, Joshi S, Emerson RG: Comparison of intracarotid 22. Struys M, Versichelen L, Mortier E, Ryckaert D, De Mey JC, and intravenous propofol for electrocerebral silence in rab- De Deyne C, Rolly G: Comparison of spontaneous frontal bits. ANESTHESIOLOGY 2003; 99:904–10 EMG, EEG power spectrum and bispectral index to monitor 39. England MR: The changes in bispectral index during a hypo- propofol drug effect and emergence. Acta Anaesthesiol volemic cardiac arrest. ANESTHESIOLOGY 1999; 91:1947–9 Scand 1998; 42:628–36 40. Hayashida M, Chinzei M, Komatsu K, Yamamoto H, Tamai H, 23. Heyse B, Van Ooteghem B, Wyler B, Struys MM, Herregods L, Orii R, Hanaoka K, Murakami A: Detection of cerebral hypo- Vereecke H: Comparison of contemporary EEG derived perfusion with bispectral index during paediatric cardiac depth of anesthesia monitors with a 5 step validation pro- surgery. Br J Anaesth 2003; 90:694–8 cess. Acta Anaesthesiol Belg 2009; 60:19–33 41. Detsch O, Schneider G, Kochs E, Hapfelmeier G, Werner C: 24. Borgeat A, Dessibourg C, Rochani M, Suter PM: Sedation by Increasing isoflurane concentration may cause paradoxical propofol in tetanus–is it a muscular relaxant? Intensive Care increases in the EEG bispectral index in surgical patients. Br J Med 1991; 17:427–9 Anaesth 2000; 84:33–7 25. Ummenhofer WC, Kindler C, Tschale`r G, Hampl KF, Drewe 42. Hart SM, Buchannan CR, Sleigh JW: A failure of M-Entropy to J, Urwyler A: Propofol reduces succinylcholine induced in- correctly detect burst suppression leading to sevoflurane crease of masseter muscle tone. Can J Anaesth 1998; 45: overdosage. Anaesth Intensive Care 2009; 37:1002–4 417–23 43. Bullock TH: Biology of Brain Waves: Natural History and 26. Bonhomme V, Hans P: Muscle relaxation and depth of an- Evolution, Brainwaves and Mind: Recent Advances. Istanbul, aesthesia: Where is the missing link? Br J Anaesth 2007; Kjellberg, 2004, pp 3–13 99:456–60 44. McKay ID, Voss LJ, Sleigh JW, Barnard JP, Johannsen EK: 27. Sleigh JW, Steyn-Ross DA, Steyn-Ross ML, Williams ML, Smith Pharmacokinetic-pharmacodynamic modeling the hypnotic P: Comparison of changes in electroencephalographic mea- effect of sevoflurane using the spectral entropy of the elec- sures during induction of general anaesthesia: Influence of troencephalogram. Anesth Analg 2006; 102:91–7

Anesthesiology 2011; 115:303–14 314 Silva et al.