Anesth Analg 2006;102:1816-1822
© 2006 International Anesthesia Research Society
doi: 10.1213/01.ane.0000202523.94060.47
NEUROSURGICAL ANESTHESIA
Section Editor:
Adrian W. Gelb
Bolus Configuration Affects Dose Requirements of Intracarotid Propofol for Electroencephalographic Silence
Shailendra Joshi, MD*,
Mei Wang, MPH*,
Joshua J. Etu, BA*, and
John Pile-Spellman, MD
Departments of *Anesthesiology and Radiology and Neurosurgery, College of Physicians and Surgeons of Columbia University, New York, New York.
Address correspondence to Shailendra Joshi, MD, Irving Assistant Professor, Department of Anesthesiology, P&S P Box 46, College of Physicians and Surgeons of Columbia University, 630 West 168th Street, New York, NY 10032. Address e-mail to sj121{at}columbia.edu.
 |
Abstract
|
|---|
We hypothesized that an intracarotid bolus injection of propofol to produce electroencephalographic (EEG) silence would require a smaller dose of the drug compared with the continuous infusion of the drug. Furthermore, the bolus propofol dose will be a function of the bolus characteristics in each bolus (mass/volume). We compared the dose requirements of intracarotid propofol needed to maintain EEG silence when delivered as bolus injections to continuous infusions in rabbits. Subsequently, we compared whether four different bolus characteristics (concentration and volume) of propofol (0.33% x 0.1 mL, 0.33% x 0.3 mL, 1% x 0.1 mL, and 1% x 0.3 mL) affected the dose required to produce EEG silence. We found that the infusion rate of propofol required to sustain EEG silence was three-fold larger than the dose required by bolus injections, 22.8 ± 11.9 vs 6.2 ± 2.9 mL/h for infusion versus bolus, respectively (n = 7, P < 0.004). Furthermore, during bolus injection, the doses of propofol required to produce EEG silence were a direct function of the bolus volume and the mass of drug in each bolus, total dose = 3.6 + 29 x mg/bolus, n = 32, r = 0.85. For maximum regional effects of the bolus intracarotid drug injection, the bolus characteristics (volume and drug concentration) have to be optimized.
|
Introduction
|
|---|
Bolus intracarotid injections of anesthetic drugs are extensively used in diagnostic radiology during Wada testing for the location of neurological functions (1). Propofol is a highly protein-bound, lipid soluble, short-acting anesthetic drug that in recent years has been used for the Wada test (2–4). However, the theoretical factors that affect dose-response to bolus injection of propofol and other anesthetic drugs have not yet been fully investigated. In theory, bolus injections of drugs could transiently overwhelm the cerebral blood flow to minimize, or even eliminate, any protein binding. Thus, bolus intraarterial injections could produce disproportionately large free drug concentrations (5). Intraarterial bolus injection of benzodiazepines produces 5–25-fold larger cerebral arterial blood and brain tissue concentrations than those predicted by the kinetic models (6). Therefore, we hypothesized that bolus injections of a highly protein-bound, lipid-soluble drug, such as propofol, will be superior to continuous intracarotid infusion. Furthermore, because of the relatively small arterial dead space, the effects of bolus intracarotid injections of anesthetic will be a function of the volume of the injection and the concentration of the drug therein.
During the Wada test, the volume of injection is essentially determined by angiographic distribution of the drug and/or the occurrence of neurological symptoms. A considerable variation in dose-response and drug-distribution of intracarotid anesthetics has been reported during the Wada test (7). Thus, it is important to investigate the dose response to bolus intracarotid injections of anesthetic drugs. Nonanesthetic drugs are also frequently delivered via the intracarotid route for research, diagnostic, and therapeutic purposes (8,9). If bolus characteristics significantly affect the drug response, then there would also be a need to optimize these characteristics for maximum therapeutic results.
To test our hypothesis, we evaluated the effects of intracarotid bolus injection of propofol on electrocerebral activity in New Zealand White rabbits (10,11). These animals have a primate-like separation of the internal carotid and the external carotid arteries (12). We elected to use propofol because of the recent interest in this drug for Wada testing. In addition, intraarterial injection of propofol is fairly well tolerated by the vascular endothelium (13).
|
Methods
|
|---|
After approval of the protocol by the institution's animal care and use committee, the study was conducted on New Zealand White rabbits weighing approximately 3–4 lbs. The animals were given full access to food and water before the experiment. The animals were sedated with intramuscular ketamine (50 mg/kg) and anesthesia was maintained with IV infusion of propofol (Diprivan 1%; Astra-Zeneca Pharmaceuticals Inc. Wilmington, DE) 1–2 mL x kg–1 x h–1, fentanyl 1–2 µg x kg–1 x h–1 and vecuronium bromide (10–20 µg x kg–1 x h–1). This infusion rate of IV propofol is significantly smaller than the dose of the drug required to produce sustained EEG silence (24 ± mL/h) and provides a level of anesthesia that enables monitoring of electroencephalographic (EEG) activity over several hours (11). Hence, the background sedation is unlikely to affect the outcome of this study. A certain degree of background sedation is mandatory for animal experiments thus the minimal confounding effect, if any, of the background anesthetics cannot be avoided during the experiments. Our experimental preparation consisted of tracheostomy, femoral arterial cannulation, isolation of the right internal carotid artery, placement of bilateral EEG leads, and bilateral skull shaving to the inner table for the placement of laser Doppler probes to monitor cerebral blood flows. All surgical sites were infiltrated with 0.25% bupivacaine with epinephrine, and an esophageal temperature probe monitored the core temperature. The technical details of our preparation have been described in earlier publications (10,11). Hemodynamic and cerebral blood-flow variables for each drug challenge were evaluated at three stages of the experiment: (1) at baseline, (2) EEG silence with propofol, and (3) EEG recovery to baseline amplitude and morphology.
Protocol 1: Bolus versus Continuous Infusion
Before our initial studies we tested the capacity of the syringe drive Medfusion 2010iTM infusion (Medex Inc., Duluth, GA) pump to deliver the set volume given the resistance of the P-50 catheter, Figure 1, a–c. Both the "displayed" and the "measured" volumes correlated well with the volume rate selected for infusion in the range of 6–48 mL/h. We then set up a drug infusion system that was capable of delivering preset volume rates of infusion almost instantaneously. Between experiments, the syringe drive pump was continuously discharging externally through a parallel P-50 catheter. The dead space of the syringe pump tubing to the stop-cock was full of propofol during this period. Furthermore, as a result of bolus injection, the dead space of the catheter (0.2 mL) was also primed with the drug. When propofol infusion was required, a three-way stop-cock was turned toward the animal such that there was no delay in drug delivery.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 1. Bivariate scattergram (A–C) showing the performance of Medfusion 2010i infusion pump against the resistance imposed a P-50 arterial catheter. The pump was tested at four set volume infusion rates: 6, 12, 24, and 48 mL/h. Each flow rate was tested three times. There was a positive linear correlation between the set rate and the measured and displayed volumes.
|
|
In four preliminary studies for this protocol, we observed that the dose requirement of intraarterial propofol infusion was exceedingly large (
48 mL/h). At this intraarterial infusion rate, the recirculating concentrations of the drug were of concern to us. However, we observed that slower infusion rates (24 mL/h) were required to maintain EEG silence when the initial loading dose was delivered as a bolus, followed by a steady-state infusion. During the definitive study, therefore, we focused only on the maintenance dose of propofol required to sustain EEG silence. The initial loading dose was given in small boluses (0.1 mL) of 1% propofol given 10 s apart until EEG silence was achieved for 5–10 s. Subsequently, we maintained EEG silence with repeat bolus injections of 1% propofol (0.1 mL). The supplemental boluses were injected when there was any evidence of EEG activity (i.e., return of any low frequency EEG amplitude >5% of baseline or a spiking pattern). We first determined the maintenance dose of propofol required to sustain EEG silence for 5 min with bolus injections alone (i.e., the bolus dose). We allowed the preparation to recover for 30 min. We gave the initial loading doses as described above then started the continuous propofol infusion at twice the volume required for EEG silence with bolus injections (i.e., two times the bolus dose). Precautions were taken to insure that there was no delay in propofol delivery. Failure of dose was described as the return of progressively increasing EEG activity for at least 15 s. The infusion of the drug was immediately stopped once the failure was detected. The preparation was allowed to rest for another 30 min and a more rapid infusion rate (in the range of three to six times the bolus dose) was attempted depending on the response of the preparation. The experimental cycles were repeated to test a maximum of three infusion doses so as to minimize any time-related changes in the preparation.
Protocol 2: Bolus Configuration Studies
In preliminary studies, we determined the dose requirement of intracarotid propofol in eight rabbits ranging from a concentration of 0.25, 0.5, and 1% and volumes of 0.05, 0.1, 0.2, and 0.4 mL. We tested 12 bolus configurations in each animal that were aimed to produce 10 min of EEG silence. To determine the initial loading dose, bolus doses were delivered every 10 s until we achieved 5–10 s of EEG silence. Thereafter, maintenance doses were immediately delivered when EEG activity was recovered to more than 5% of baseline amplitude or when there was any spiking activity. The total dose was a sum of initial loading and maintenance doses. During the preliminary studies, we observed an initial loading dose failure in four of eight animals when 0.25% propofol was used with a bolus volume of 0.05 mL. The preliminary studies also suggested that multiple experiments were possible in the same animal because of a relatively rapid recovery of EEG activity after intracarotid propofol. The recovery of EEG activity, as well as the systemic changes (i.e., arterial blood pressure) were complete within 10 min of the last intracarotid injection.
Based on the preliminary experiments, we designed the study protocol to safely undertake repeat experiments on the same animal. We reduced the number of challenges to give additional recovery time to the preparation. We selected four doses that used two concentrations (0.33% and 1%) and two volumes (0.1 and 0.3 mL). We then generated three doses 0.33, 1.0, and 3 mg. Using balanced randomization each bolus configuration was injected twice either first, second, third, or fourth. We also reduced the period of EEG silence from 10 to 5 min to decrease the amount of intracarotid propofol. There was a 30-min rest period between each intracarotid drug challenge. The initial loading, maintenance, and total doses were determined as described above.
Data Analysis
The data are presented as mean ± sd. The hemodynamic and laser Doppler flow data were recorded at the three points of time (baseline, silence, and recovery). A P value of <0.0083 was considered significant among the four challenges (0.33% x 0.1 mL, 0.33% x 0.3 mL, 1% x 0.1 mL, and 1% x 0.3 mL). A P < 0.0167 was considered significant among the three stages of each challenge (baseline, drug, and recovery) that was determined by repeated measures analysis of variance with Bonferroni-Dunn test for multiple comparisons. Linear regression analysis was used to determine the relation between bolus dose and dose requirements as well as electrocerebral variables.
|
Results
|
|---|
Protocol 1
The experiments were conducted on seven animals with a mean weight of 3.0 ± 0.6 lbs. The initial propofol loading dose was similar (0.30 ± 0.19 and 0.27 ± 0.17 mg) in the bolus and infusion groups, respectively, P = 0.77, n = 7. The propofol doses that were required to sustain EEG silence significantly differed with the mode of drug delivery, 6.2 ± 2.9 versus 22.8 ± 11.9 mL/h for bolus versus infusion, respectively (n = 7, P < 0.004). The hemodynamic variables did not differ with either of the two modes of propofol delivery, Table 1.
View this table:
[in this window]
[in a new window]
|
Table 1. Changes in Physiological Variables with Different Bolus versus Continuous Infusions of Intracarotid Propofol
|
|
Protocol 2
The study was conducted on eight rabbits weighing 3.2 ± 0.6 lbs. Satisfactory data were obtained from all animals. Four doses were tested in each animal yielding 32 data points. The total dose of propofol (y, mg) required to produce 5 min of EEG silence was directly related to the amount of drug in each bolus (x), y = 3.6 + 29 x x, n = 32, r = 0.85. Both the drug concentration and the bolus volume affected the total dose requirements (Figure 2). There was a three-fold variation in the dose requirement between the smallest and the largest bolus dose (i.e., between 0.33% x 0.1 vs 1% x 0.3 mL from 4.3 ± 1 vs 12.3 ± 2.1 mg, respectively, P < 0.0001). The total recovery time (y, y [s] = 154 + 38 x bolus dose [x, mg/bolus], n = 32, r = 0.64), and the postdrug silence time (y, y [s] = 59 + 34 x bolus dose [x, mg/bolus], n = 32, r = 0.64) was a direct function of the bolus dose. However, postsilence recovery time (y, y [s] = 95 + 4.5 x bolus dose (x, mg/bolus), n = 32, r = 0.16) was not related to the bolus dose, Table 2.
View larger version (14K):
[in this window]
[in a new window]
|
Figure 2. Bivariate scattergram showing the effects of bolus dose on the total dose required to produce 5 min of electroencephalographic silence.
|
|
Baseline hemodynamic variables were similar across animals and across the four drug challenges, Table 3. There was no difference in the hemodynamic effects of the four challenges, Table 3. The mean arterial blood pressure decreased during EEG silence with all three challenges but was not different among the different bolus configurations. Laser Doppler bloodflow showed no consistent relationship with bolus doses of propofol, although flows were more rapid during silence with 1% drug concentrations, but this was not significant (Table 3).
|
Discussion
|
|---|
This study yielded two notable results. First, the mode of intracarotid drug delivery has a significant impact on the propofol dose required to maintain EEG silence. Compared with steady-state infusion, bolus injection of propofol was able to achieve EEG silence with a significantly smaller dose. Second, we observed that the intracarotid propofol dose required to produce 5 min of EEG silence was a direct function of the bolus dose. Both the concentration and volume of the drug bolus had a significant effect on the dose requirements of intracarotid propofol.
The fundamental kinetics of intracarotid bolus injection of drugs, despite extensive use in clinical neuroradiology, remains relatively uninvestigated. There are few tools available to track rapid changes in drug concentrations after bolus injections of intracarotid drugs. Our model that monitors the EEG response to intracarotid anesthetics provides valuable insights into the kinetics of intracarotid drugs particularly after bolus delivery. Despite factors such as acute tolerance to anesthetic effects, monitoring the dose of anesthetics required to achieve and maintain EEG activity at the threshold of silence provides a convenient indirect measure of tissue drug concentration that can be tracked in real time (14,15). To compare the effects of bolus versus continuous infusion of intracarotid propofol, we had to ensure that there was no delay in the delivery of the drug after the initial loading dose had been injected. We took several precautions to prevent any delay in drug injection. We assessed the pump functions against the resistance imposed by the P-50 arterial cannula. To prevent any delay due to the engagement of the clutch of the infusion pump, we primed the catheter and continuously discharged the infusion pump externally through a parallel cannula. Thus, during drug infusions, propofol was delivered as soon as the stopcock was turned on. It takes 1–2 s for an intraarterial injection to transit the cerebral tissue and reach the venous side. Thus, if the catheter is adequately primed and the pump is operational there will be negligible (1 s) time difference with bolus or infusion drug delivery to reach the cerebral tissue (16).
Jones et al. (6) investigated the kinetics of bolus injection of benzodiazepines and found that the tissue drug concentrations were 5–25-fold larger than those predicted by parallel in vitro studies. They offered several kinetic explanations for their results; however, they overlooked the possibility that they could have injected pure drugs to the brain. Their experiment involved bolus injection of 0.15–0.2 mL of drug volume into a rat's common carotid artery. The rat internal carotid probably irrigates less than 1 g of brain tissue and the internal carotid flow is probably 0.5 mL/min (5). Clearly, a 0.15-mL drug volume, if delivered in 1–3 s, would overwhelm the regional blood flow to deliver pure drug to the brain.
Our results show a very significant difference in dose requirements of intracarotid propofol when given as an infusion compared with bolus injection. Propofol is protein-bound (98%) and lipid-soluble with an octanol/water partition coefficient of 7000: 1. High protein binding and high lipid solubility make propofol particularly suited for bolus injection (17). In theory, transient delivery of bolus propofol would result in large free drug concentrations in the cerebral arterial blood because of the absence of any protein binding. In addition, high lipid solubility will result in a rapid uptake by the brain tissue. On the other hand, when steady-state infusions are used, a significant proportion of the drug will be bound to plasma proteins. The normal cerebral transit time in a rabbit is 1.8 s (16). Although we do not know the time-constant for propofol-protein binding, for benzodiazepine it is 25–50 ms (6). Thus, a significant degree of protein binding will occur during the cerebral transit of the drug that would reduce the free drug concentrations. Therefore, in theory, bolus injection of certain drugs, like propofol, that have high lipid solubility and protein binding, would be more effective than continuous intraarterial infusions.
In the second part of this study (protocol 2), we observed that the concentration of the drug and the volume of the bolus affected the initial loading, maintenance and total intracarotid propofol dose required to produce 5 min of EEG silence. It should be noted that in protocol 1, when we used the same volume of the bolus (0.1 mL) there was no difference between bolus or infusion challenges. However, in protocol 2, different volumes of initial loading doses were used, which resulted in a difference in the amount of initial loading dose needed to achieve EEG silence. This probably reflects the tissue propofol concentrations generated by different bolus volume to arterial dead space ratios. In theory, a smaller volume better provides the target dose if the time lapse between boluses is not excessive.
Critical to the understanding of bolus kinetics of the drugs is understanding the mechanics of the injection. Measurements in healthy rabbits indicate a cerebral blood volume of 1.93 mL/100g (16). Further assuming that the intracarotid injection irrigates 5 g of tissue, we estimate that the total blood volume in a unilateral internal carotid irrigation is <0.1 mL. However, in addition to cerebral tissue blood volume, there is a comparable amount of blood in extracranial and intracranial arteries. We, therefore, estimated that the total blood volume in our experimental conditions was between 0.2 to 0.3 mL. Thus, injection of 0.1 and 0.3 mL used in our studies would have largely been contained in the arterial dead space and would have certainly delivered relatively undiluted drug to the brain. In clinical settings, drugs are regularly administered to humans at the rates of 1–10 mL/s during cerebral angiography. Such a rate of injection is sufficient to transiently overwhelm carotid blood flow and deliver relatively pure drug to the brain. In concurrent studies in our laboratory, we are assessing the possibility of videoimaging drug concentrations in the brain after intracarotid delivery. Figure 3, a–c shows transit of 0.1 mL of a propofol bolus through the arteries and the veins. The pictures were taken 1 s before, during, and 1 s after the completion of the injection. These pictures show that even a 0.1 mL volume transiently, although completely, displaces the blood in the cerebral arteries. Such images would suggest that protein binding would be relatively unimportant with bolus drug delivery.
View larger version (81K):
[in this window]
[in a new window]
|
Figure 3. Videomicroscopy images taken 1 s before (3-a), during (3-b), and 1 s after (3-c) the injection of 0.1 mL bolus of propofol. A single artery lies transversely at the bottom of the picture, and multiple veins drain the arterial irrigation. Under green illumination, blood appears to be black, whereas propofol appears white. Note that the bolus of 0.1 mL completely displaces the blood contained in the artery and most of the veins. This displacement is exceedingly transient. The bolus of the drug is washed out within 1 s of intracarotid injection.
|
|
The primary advantage of bolus injection of drugs is the fairly consistent regional distribution of the drug (18). Bolus injections also deliver a consistently large concentration and avoid regional variations in drug concentrations due to streaming (19–21). However, the disadvantage of bolus delivery is the limited uptake through the blood-brain barrier during the short time it transits the brain. In theory, if the concentration of drug exceeds the maximum uptake by the brain in that transit period, then the extra amount of drug will simply overflow to the venous side. This would increase systemic side effects and decrease regional selectivity. Delivering bolus drugs in amounts that exceed brain uptake and in volumes that exceed cerebral arterial blood volume would decrease the efficiency of bolus injection. Thus, ideally, for maximum regional delivery, the bolus of the drug has to be tailored to the kinetic factors that determine brain uptake; the anatomical factors, such as arterial dead space, that determine bolus volume; and the flow factors that determine effective concentrations and transit times.
Our results suggest that a bolus dose of intracarotid propofol is more effective in causing EEG silence than continuous infusion. This study also shows that the configuration of the bolus has a significant effect on the dose requirement of intracarotid drugs.
Thanks are due to Richard Arrington, BA, Cert. AT, Manager of Anesthesiology Services, for helping with technical aspects of the experiments.
|
Footnotes
|
|---|
Accepted for publication December 16, 2005.
This work was supported in part by National Institutes of Health Grant K08-00698 and the Irving Clinical Research Career Award from the Irving Center for Clinical Research (to SJ).
This work was presented in part at the annual meeting of the Society of Neurosurgical Anesthesia and Critical Care in Las Vegas, Nevada on October 21, 2004.
|
References
|
|---|
- Wada J, Rassmussen T. Intracarotid injection of sodium amytal for the lateralization of cerebral speech dominance: experimental and clinical observations. J Neurosurg 1960;17:266–82.[ISI]
- Bazin JE, Picard P, Gabrillargues J, Dordain M. Propofol administered via the carotid artery to achieve a Wada test. Can J Anaesth 1998;45:707–8.[Free Full Text]
- Silva TM, Hernandez-Fustes OJ, Bueno ML, et al. The Wada test with propofol in a patient with epilepsy. Ar Qneuropsiquiatr 2000;58:348–50.
- Takayama M, Miyamoto S, Ikeda A, et al. Intracarotid propofol test for speech and memory dominance in man. Neurology 2004;63:510–5.[Abstract/Free Full Text]
- Joshi S, Pile-Spellman J, Young WL. Does papaverine cause endothelial injury in clinically relevant doses [letter]? Am J Neuroradiol 1997;18:1797–800.[Medline]
- Jones DR, Hall SD, Jackson EK, et al. Brain uptake of benzodiazepines: effects of lipophilicity and plasma protein binding. J Pharmacol Exp Ther 1988;245:816–22.[Abstract/Free Full Text]
- Trenerry MR, Loring DW. Intracarotid amobarbital procedure. The Wada test. Neuroimaging Clin N Am 1995;5:721–8.[Medline]
- Pardridge WM. Drug delivery to the brain. J Cereb Blood Flow Metab 1997;17:713–31.[ISI][Medline]
- Jolliet-Riant P, Tillement JP. Drug transfer across the blood-brain barrier and improvement of brain delivery. Fundam Clin Pharmacol 1999;13:16–26.[Medline]
- Wang M, Joshi S, Emerson RG. Comparison of intracarotid and intravenous propofol for electrocerebral silence in rabbits. Anesthesiology 2003;99:904–10.[ISI][Medline]
- Joshi S, Wang M, Nishanian EV, Emerson RG. Electrocerebral silence by intracarotid anesthetics does not affect early hyperemia after transient cerebral ischemia in rabbits. Anesth Analg 2004;98:1454–9.[Abstract/Free Full Text]
- Joshi S, Wang M, Hartl R. Retinal discoloration test. J Cereb Blood Flow Metab 2004;24:305–8.[ISI][Medline]
- MacPherson RD, Rasiah RL, McLeod LJ. Intraarterial propofol is not directly toxic to vascular endothelium. Anesthesiology 1992;76:967–71.[ISI][Medline]
- Ihmsen H, Tzabazis A, Schywalsky M, Schwilden H. Propofol in rats: testing for nonlinear pharmacokinetics and modelling acute tolerance to EEG effects. Eur J Anaesthesiol 2002;19:177–88.[ISI][Medline]
- Larsson JE, Wahlstrom G. Age-dependent development of acute tolerance to propofol and its distribution in a pharmacokinetic compartment-independent rat model. Acta Anaesthesiol Scand 1996;40:734–40.[ISI][Medline]
- Cenic A, Nabavi DG, Craen RA, et al. Dynamic CT measurement of cerebral blood flow: a validation study. Am J Neuroradiol 1999;20:63–73.[Abstract/Free Full Text]
- Fragen RJ, Avram MJ, et al. Nonopioid intravenous anesthetics. Clin Anesth 1992; 385–412.
- Saris SC, Shook DR, Blasberg RG, et al. Carotid artery mixing with diastole-phased pulsed drug infusion. J Neurosurg 1987;67:721–5.[ISI][Medline]
- Saris SC, Wright DC, Oldfield EH, Blasberg RG. Intravascular streaming and variable delivery to brain following carotid artery infusions in the Sprague-Dawley rat. J Cereb Blood Flow Metab 1988;8:116–20.[ISI][Medline]
- Lutz RJ, Dedrick RL, Boretos JW, et al. Mixing studies during intracarotid artery infusions in an in vitro model. J Neurosurg 1986;64:277–83.[ISI][Medline]
- Dedrick RL. Arterial drug infusion: pharmacokinetic problems and pitfalls. J Natl Canc Inst 1988;80:84–9.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
M. Wang and S. Joshi
Electrocerebral Silence After Intracarotid Propofol Injection Is a Function of Transit Time
Anesth. Analg.,
June 1, 2007;
104(6):
1498 - 1503.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
