Anesth Analg 2004;98:178-184
© 2004 International Anesthesia Research Society
PAIN MEDICINE
Amiodarone Decreases Heat, Cold, and Mechanical Hyperalgesia in a Rat Model of Neuropathic Pain
Sukdeb Datta, MD, DABPM,
Taruna Waghray, MD,
Maria Torres, MD, and
Silvio Glusman, MD, PhD
From the Department of Anesthesiology and Pain Management, Cook County Hospital, Chicago, Illinois
Address correspondence and reprint requests: Sukdeb Datta, MD, DABPM, Assistant Professor, Department of Anesthesiology, University of Cincinnati College of Medicine, P.O. Box 670531, Cincinnati, OH 45267–0531. Address email to sukdeb.datta{at}uc.edu
 |
Abstract
|
|---|
Lidocaine is effective in controlling ventricular dysrhythmia and neuropathic pain. Amiodarone, like lidocaine, has sodium channel blocking properties. In the present study we explore whether amiodarone has a similar effect as lidocaine on the heat, cold, and mechanical hyperalgesia seen in the rat model of neuropathic pain. Ten male Sprague-Dawley rats were anesthetized. Four loose ligatures were placed on the sciatic nerve of the right hindpaw. A sham operation was performed on the contralateral hindpaw (control). Heat hyperalgesia was determined by comparing each paw withdrawal latency to heat stimulation (radiant heat source, 50°C). Cold hyperalgesia was assessed with acetone application. Mechanical hyperalgesia was determined by comparing the mechanical threshold in the ligated and control hind paws using calibrated von Frey filaments. Amiodarone was intraperitoneally administered at doses of 1, 5, 10, 20, 50, and 100 mg/kg after the development of hyperalgesia. The animals were tested for hyperalgesia before and 1, 3, and 24 h after the administration of a single dose of amiodarone. Intrathecal catheters were implanted in 5 new rats, and amiodarone 5 mg/kg was injected. Testing for heat, mechanical, and cold hyperalgesia was performed similarly in the intrathecal amiodarone administration group. Amiodarone produces statistically significant decreases of heat, cold, and mechanical hyperalgesia after intraperitoneal administration. Results are statistically significant at 10 mg/kg (heat hyperalgesia), 20 mg/kg (mechanical hyperalgesia), and 100 mg/kg (cold hyperalgesia) intraperitoneally. Hyperalgesia returns 24 h after a dose. The intrathecal administration of amiodarone produces a nonstatistically significant reduction of hyperalgesia. Amiodarone seems to have a similar effect as lidocaine on the hyperalgesia seen in the rat model of neuropathic pain. As the half-life of amiodarone is significantly longer that that of lidocaine (mean, 53 days versus 90 min) in humans, it may have the potential to provide a longer lasting (and perhaps more effective) effect than lidocaine on neuropathic pain states.
IMPLICATIONS: Amiodarone was found to produce a statistically significant decrease in heat, cold, and mechanical hyperalgesia in a rat model of neuropathic pain after intraperitoneal injection. Considering its long half-life in humans, amiodarone has the potential to provide long lasting pain relief in neuropathic pain states.
|
Introduction
|
|---|
The management of patients with neuropathic pain is challenging. The local anesthetic lidocaine is well recognized as both a diagnostic and therapeutic drug for neuropathic pain (1–5). Repetitive C-fiber stimulation induces a state of facilitated processing of sensory information in the dorsal horn, whereas chronic nerve compression gives rise to a hyperalgesic state, characterized by spontaneous neuronal activity generated by voltage sensitive sodium channels, as well as the still not well understood state of central facilitation. Possible mechanisms of action of lidocaine may include effects on impulse generation in injured nerves by blocking these sodium channels (6), inhibiting dorsal horn neural transmission (7), modulating the levels of endogenous peptides in the spinal cord (cholecystokinin, neuropeptide Y), and modifying cerebral pain perception (8).
Amiodarone is a structural analog of thyroxine, originally developed in the 1960s as an antianginal coronary vasodilator (9). The pharmacologic and electrophysiologic properties of this drug are extraordinarily complex, and our understanding of its mechanism of action is still evolving. It has been described as the prototype Vaughn-Williams class III drug, but in fact, it exhibits the electrophysiologic characteristics of all four classes (9). It is useful to consider the effects of acute versus chronic administration. When administered acutely, the initial effects involve antisympathetic and calcium channel blocking properties (10); whether amiodarone acutely prolongs the action potential is controversial (11). The most prominent effect of chronic amiodarone administration is the class III action of prolonged action potential duration (10,11).
Our experimental working hypothesis was that because amiodarone shares with lidocaine the same sodium channel blocking properties, it might also be effective in the management of neuropathic pain. In the present study, we sought to examine whether amiodarone could produce an effect similar to lidocaine in reducing the hyperalgesia to heat, mechanical, and cold stimuli in a rat model of neuropathic pain induced by chronic constriction of the sciatic nerve.
|
Methods
|
|---|
The study was performed under a protocol approved by the Institutional Animal Care and Utilization Committee of the Cook County Hospital. Ten male Sprague-Dawley rats (Harlan Industries, Indianapolis, IN) weighing 224–250 g at the beginning of the experiment were used. Laboratory chow and water were available to the rats at all times. Rats were individually housed and maintained on a 12-h light and 12-h dark cycle. The model of neuropathic pain previously described by Bennett and Xie (12) was used. In brief, animals were anesthetized with a mixture of ketamine and xylazine given IM. After an incision was made in the skin, the biceps femoris of each leg was dissected bluntly at the midthigh level to expose the sciatic nerve. The nerve was carefully mobilized, with care taken to avoid undue stretching. Four 4-0 chromic sutures were tied with a square knot around the right sciatic nerve, approximately 2 mm apart, with enough tension to produce a single brief twitch of the innervated muscles. The left sciatic nerve was mobilized but not ligated. The muscles were reapproximated and the skin closed with 3-0 silk sutures. Animals were allowed to recover for 7–10 days before nociceptive threshold testing. After sciatic nerve ligation, animals were maintained individually in clear plastic cages with solid floors covered with 3–6 cm sawdust. Animals appropriately prepared showed mild flexion of the ligated leg with slight curling of the digits.
For the assessment of heat, mechanical, and cold hyperalgesia, the rats were placed on a metal mesh covered with a plastic dome at room temperature, and they were allowed to habituate until exploratory behavior diminished. Heat nociceptive threshold was measured with a device similar to that used by Hargreaves et al. (13). A radiant heat source (halogen projector lamp, 50 W, 8 V) with a 4-mm aperture contained in a movable holder beneath the wire mesh floor was used. Voltage to the lamp was controlled with a constant voltage regulator. The voltage of the lamp was calibrated to produce a mean response latency of 10 s in the untreated animal. Before each latency test, the lamp was positioned directly under the midpalmar aspect of the hindpaw, with low voltage illumination used to target the beam. When the radiant heat source was turned on it activated a timing circuit. Time to brisk withdrawal of the hindpaw was measured to the nearest 0.1 s. The cutoff time to prevent tissue injury in absence of a response was 20 s, and this latency was assigned to nonresponders. The data for each hindpaw at each particular testing period were recorded and the mean Paw Withdrawal Latency (PWL) was determined from the average of 5 separate trials, taken at 10- to 15-min intervals. A difference score between PWL of each foot was calculated by subtracting the mean withdrawal score of the left (control) hindpaw from the mean PWL of the right (sciatic nerve ligated) hindpaw divided by the PWL of the control hindpaw. A negative difference score indicated development of a faster PWL on the sciatic nerve ligated side and therefore gave an index of heat hyperalgesia. Cold hyperalgesia was measured as the number of foot withdrawal responses after application of cold stimuli (acetone) to the plantar surface of the paw (14). Acetone (0.1 mL) was applied to each hindpaw through a polyethylene (PE) 10 plastic tubing connected to a 1-mL syringe. A brisk foot withdrawal response after the spread of acetone over the plantar surface of the paw was considered a sign of cold hyperalgesia. The testing was started with the paw contralateral to the nerve injury and repeated 5 times for both hindpaw with an interval of approximately 2 min between each test. A difference score between paw withdrawals of each foot was calculated by subtracting the mean withdrawal responses of the right (sciatic nerve ligated) hind paw from left (control) hindpaw. A positive difference score indicated development of cold hyperalgesia. The threshold for mechanical hyperalgesia was measured by using a series of calibrated von Frey hairs (Semmes-Weinstein, Stoelting, IL) (15,16). The plantar surface of the control (left) hindpaw was touched with different von Frey hairs with a bending force of 0.217–12.5 g. Ten trials were done on the left paw. If the rat responded to the stimulation by withdrawing the paw 2 times out of the 10 trials then it was taken as a threshold. If the rat responded to the stimulation by withdrawing the paw more than two times, the next weaker hair was used until the threshold was found. To avoid excessive stimulation, the testing was started in the following sessions with the weakest hair that had elicited withdrawal responses in the previous session. Next, the same hair was used to test the right (sciatic nerve ligated) paw. Mechanical hyperalgesia was determined by comparing the number of withdrawal responses out of 10 trials in the right paw versus the number of withdrawal responses in the control paw.
Intrathecal catheter insertion was performed in five new rats. The rats were anesthetized with a mixture of ketamine and xylazine. An incision was made in the atlanto-occipital membrane, and a sterile PE 10 catheter was inserted so that the tip of the catheter reached L2-3 spinal segments (7.5 cm). The catheter was exteriorized at the incision site and the animals were allowed to recover for 5–6 days. Thereafter, testing in this group was performed similarly for determination of heat, cold, and mechanical hyperalgesia.
Amiodarone hydrochloride (Sigma Corporation, St. Louis, MO) was dissolved in sterile physiological saline and stored at 4°C before use. The rats were allowed to develop changes consistent with the development of heat, mechanical, and cold hyperalgesia. Once the rats developed hyperalgesia (10 rats, approximately 18 days postsurgery), amiodarone was administered intraperitoneally (1, 5, 10, 20, 50, and 100 mg/kg). The animals were tested for hyperalgesia before (control) and 1 and 3 h after administration of a single dose of amiodarone. In the 50 and 100 mg/kg group, rats were also tested at 24 h to determine whether the effects of amiodarone persisted past its elimination t1/2 and to see whether cumulative effects were present. A single dose of the drug was administered in each session and incremental doses were used in subsequent sessions. A log dose response curve was constructed and the ED50 of the dose decreasing thermal hyperalgesia was determined. Intrathecal amiodarone (5 mg/kg) was administered in 5 rats. The animals were tested for hyperalgesia before (control) and 1 and 3 h after administration of this dose of amiodarone.
Analysis of variance for repeated measures followed by Student’s t-test with Bonferroni’s correction when appropriate was used for statistical analysis of the behavioral symptoms over time and between the treatment groups and the control groups. Results were expressed as mean ± SEM and considered significant at P 
0.05.
|
Results
|
|---|
Significant heat, mechanical, and cold hyperalgesia developed approximately 18 days postsurgery and remained relatively stable over the observation period. Heat, cold, and mechanical hyperalgesia were maximum between 24 and 28 days after the operation and slightly decreased afterwards. However, this change was not statistically significant. The animals did not show any autotomy, and there were no apparent signs of distress, altered social behavior, or loss of appetite. The animals gained weight appropriately throughout the study period.
Heat Hyperalgesia
In the heat nociceptive test (Fig. 1), a decrease in PWL after the intraperitoneal amiodarone administration of amiodarone was statistically significant at 10 mg/kg (P = 0.004), 20 mg/kg (P = 0.02, from -29.19 ± -6.61% (SEM) to -8.41 ± -4.19% at 1 h), 50 mg/kg, and 100 mg/kg (P < 0.0001 for both the 1 h and 3 h postamiodarone administration). Although the decrease in PWL was present at 24 h after the administration of 50 and 100 mg/kg of amiodarone, the results were not statistically significant. Thus, a return to the hyperalgesic state was seen at 24 h in both the 50 and 100 mg/kg amiodarone groups. The 5 mg/kg intraperitoneal dose seemed to increase the heat hyperalgesia. We believe these latter observations may have been attributable to normal variations.
Cold Hyperalgesia
A statistically significant decrease in cold hyperalgesia was demonstrated with the intraperitoneal amiodarone administration at 100 mg/kg (P = 0.0001) for both the 1 h and 3 h postamiodarone administration (Fig. 2). Although cold hyperalgesia was decreased at 24 h after amiodarone administration of 50 and 100 mg/kg of amiodarone compared with the control group, results did not reach statistical significance. The smallest dose of intraperitoneal amiodarone administered (1 mg/kg) seemed to increase the cold hyperalgesia at both 1 h and 3 h postamiodarone administration. We believe that these results may be explained by normal variations.
Mechanical Hyperalgesia
A statistically significant decrease in mechanical hyperalgesia was observed after the intraperitoneal administration of amiodarone at 20 mg/kg (P = 0.043, at 1 h postamiodarone administration, not significant at 3 h postamiodarone administration), 50 mg/kg (P = 0.139, not significant values at 1 h and 3 h postamiodarone administration) and 100 mg/kg (P = 0.012, 1 h and 3 h postamiodarone administration) (Fig. 3). Although mechanical hyperalgesia was decreased at 24 h after amiodarone administration of 50 and 100 mg/kg of amiodarone compared with the control group, results did not reach statistical significance. These observations may also be explained by normal variations. ED50 for decreasing all 3 forms of hyperalgesia was determined to be between 40 to 60 mg of amiodarone.
View larger version (21K):
[in this window]
[in a new window]
|
Figure 3. Response to intraperitoneal amiodarone (mechanical hyperalgesia). Positive difference score indicates development of mechanical hyperalgesia in the sciatic nerve ligated leg. Values are mean ± SEM. *P 0.05.
|
|
Intrathecal administration of 5 mg/kg amiodarone in 5 rats revealed a decrease in heat, mechanical, and cold hyperalgesia in both the 1 h and 3 h values. However, results did not reach statistical significance (Fig. 4 and 5). No behavioral changes or motor impairment were observed after the intraperitoneal or intrathecal administration of amiodarone.
View larger version (16K):
[in this window]
[in a new window]
|
Figure 4. Response to intrathecal amiodarone administration (heat hyperalgesia test). Negative paw withdrawal latency (PWL) % indicates heat hyperalgesia in the sciatic nerve ligated leg. Values are mean ± SEM.
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Figure 5. Response to intrathecal amiodarone administration (mechanical and cold hyperalgesia). Positive difference score indicates development of mechanical and cold hyperalgesia in the sciatic nerve ligated leg. Values are mean ± SEM.
|
|
|
Discussion
|
|---|
Amiodarone is a "broad spectrum" antiarrhythmic drug with multiple and complex electrophysiological effects. It is a sodium channel blocker that thus reduces the rate of membrane depolarization and impulse conduction. It also noncompetitively inhibits 
- and ß-adrenergic receptors and possesses both vagolytic and calcium channel blocking properties (17). Du et al. (18) demonstrated a sympatholytic action of amiodarone in rat hearts by showing it induces partial norepinephrine (NE) depletion in the heart by interfering with vesicular NE storage and enhancing intraneuronal NE metabolism. Haikerwal et al. (19) also demonstrated similar presynaptic "reserpine like" sympatholytic action of amiodarone and its metabolite desethylamiodarone. Reserpine acts on the storage vesicle in the presynaptic nerve terminals to inhibit the vesicular uptake of monoamines, principally NE and dopamine, leading to axoplasmic monoamines being metabolized and a depletion of sympathetic monoamine stores. When amiodarone was administered intravascularly both in vivo and in vitro, there was an increase in overflow of the intraneuronal metabolite of NE, dihydroxyphenylglycol (DHPG), a decrease in tissue NE content together with an increase in tissue DHPG level, and a decrease in neuronal NE release in response to sympathetic nerve stimulation (18). Amiodarone also inhibits production of certain cytokines, such as tumor necrosis factor- (20) and interleukin-6 (21), involved in production of pain. In view of these actions, we hypothesized that amiodarone could become a useful drug in the management of neuropathic pain states.
Our experimental results showed an intraperitoneal dose of amiodarone of 100 mg/kg produces a statistically significant decrease on all three types of hyperalgesia studied. Heat hyperalgesia appeared to be most easily affected after intraperitoneal amiodarone, followed by mechanical and cold hyperalgesia (right shift in the dose response curve) with an ED50 between 40 mg/kg and 60 mg/kg of amiodarone. Also, a smaller dose of amiodarone is required to see a statistically significant decrease of heat hyperalgesia (10 mg/kg) than of mechanical (20 mg/kg) or cold hyperalgesia (100 mg/kg).
The site of action of amiodarone may be peripheral, as amiodarone has very poor penetration in the brain (22,23), and the intrathecal administration of amiodarone produced a nonstatistically significant decrease in all three forms of hyperalgesia. Abdi et al. (24) used electrophysiologic measurements in a rat model of neuropathic pain and suggested that the mechanism of action of a similar drug, lidocaine, may be peripheral. However, other workers have suggested a more central site of action of lidocaine. The fact that the C-fiber evoked polysynaptic reflex generated by sural nerve stimulation is suppressed by lidocaine at doses that do not block impulse conduction (5) provides evidence that this drug also has central inhibitory effects. Bach et al. (1) showed that lidocaine increased response thresholds of nociceptive flexion reflexes in diabetic patients, also suggesting a central action. As amiodarone shares some of the same mechanisms of action as lidocaine, it may well have both a central and a peripheral mechanism of action. To conclusively demonstrate the site of action of amiodarone, electrophysiologic studies are required.
In previous studies using rat models, the largest concentrations of amiodarone after an IV injection of amiodarone 50 mg/kg were found within 5–30 minutes of administration (25). Previous workers have demonstrated that the terminal half-life of amiodarone ranges from 17 to 20 hours (26) after a single IV bolus dose, 15 hours after a 100 mg/kg dose, and 105 hours after 200 mg/kg oral dosage (23). We therefore decided to test our rats at 1-hour and 3-hour intervals and at 24 hours to see if the effects of intraperitoneal amiodarone correlate with its elimination half-life.
Previous workers have used amiodarone in rats in dose ranges from 25 mg/kg to 200 mg/kg (22,25,27). However, increasing doses are associated with increasing toxicity, with pulmonary and hepatic toxicity as two major side effects of chronic amiodarone therapy (27).
One main area of concern is the effect of accumulation of amiodarone with chronic administration (28). We therefore monitored the animals for signs of toxicity, such as loss of weight or decreased appetite, and none was apparent over the trial period.
In conclusion, amiodarone decreases the heat, cold, and mechanical hyperalgesia seen in a rat model of neuropathic pain. Considering its long half-life in humans, amiodarone may have the potential to provide long lasting pain relief in neuropathic pain states.
|
Acknowledgments
|
|---|
Supported, in part, by the Department of Anesthesiology and Pain Management, Cook County Hospital, Chicago, Illinois.
|
Footnotes
|
|---|
Presented, in part, at the Midwest Anesthesiology Residents’ Conference, April 16–18, 1999, Columbus, Ohio, the Rush University Forum for Research and Clinical Investigation, Chicago, Illinois, May 18–20, 1999, and the IARS 74th Clinical and Scientific Congress, March 10–14, 2000, Honolulu, Hawaii.
|
References
|
|---|
- Bach FW, Jensen TS, Kastrup J. The effects of intravenous lidocaine on nociceptive processing in diabetic neuropathy. Pain 1990; 40: 29–34.[ISI][Medline]
- Boas RA, Shahnarian A. Analgesic response to intravenous lidocaine. Br J Anaesth 1982; 54: 501–5.[Abstract/Free Full Text]
- Kastrup J, Peterson P, Dejgard A, et al. Intravenous lidocaine infusion: a new treatment for chronic painful diabetic neuropathy? Pain 1987; 28: 69–75.[ISI][Medline]
- Marchettini P, Lacerenza M, Marangoni C, et al. Lidocaine test in neuralgia. Pain 1992; 48: 377–82.[ISI][Medline]
- Abram SE, Yaksh TL. Systemic lidocaine blocks nerve injury- induced hyperalgesia and nociceptor-driven spinal sensitization in the rat. Anesthesiology 1994; 80: 383–91.[ISI][Medline]
- Catterall WA. Common modes of drug action on Na+ channels: local anesthetics, antiarrhythmics and anticonvulsants. Trends Pharmacol Sci 1987; 8: 57–65.
- Woolf CJ, Wiesenfeld-Hallin Z. The systemic administration of local anesthetics produces a selective depression of C-afferent fibre evoked activity in the spinal cord. Pain 1985; 23: 361–74.[ISI][Medline]
- Garfield JM, Gugino L. Central effects of local anesthetic agents. In: Strichartz GR, ed. Local anesthetics. Berlin: Springer-Verlag, 1987: 253–84.
- Pollack PT. Oral amiodarone: historical overview and development. Pharmacotherapy 1998; 18: 121S–6.[ISI][Medline]
- Naccarelli GV, Jalal S. Intravenous amiodarone: another option in the acute management of sustained ventricular tachyarrhythmias. Circulation 1995; 92: 3154–5.[Free Full Text]
- Kodama I, Kamiya K, Toyama J. Amiodarone: Ionic and cellular mechanisms of action of the most promising class III agent. Am J Cardiol 1999; 84: 20R–8.[ISI][Medline]
- Bennett GJ, Xie YK. A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain 1988; 33: 87–107.[ISI][Medline]
- Hargreaves K, Dubner R, Brown F, et al. A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia. Pain 1988; 32: 77–88.[ISI][Medline]
- Choi Y, Yoon YW, Na HS, et al. Behavioral signs of ongoing pain and cold allodynia in a rat model of neuropathic pain. Pain 1994; 59: 369–76.[ISI][Medline]
- Ren K, Dubner R. NMDA receptor antagonists attenuate mechanical hyperalgesia in rat with unilateral inflammation of the hindpaw. Neurosci Lett 1993; 63: 22–6.
- Kontinen VK, Aarnisalo AA, Idanpaan-Jeikkila JJ, et al. Neuropeptide FF in the rat spinal cord during carrageenan inflammation. Peptides 1997; 18: 287–92.[ISI][Medline]
- Wagner JA, Weisman HF, Levine JH, et al. Differential effect of amiodarone and desethylamiodarone on calcium antagonist receptors. J Cardiovasc Pharmacol 1990; 15: 501–7.[ISI][Medline]
- Du XJ, Esler MD, Dart AM. Sympatholytic action of intravenous amiodarone in the rat heart. Circulation 1995; 91: 462–70.[Abstract/Free Full Text]
- Haikerwal D, Du XJ, Turner A, et al. Presynaptic antisympathetic action of amiodarone and its metabolite desethylamiodarone. J Cardiovasc Pharmacol 1999; 33: 309–15.[ISI][Medline]
- Matsumori A, Ono K, Nishio R, et al. Amiodarone inhibits production of tumor necrosis factor- by human mononuclear cells: a possible mechanism for its effect in heart failure. Circulation 1997; 96: 1386–9.[Abstract/Free Full Text]
- Ito H, Ono K, Nishio R, et al. Amiodarone inhibits interleukin 6 production and attenuates myocardial injury induced by viral myocarditis in mice. Cytokine 2002; 17: 197–202.[ISI][Medline]
- Plomp TA, Wiersinga WM, Maes RA. Tissue distribution of Amiodarone and desethylamiodarone in rats after multiple intraperitoneal administrations of various amiodarone doses. Arzneimittelforschung 1985; 35: 122–9.[Medline]
- Plomp TA, Wiersinga WM, Van Rossum JM, Maes RA. Pharmacokinetics and body distribution of amiodarone and desmethylamiodarone in rats after oral administration. In Vivo 1987; 1: 265–79.[Medline]
- Abdi S, Lee DH, Chung JM. The anti-allodynic effects of amitriptyline, gabapentin and lidocaine in a rat model of neuropathic pain. Anesth Analg 1998; 87: 1360–6.[Abstract/Free Full Text]
- Riva E, Gerna M, Neyroz P, et al. Pharmacokinetics of amiodarone in rats. J Cardiovasc Pharmacol 1982; 4: 270–5.[ISI][Medline]
- Weir SJ, Ueda CT. Amiodarone pharmacokinetics. I. Acute dose-dependant disposition studies in rats. J Pharmacokinet Biopharm 1986; 14: 601–13.[ISI][Medline]
- Kannan R, Sharma JS, Guha M, Venkataraman K. Tissue drug accumulation and ultrastructural changes during amiodarone administration in rats. Fundam Appl Toxicol 1989; 13: 793–803.[ISI][Medline]
- Latini R, Bizzi A, Cini M, et al. Amiodarone and desethylamiodarone tissue uptake in rats chronically treated with amiodarone is non-linear with the dose. J Pharm Pharmacol 1987; 39: 426–31.[ISI][Medline]
Accepted for publication August 13, 2003.
Top
Abstract
Introduction
