Severe Toxic Damage to the Rabbit Spinal Cord after Intrathecal Administration of Preservative-free S  (+)-Ketamine

The study protocol was approved by the Animal Research Committee of the Academic Hospital at the University of Amsterdam (Amsterdam, The Netherlands). The rabbits were kept according to the National Guidelines for Care of Laboratory Animals in The Netherlands. Eighteen New Zealand white female rabbits weighing 3.6 ± 0.2 kg (mean ± SD) were used in this study. Before commencing the study, the rabbits were adapted (for 5 days) to wearing a protective vest.

Upon receipt, the animals were randomly assigned into two groups, consisting of 6 rabbits in the control group and 12 rabbits in the S  (+)-ketamine group. The control group received 0.9% saline solution, 0.5 ml (pH 5.0), and the S  (+)-ketamine group received commercially available, preservative-free S  (+)-ketamine (Ketanest-S, 5 mg/ml; Pfizer, Capelle a/d Ijssel, The Netherlands), 0.5 ml (pH 4.5), equivalent to 0.7 mg/kg. The drugs were administered very slowly (lasting 3 min) intrathecally, once a day in the morning, for 7 consecutive days. After each administration, the dead space of the catheter was cleared with 0.2 ml isotonic saline.

The rabbits were anesthetized with intramuscular ketamine (50 mg/kg) and xylazine (10 mg/kg). Subsequently, anesthesia was maintained with 1.5% isoflurane by mask in a mixture of 50% oxygen in air. Enrofloxacin (25 mg) was given before incision and continued until 2 days after surgery. During anesthesia, oxygen saturation and heart rate were measured using a tail oximeter.

Under sterile conditions, a straight midline incision was performed between the fifth and sixth spinous processes at lumbar level. After a laminectomy and removal of ligamentum flavum and epidural fat, the transparent dura mater with the spinal cord was exposed. After a small slit in the dura mater with leakage of cerebrospinal fluid, a 23-gauge polyamide catheter (Pajunk, Geisingen, Germany) was inserted 3 cm cephalic into the intrathecal space. The right position of the catheter was confirmed by cautious aspiration of cerebrospinal fluid. Subsequently, the free end of the catheter was tunneled subcutaneously and externalized at the nape of the neck. To protect the catheter, the animal wore a special protective vest. After anesthetic recovery, the neurologic status of the animals was assessed. Any evidence of neurologic (motor) dysfunction due to the implantation of the catheter itself resulted in exclusion from the study. Postoperative analgesia was accomplished by infiltration of the wound with 0.25% bupivacaine and intramuscular administration of buprenorphine (0.05 mg/kg), and 10 mg flunixin (Finadyne; Schering-Plough, Utrecht, The Netherlands). The intrathecal catheter remained in situ  for 7 days.

The primary outcome parameter was the observed histopathologic changes in the spinal cord and spinal nerves after repeated intrathecal administration of S  (+)-ketamine. Neurologic status of the animals (secondary outcome) was evaluated before start of treatment and 1 week after placement of the intrathecal catheter.

During treatment, the animals were observed for any sign of neurologic impairment while moving around freely in the cages. The animals were scored neurologically on the five-point scale of Tarlov before the procedure and 8 days (end of treatment) after insertion of the catheter: 0 = paraplegia with no lower-extremity function; 1 = poor lower-extremity function, weak antigravity movement only; 2 = some lower-extremity function with good antigravity movement but inability to draw legs under body and/or hop; 3 = ability to draw legs under body and hop but not normally; 4 = normal motor function.10Because intrathecal administration of S  (+)-ketamine could result in temporary gait deficits, neurologic status of the animals was assessed in the afternoon by an observer, blinded to the treatment allocation (and not responsible for the intrathecal administrations).

Eight days after insertion of the intrathecal catheter, the animals were anesthetized with intramuscular ketamine, 50 mg/kg, and xylazine, 10 mg/kg, followed by 1.5% isoflurane by mask in a mixture of 50% oxygen in air. After administration of heparin (2,500 U), animals were killed with pentobarbital (100 mg intravenously) and perfusion fixated (venous cannulas were placed transcardially and in the aorta) with 3.6% formalin after flushing with normal saline. The spinal cord was removed in toto  (dorsal laminectomy was performed from S1 to T1) and immersed in 10% formalin for at least 10 days. The spinal cord in rabbit has a length of approximately 10–11 cm. The cord was therefore cut in three equal parts (cervical, thoracic, lumbar). From each proximal third part, sagittal sections of the cord were taken ±0.5 mm thick and embedded in paraffin. The remainder was longitudinally cut and also embedded. Per level, at least 24 slides, 6 for each transverse section and 6 for each longitudinal section, 6 μm thick, were cut and stained with hematoxylin and eosin, Klüver-Barrera, and Nissl. The slides were examined by a neuropathologist blinded to the drug (placebo or S  (+)-ketamine) used. Sections were evaluated for the presence or absence of histologic alterations including myelin pallor, myelin loss, axonal swellings, central chromatolysis of neurons, subpial lymphocytic infiltration, infarction, gliosis, and leukodiapedesis. Neurohistopathologic damage was present if at least one out of six slices showed any of these changes.

Normally distributed data were expressed as mean ± SD. Nonnormally distributed data were expressed as median and interquartile range (25th–75th percentile). Baseline weight and temperature data of the two groups were compared bidirectionally using a Student t  test. For neurologic status (modified Tarlov score) of the animals after intrathecal treatment, two-sided statistical comparisons were performed using the Mann–Whitney U test. In case of histopathologic lesions, a two-sided Fisher exact test was performed. To assess the strength of the association between neurologic status and the presence of histopathologic lesions in the spinal cord, the Spearman rank correlation coefficient was calculated and tested unidirectionally for significance. For all analyses, significance was set at a value of P < 0.05.

Eighteen rabbits were randomized and treated in this study. The two groups (6 rabbits in the placebo group, 12 rabbits in the S  (+)-ketamine group) were similar with respect to sex (all rabbits were female), baseline body weight, and rectal temperature (table 1). One animal in the control group was killed during placement of the catheter because of intraoperative systemic complications (severe hypotension).

No rabbits were excluded because of neurologic deficits after catheter implantation. In one animal from the placebo group, neurologic evaluation was not possible 7 days after the start of treatment because of an accidental removal of the intrathecal catheter on day 5. In this animal, histopathologic examination was performed 6 days after intrathecal placement of the catheter. In the other rabbits, the intrathecal catheter remained in situ  during the course of the study. These animals were available for neurologic assessment and postmortem examination of the spinal cord after 7 days of treatment.

Motor function was affected temporarily (±20 to 30 min, modified Tarlov scores ranging from 1 to 3) in all of the rabbits of the ketamine group after each daily intrathecal administration of S  (+)-ketamine. The rabbits receiving saline displayed no signs of motor impairment or other neurologic deficits. Modified Tarlov scores, before and 7 days after the start of treatment, are shown in table 1. Neurologic deficit (modified Tarlov score lower than 4) was observed in six rabbits receiving S  (+)-ketamine (table 1). There was, however, no significant difference in modified Tarlov scores between the control group and the S  (+)-ketamine group after 7 days of treatment (P = 0.063).

The histopathologic results are shown in table 1. Eleven rabbits from the group of animals receiving S  (+)-ketamine presented histopathologic lesions suggestive for neurotoxicity. These results were significantly different compared with the control animals (no histologic changes observed; P = 0.001).

Histology of the spinal cords showed mild to severe gray and white matter damage. Most lesions were present around the spinal canal and consisted of small severely damaged areas resembling small infarctions. Identical subpial lesions were a frequent finding. In some cords, focal white matter damage was observed. The gray matter damage ranged from central chromatolysis of motor neurons to obvious necrosis of small areas. The damage around the central canal consisted of ependymal loss and subependymal necrotic lesions. The subpial lesions showed mainly myelin loss combined with axonal swellings, necrosis, and reactive leukodiapedesis. Vascular damage was not observed. Whether these lesions were in continuity with the subpial region or with the ependyma could not clearly be established (fig. 1).

Combining the data from both groups, there was a significant correlation between the presence of neurologic injury (assessed by the modified Tarlov score) and the presence of histopathologic lesions in the spinal cord (r = 0.533; P = 0.014). In addition, there was no significant correlation between neurologic impairments and histopathologic changes of the spinal cord in the ketamine group.

The results of the current blinded, placebo-controlled study in rabbits showed morphologic evidence of neurotoxicity after 7 days of repeated (once a day) subarachnoid administration of commercially available, preservative-free 0.5% S  (+)-ketamine. Postmortem examination demonstrated severely damaged spinal cords in eight animals. Apparently, there were predilection sites for spinal cord damage caused by S  (+)-ketamine. The lesion sites, subpial and around the central canal, suggested both that there was an immediate toxic effect on the tissue exposed to S  (+)-ketamine and that the subpial and subependymal regions were the first to be affected. Both axonal swellings and the chromatolysis suggested also a more profound noxious neuronal effect. In some instances, the extent of the lesions looked vascular, because of the extent and sharp borders of some of the lesions, but vascular changes were not observed. Obviously, the damage resulted finally in necrosis.

Questions have been raised about the potential neurotoxicity of the neuraxial use of ketamine,4although the safety of this approach is supported by a number of clinical studies and case reports with no neurologic or behavioral changes suggesting neurotoxicity.2,11,12In addition, in one study with rabbits, single intrathecal administration of 0.3 ml preservative-free S  (+)-ketamine, 1% (not commercially available), produced spinal analgesia with no histopathologic lesions in the spinal cord.4However, the use of single-injection techniques does not reveal any information about long-term use of a drug.13Because single-shot administration of S  (+)-ketamine has a short-term effect only, safety studies on prolonged administration of this analgesic are necessary before widespread use in the treatment of pain in humans, which may last several days (postoperative pain) to several months (neuropathic cancer pain).

In this study, repeated intrathecal administration of S  (+)-ketamine resulted in histopathologic changes in the spinal cord of the rabbit. These results were significantly different compared with the control animals. The presence of histopathologic lesions in the spinal cord did not correlate significantly with neurologic impairments assessed with the modified Tarlov score in the ketamine group. There was no significant difference in neurologic status between the two groups after 7 days of intrathecal treatment.

In addition, not all animals receiving intrathecal S  (+)-ketamine developed deterioration in neurologic function despite evidence of severe histologic toxicity. A possible explanation may be the lack of precision of the modified Tarlov score to assess neurologic function. However, these data may also imply that clinical observations alone may not provide sufficient evidence for assessing the safety index of intrathecal administration of S  (+)-ketamine. Therefore, this observed lack of convergence emphasizes the importance of assessing spinal histopathology to evaluate the neurotoxicity of spinal drugs.14 

The histopathologic lesions detected in our study may be due to an excessive N -methyl-d-aspartate receptor antagonism by S  (+)-ketamine. This antagonism may produce neurotoxicity attributed to an inactivation of an inhibitory mechanism (involving a blockade of glutamate N -methyl-d-aspartate receptors on γ-aminobutyric acid–mediated interneurons, which are responsible for a tonic inhibition of excitatory pathways) resulting in an excitotoxic injury with cell necrosis and apoptosis.15In addition, there is evidence to suggest that prolonged N -methyl-d-aspartate receptor antagonism hinders endogenous mechanisms for neuronal survival and neuroregeneration.16Factors other than S  (+)-ketamine including low pH and the injection volume of the intrathecal solution and mechanical damage caused by the intrathecal catheter may be responsible for the observed histopathologic changes. However, previous reports with intrathecal catheters administering solutions with similar pH and injection volume did not demonstrate histologic findings suggestive of neurotoxicity.5,17,18Moreover, we did not observe histopathologic changes in our control group receiving saline.

Intrathecal administration of S  (+)-ketamine produced signs of neurotoxicity in this preclinical model. However, the spinal cord of the rabbit may be more sensitive than the human spinal cord in demonstrating neurotoxicity after intrathecal administration of S  (+)-ketamine.19Because the observed histopathologic lesions may be model dependent, translation of our findings into the equivalent clinical (human) situation may be difficult. In previous studies, however, the rabbit model did demonstrate the ability to show spinal cord pathology after intrathecal drug treatments (ketamine and midazolam) compared with saline.7,18In addition, in this study, we used the commercially available, preservative-free S  (+)-ketamine formulation in a relevant concentration (0.5%) and dosage (0.7 mg/kg once a day), also clinically used for intrathecal pain treatment in humans (daily doses ranging from 50 to 75 mg S  (+)-ketamine).13,20,21 

A drawback of our study was the intrathecal bolus administration of S  (+)-ketamine that could have been responsible for an increased cerebrospinal fluid level of S  (+)-ketamine in combination with a decreased spinal cord flow, responsible for the observed neurotoxicity. However, intrathecal administration of a comparable daily dose of S  (+)-ketamine by continuous infusion resulted also in histopathologic changes of the spinal cord.9 

Treatment of acute perioperative pain (improved and prolonged analgesia, especially after caudal administration in pediatric anesthesia) and pain relief in neuropathic pain syndromes could be improved after neuraxial administration of S  (+)-ketamine. This study was primarily designed to evaluate neurotoxicity after prolonged intrathecal administration of S  (+)-ketamine. However, in cases of an unintentional dura puncture during caudal or epidural administration of S  (+)-ketamine (bolus administration of 1 mg/kg), the concentration of this drug could reach toxic levels in the cerebrospinal fluid. In this view, more safety studies must be performed to evaluate acute neurotoxicity after high doses of intrathecal S  (+)-ketamine.

In summary, there is no doubt, considering the extent and severity of the lesions, that prolonged intrathecal use of S  (+)-ketamine has a serious toxic effect on the central nervous system in rabbits. In this study, however, neurologic assessment of the animals during treatment provide insufficient evidence for proving these neurotoxic effects. Our results underline the necessity of assessing spinal histopathology before neuraxial administration of drugs may be used for clinical practice. Based on our findings we conclude that the use of neuraxial S  (+)-ketamine in humans should be avoided because the histologic lesions caused by this agent suggest a neurotoxic effect. The possible benefits do not justify clinical application, and we should look for a replacement of S  (+)-ketamine, administered neuraxially, to improve pain management.

The authors thank Marloes Klein (Technician, Department of Experimental Surgery, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands) for expert technical assistance in performing this study.