NEUROLOGICAL TOXICITY OF THE SUBARACHNOID INFUSION OF BUPIVACAINE, LIGNOCAINE OR 2-CHLOROPROCAINE IN THE RAT

Transcription

1 Br.J. Anaesth. (98),, -9 NEUROLOGICAL TOXICITY OF THE SUBARACHNOID INFUSION OF BUPIVACAINE, LIGNOCAINE OR -CHLOROPROCAINE IN THE RAT D. F. LI, M. BAHAR, G. COLE AND M. ROSEN Neurological deficit after administration of local anaesthetics is a potential complication of extradural or spinal anaesthesia. Previous animal studies reported on the neurotoxic effects of a single bolus injection of a large volume of local anaesthetics to the subarachnoid space (Ravindran, Turner and Muller, 98; Rosen et al., 98). Artefacts determining the methodology included traumatic injury to the spinal cord and the need for prolonged cardiovascular and ventilatory support, which might affect perfusion of the spinal cord. At present there is interest in the use of continuous infusions of local anaesthetics for extradural analgesia but, with this method, the safety of nervous tissue constantly exposed to local anaesthetics which may be accidentally infused intrathecally, remains to be established. The present experiments were performed on unanaesthetized, mobile rats to study the neurological sequelae of intrathecal infusion of bupivacaine (Marcain), lignocaine (Xylocaine), -chloroprocaine (Nesacaine-CE) and Hartmann's Solution (Travenol). All solutions were free of preservative. MATERIALS AND METHODS Preparation of rats Three to five days before the experiment, the subarachnoid spaces of Wistar rats (8- g) were cannulated with fine polyethylene tubing (Portex) via the lumbar vertebrae as previously reported (Bahar, Rosen and Vickers, 98). The location of DOMINIC F. LI,* M.B., B.S., M.R.C.O.G., M. BAHAR, M.D., M. ROSEN, M.B., CH.B., F.F.A.R.C.S. (Department of Anaesthetics); G. COLE, M.B., CH.B., D.P.M., M.D. (Department of Pathology); Welsh National School of Medicine, Health Park, Cardiff CF XN. Present address: Department of Obstetrics and Gynaecology, University of Hong Kong, Hong Kong. Correspondence to M. Rosen. SUMMARY Neurotoxicity after subarachnoid infusion of bupivacaine, lignocaine and -chloroprocaine was studied in a chronic rat model. Hartmann's solution \il h' was infused as a control, and.% bupivacaine,.% lignocaine and.% - chloroprocaine were infused at \il h~ for, 6 or h, to five rats in each group. No residual paralysis occurred in the control group, but of rats (6%) which received an infusion of local anaesthetic had residual paralysis lasting until sacrifice at days. The incidence of paralysis was dependent on the duration of exposure to the local anaesthetic, but there were no significant differences in incidence between any of the local anaesthetics tested. Abnormal histology, in the form of neuronal vacuolation, was not a sensitive index, being present in control rats, but more intense in those receiving lignocaine and - chloroprocaine than in those given bupivacaine; no correlation with clinical findings could be established. The neurotoxic effects of each local anaesthetic tested as a continuous intrathecal infusion were dose related in the rat, which maybe a useful model for screening other local anaesthetics. the catheter was confirmed by injecting % lignocaine ulitre, which invariably paralysed both hind limbs of the rat and conferred lumbar analgesia. Those rats in which there was any doubt about the location of the catheter or which showed any evidence of traumatic nervous damage were not included in the experiment. Plan of investigation On the day of the experiment, a jacket was put on the rat which was under inhalation anaesthesia

2 NEUROTOXICITY OF SUBARACHNOID ANAESTHETICS (.% halothane in oxygen ml min" ). The intrathecal cannula was passed through a swivel clamped to the cage and connected to a syringe pump. The whole system was protected by a lightweight metal tether (Harvard Apparatus). The animal recovered from anaesthesia and was fully mobile. Five rats were randomly allocated to each of groups receiving an intrathecal infusion at ul Group : Control group with Hartmann's solution (ph.9-6.9) for h. Groups -:.% Bupivacaine (ph.6-.6) for, 6 and h, respectively. Groups -:.% Lignocaine (ph.-.9) for, 6 and h, respectively. Groups 8-:.% -Chloroprocaine (ph.9-.) for, 6 and h, respectively. All drugs were free of preservative with no adrenaline, and the local anaesthetics consistently paralysed both hind limbs but left the front limbs mobile. During the period of infusion all rats were free to move, with access to food and water. After the designated period the infusion was disconnected and the rats returned to separate cages, where they were examined daily for any residual neurological deficit, for days before sacrifice for pathological examination. Histology Immediately after the rat was killed, ulitre of % formalin was injected slowly through the intrathecal catheter. The vertebral column was then removed in one piece and fixed in % formalin in saline for days. The specimen was then decalcified for week. Sections of the spinal cord were taken from the region adjacent to the tip of the catheter, which should have been subjected to the maximum effects of the drugs. Sections were embedded in paraffin, cut at -um intervals and stained with haematoxylin and eosin (H/E) and Kluver-Barrera stain (KLB) for myelin. Microscopic examination was carried out by a neuropathologist (G.C.) who was unaware of which drug had been infused. Statistical analysis was by Chi-square test. RESULTS Clinical findings Apart from the group receiving Hartmann's solution, a steady block with lumbar analgesia and paralysis of both hind limbs were observed in all rats receiving local anaestheic, throughout the infusion. No residual neurological deficit occurred in the group receiving Hartmann's solution (table I). Of the rats which received an infusion of local anaesthetic, had residual paralysis of at least one hind limb, usually the right one: nine in the bupivacaine groups, eight in the lignocaine groups and in the -chloroprocaine groups (table I). The neurotoxic effects were dose dependent and related to the duration of exposure: of which received the infusion for h and of which received the infusion for 6 h, but only four of which received the infusion for h were paralysed (x =.66, P <.). No difference was observed between the effects of the three local anaesthetics (X =.6, ns). Histological findings In all animals, the white matter showed spongiose changes, but no obvious cellular reaction. This appearance was present in control animals, and was therefore probably an artefact related to the method of fixation. The grey matter of some animals showed varying degrees of intra- and perineuronal vacuolation (fig. ), the occurrence of which was more common in the lignocaine and -chloroprocaine groups than in the bupivacaine groups (tables II and III) (X = 6.6, P <.). However, two rats receiving Hartmanns's solution also had similar TABLE I. Neurological sequelae of -\d h~' subarachnoid infusions. Five rats in each of the groups. (Mean weights of the control group, bupivacaine group, lignocaine group and -chloroprocaine group were.8,., 8. and.9 g respectively) Infusion and duration Hartmann's solution h.% Bupivacaine h 6h h.% Lignocaine h 6h h.% -Chloroprocaine h 6h h Totals Total dose infused (mg) Residual Yes paralysis No

3 S BRITISH JOURNAL OF ANAESTHESIA TABLE II. Histological findings in rats with no residual paralysis TABLE III. Hislological findings in rats with residual paralysis Infusion Hartmann's solution.%bupivacaine.%lignocaine.% -Chloroprocaine Totals Abnormal Histological findings Normal 9 Total 6 neuronal changes. Otherwise, neurones looked healthy, with central nuclei and well stained Nissl's substance (fig. ). No correlation could be found between these histological changes and the clinical findings. A mild inflammatory and granulomatous reaction was present around some of the catheter sites, but in no case was this severe (fig. ). In some animals the catheter appeared to impinge on the lateral white columns of the cord, but these animals had not shown any neurological deficit before the infusion, so that this finding could not be important. The ganglia and nerve roots appeared normal and there was no evidence of demyelination. DISCUSSION There has been a gradual move towards the use of Infusion Hartmann's solution.%bupivacaine.%lignocaine.% -Chloroprocaine Totals Histological findings Abnormal Normal 6 Total 9 8 continuous extradural infusion of local anaesthetics for obstetric and postoperative pain relief (Evans and Carrie, 99; Matouskova, Hanson and Elmen, 99; Ross, Clarke and Armitage, 98; Davies and Fettes, 98; Taylor, 98). The alleged advantage is that "soaking the nerve roots in local anaesthetic might reduce the incidence of unblocked segments" (Evans and Carrie, 99). However, the potential problem of neurotoxicity resulting from such continuous exposure in the extradural space has not been tested adequately, especially if inadvertent intrathecal infusion occurred after accidental dural puncture. After case reports of neurological sequelae resulting from the use of -chloroprocaine (Ravindran et al., 98; Reisner, Hochman and Plumer, 98), both in vitro and in vivo studies were performed to FIG.. Neurones of anterior horn with perineuronal and intracytoplasmic vacuoles. Kluver-Barrera stain. Horizontal bar represents. mm.

4 NEUROTOXICITY OF SUBARACHNOID ANAESTHETICS FlG.. Well preserved neurones of anterior horn. Kluver-Barrera stain. Horizontal bar represents. mm. study the neurotoxicity of local anaesthetics in common use. In vitro studies which exposed the vagus nerve of rabbits to local anaesthetics in air showed conflicting results (Riggi, Harding and Hashim, 98; Barsa et al., 98), possibly because the process of oxidation of the drugs was not controlled (James, 98). Results from in vivo studies were also inconclusive. -Chloroprocaine was more FlG.. Catheter site with granulomatous reaction lateral to spinal cord in the subdural space. Haematoxylin and eosin. Horizontal bar represents mm.

5 8 BRITISH JOURNAL OF ANAESTHESIA neurotoxic in two studies (Ravindran, Turner and Muller, 98; Wang etal., 98), while others found no difference between the various local anaesthetics and control solutions (Feldman and Covino, 98; Ready et al., 98; Rosen et al., 98). We reexamined the effects of -chloroprocaine and compared them with the effects of two other agents (bupivacaine and lignocaine) which had not previously been tested in this controlled way, and found all three to behave similarly. We carried out experiments with continuous infusion intrathecally, since it was assumed that this would result in the most constant exposure to potential toxicity. Results from these experiments support this concept and show that, in unanaesthetized rats, prolonged exposure of nervous tissue to the local anaesthetics studied produced neurological deficit if the dose was large enough. The three agents were equally likely to cause such deficit, and showed a similar relationship between the duration of exposure and clinical effects: the longer the exposure, the worse were the neurological sequelae. In most of the earlier experiments, large volumes of local anaesthetics were administered by bolus injection to the subarachnoid space of animals. This invariably caused apnoea and wide cardiovascular fluctuations in the animals, and must have affected the vascular perfusion of the spinal cord, which could have contributed to the neurological deficit found in the animals after the experiment (Kane, 98). In our study arterial pressure was not measured because of technical difficulty in mobile, unanaesthetized animals, However, it seems reasonable to assume that arterial pressure was fairly stable, even though there was a steady block just below the thoracic region, since all the rats were eating, drinking and micturating normally during the periods of the infusions. In the present investigation the catheter was put into the subarachnoid space - days before the infusion. This avoided repeated lumbar puncture to give intrathecal injections (Ravindran, Turner and Muller, 98; Ready et al., 98; Rosen et al., 98) which could have caused traumatic injury to the spinal cord, and it allowed a period to ensure normal function after catheterization. The ph of the solutions we studied ranged from. to 6.9, which is comparable to those used in previous studies. It has been shown that low ph (.), normal saline solution did not produce residual paralysis (Ravindran, Turner and Muller, 98). It seems likely, therefore, that the local anaesthetics per se were the causative factor in the observed residual paralysis, and that -chloroprocaine behaves just like the other agents. It would be appropriate to compare the rat with other animal models to examine the consistency of response. There are still aspects to be clarified. It is likely, but untested so far in this model, that extradural infusion would require much larger doses of local anaesthetic drugs in order to cause similar neurological deficit, since the resulting intrathecal concentration will be much lower. Extradural infusion should therefore be safe in terms of neurological toxicity. Histological findings by light microscopy must be interpreted with caution. The neuronal changes did not usually correlate with the clinical findings of residual paralysis, which is in agreement with other studies (Ravindran, Turner and Muller, 98; Rosen et al., 98). Apart from cytoplasmic vacuolation, the majority of neurones looked healthy and were still viable, with no evidence of chromatolysis. Such changes have also been reported in various toxic and ischaemic conditions. Using the same method of catheter implantation, similar histological changes were found in rats which received small, repeated doses of cocaine, lignocaine, bupivacaine, normal saline and adrenaline over days, although none of the animals showed residual neurological deficit (Bahar et al., 98). At tissue level, the pharmacological behaviour of local anaesthetics given by bolus injection is probably very different from continuous infusions with lower concentration-time exposure. Although the occurrence of abnormal histology was more frequent in the lignocaine and -chloroprocaine groups, the clinical findings of paralysis did not differ significantly between these and the bupivacaine groups. A more sensitive indicator, perhaps neurochemical or electron microscopy, will be necessary to study histopathological changes in the rat spinal cord after toxic exposures. We therefore conclude that neurotoxicity can result from prolonged exposure to local anaesthetics and is dose related. Since -chloroprocaine did not give better or worse results than bupivacaine or lignocaine, its reported neurotoxicity in man may have been the result of the size of the bolus dose accidentally injected intrathecally, and such toxicity might occur with other local anaesthetics. The rat model should prove useful in screening other and new local anaesthetic agents.

6 NEUROTOXICITY OF SUBARACHNOID ANAESTHETICS 9 ACKNOWLEDGEMENTS We wish to thank Mr Ritchie and other members of the Animal Unit, the Medical Illustration Department of University Hospital of Wales, Mr C. Juniper for valuable technical assistance and Mrs H. O'Donnell for secretarial help. REFERENCES Bahar, M., Cole, G., Rosen, M., and Vickers, M. D. (98). Histopathology of the spinal cord after intrathecal cocaine, bupivacaine, lignocaine and adrenalin in the rat. Eur.J. Anaesthesiol, (in press). Rosen, M., and Vickers, M. D. (98). Chronic cannulation of the intradural or extradural space in the rat. Br. J. Anaesth., 6,. Barsa, J. E., Batra, M., Fink, B. R., and Sumi, S. M. (98). A comparative in vivo study of local neurotoxicity of lidocaine, bupivacaine, -chloroprocaine and a mixture of -chloroprocaine and bupivacaine. Anesth. Analg., 6, 96. Davies, A. O., and Fettes, I. W. (98). A simple safe method for continuous infusion epidural analgesia in obstetrics. Can. Anaesth. Soc.J., 8,8. Evans, K. R. L., and Carrie, L. E. S. (99). Continuous epidural infusion of bupivacaine in labour a simple method. Anaesthesia,,. Feldman, H. S., and Covino, B. G. (98). A chronic model for investigation of experimental spinal anesthesia in the dog. Aneslhesiology,, 8. James, F. M. (98). The anesthesiology triad in obstetrics. Aneslhesiology, 6,. Kane, R. E. (98). Neurologic deficits following epidural or spinal anesthesia. Anesth. Analg., 6,. Matouskova, A., Hanson, B., and Elmen, H. (99). Continuous mini-infusion of bupivacaine into the epidural space during labour. Part III a clinical study of parturients. Ada Obstet. Gynecol. Scand. (Suppl.), 8,. Ravindran,R. S., Bond, V. K.,Tasch,M. D., Gupta,C. D.,and Luerseen, T. G. (98). Prolonged neural blockade following regional analgesia with -chloroprocaine. Anesth. Analg., 9,. Ravindran, R. S., Turner, M. S., and Muller, I. (98). Neurological effects of subarachnoid administration of - chloroprocaine-ce, bupivacaine and low ph normal saline in dogs. Anesth. Analg., 6, 9. Ready, L. B., Plumer, M. H., Fink, B. R., and Sumi, S. M. (98). Intrathecal local anesthetic toxicity in rabbits. Anesthesiology, 9,A8. Reisner, L. S., Hochman, B. N., and Plumer, M. H. (98). Persistent neurologic deficit and adhesive arachnoiditis following intrathecal -chloroprocaine injection. Anesth. Anal., 9,. Riggi, S. J., Harding, S., and Hashim, M. (98). The toxicity of intrathecal -chloroprocaine injection. Aneslh., Analg., 9,. nerves of rabbits. Annual S.O.A.P. Meeting, April, 98, San Diego. Rosen, M. A., Baysinger, C. L., Shnider, S. M., Dailey, P. A., Norton, M., Curtis, J. D., Collins, M., and Davis, R. L. (98). Evaluation of neurotoxicity after subarachnoid injection of large volumes of local anesthetic solutions. Anesth. Analg., 6, 8. Ross, R. A., Clarke, J. E., and Armitage, E. N. (98). Postoperative pain prevention by continuous epidural infusion. Anaesthesia,, 66. Taylor, J. H. C. (98). Clinical experience with continuous epidural infusion of bupivacaine at 6 ml per hour in obstetrics. Can. Anaesth. Soc.J.,,. Wang, B. C, Spielholz, N. I., Hillman, D. E., andturndorf, H. (98). Subarachnoid sodium bisulfite (the antioxidant in Nesacaine) causes chronic neurological deficit. Anesthesiology,, A9.