No Neurotoxicity from Long-Term (>5 Years) Intrathecal Infusion of Midazolam in Humans

Abstract

Although morphine (but also fentanyl, hydromorphone, sufentanil, and meperidine) has been used in neuropathic pain, analgesia is often incomplete and the addition of bupivacaine or clonidine is often used for their potential synergistic effects. Opioids (most often morphine) have a low potential for neurotoxicity. The selective alpha-2 adrenergic agonist clonidine, a lipophilic drug with rapid onset and short duration of action, has been shown to produce analgesia in humans and animals, without clinical neurotoxicity. Intrathecal (IT) bupivacaine, too, appears to be relatively free of significant side effects. Midazolam is a benzodiazepine-class drug acting on the benzodiazepine/gamma aminobutyric acid (GABA)-A receptor complex. IT bolus doses of 2 mg midazolam result in segmental analgesia with long-term effects after a one-time application,1Serrao J.M. Marks R.L. Morley S.J. Goodchild C.S. Intrathecal midazolam for the treatment of chronic mechanical low back pain: a controlled comparison with epidural steroid in a pilot study.Pain. 1992; 48: 5-12Abstract Full Text PDF PubMed Scopus (180) Google Scholar, 2Valentine J.M.J. Lyons G. Bellami M.C. The effect of intrathecal midazolam on post-operative pain.Eur J Anaesthesiol. 1996; 13: 589-593Crossref PubMed Scopus (72) Google Scholar and IT midazolam is synergistic with bupivacaine in postoperative analgesia.3Batra Y.K. Jain K. Chari P. et al.Addition of intrathecal midazolam to bupivacaine produces better post-operative analgesia without prolonging recovery.Int J Clin Pharmacol Ther. 1999; 37: 519-523PubMed Google Scholar Borg and Krijnen4Borg P.A. Krijnen H.J. Long-term intrathecal administration of midazolam and clonidine.Clin J Pain. 1996; 12: 63-68Crossref PubMed Scopus (73) Google Scholar reported the use of IT midazolam and clonidine in four patients with chronic nonmalignant pain not responding to conventional analgesia, and demonstrated nearly complete pain relief from this regimen. There was apparently no tolerance development and no side effects. Rainov et al.5Rainov N.G. Heidecke V. Burkert W. Long-term intrathecal infusion of drug combinations for chronic back and leg pain.J Pain Symptom Manage. 2001; 22: 862-871Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar found no toxicity from drug combinations consisting of up to four drugs, including midazolam, stored in the pump reservoir for 1–4 months, in chronic back pain. Generally, doses used in clinical studies ranged from 0.3 to 12 mg/day. Despite this experience, midazolam is rarely used for the treatment of neuropathic pain, in part due to concerns about possible acute and chronic neurotoxicity. We report a patient suffering from peripheral neuropathic pain who had no evidence of neurotoxicity from long-term infusion of a combination of midazolam and clonidine. A 60-year-old woman, with a history of lupus erythematosus, developed atypical trigeminal neuralgia involving the left trigeminal branches I and II in the 1980s. This was unresponsive to carbamazepine. In 1989, thermocoagulation of the left Gasserian ganglion at another institution abolished the pain for one year. In 1990, left corneal ulcers appeared. She also developed postherpetic neuralgia. The eye was extirpated and a prosthesis inserted. Since 1991, nearly continuous pain with sporadic exacerbations has occurred in the cheekbone and forehead, along with an icy intolerable dysesthesia in the left hemiface. Carbamazepine and amitriptyline at maximally tolerated doses, among other drugs, proved ineffective. In 1993, a cervical spinal cord stimulator was installed elsewhere, with benefit lost within a month. In 1995, an extradural cortical stimulator was applied by us over the right motor cortex (face projection), with excellent benefit (60%–80%). After about six months, though, this benefit was lost. In February, 1998, after a positive IT challenge test, a Synchromed® (Medtronic) pump was implanted and preservative-free midazolam (Ipnovel®, Roche, Italy) and clonidine were infused. Within a short time, her analgesia stabilized with a reduction of about 90% of her previous pain. Her previous suicidal ideation regressed. Six and a half years later, benefit continues unabated. Her pump was replaced in 2001. Daily doses have ranged between 2 and 4.6 mg of midazolam. No toxicity whatsoever has been observed both clinically and on routine repeated hematological tests. However, side effects were noted when clonidine was increased beyond initial doses (e.g., hypotension) to obtain 100% control in the initial stages. Neuropsychological examinations were always within normal range. Animal studies of IT midazolam remain controversial. Nishiyama et al.6Nishiyama T. Matsukawa T. Hanaoka K. Acute phase histopathological study of spinally administered midazolam in cats.Anaesth Analg. 1999; 89: 717-720PubMed Google Scholar infused midazolam IT (10 mg bolus dose) in cats and were unable to find any acute histological damage or inflammatory reaction of the spinal cord, thus concluding that spinally administered midazolam, even in large doses, does not cause acute spinal neurotoxicity or inflammation. In sheep and pigs, IT midalozam at 5–15 mg/day for 43 days produced no signs of neurotoxicity.7Johansen M.J. Satterfield W.C. Baze W.B. et al.Toxicity and efficacy of intrathecal midazolam.Pain Med. 2002; 3: 188-189Google Scholar On the other hand, histological/morphological signs of neurotoxicity in the rabbit or rat spinal cord, such as neuronal death and degeneration of cell somata and fibers, have been reported with electron and modified light microscopy. Erdine et al.8Erdine S. Yucel A. Ozyalcin S. et al.Neurotoxicity of midazolam in the rabbit.Pain. 1999; 80: 419-423Abstract Full Text Full Text PDF PubMed Scopus (70) Google Scholar administered 0.3 mg IT to rabbits and Svensson et al.9Svensson B.A. Welin M. Gordh Jr., T. Westman J. Chronic subarachnoid midazolam (Dormicum) in the rat. Morphological evidence of spinal cord neurotoxicity.Reg Anesth. 1995; 20: 426-434PubMed Google Scholar injected 0.1 mg IT in rats for 20 days. Both studies found significant histologic changes in the spinal cords of the animals, which were interpreted as drug-related neurotoxic effects and led these authors to contraindicate midazolam for long-term management. These animal studies, however, are not comparable to clinical investigations. First, midazolam was given in animals at higher concentrations than the doses used in human trials. Secondly, the subarachnoid space of a rabbit or a rat is much smaller than in humans, with much less dilution effect than humans receiving minimal volumes of an IT drug into a large cerebrospinal fluid (CSF) pool. Third, preservative-free formulations were not always used. Moreover, possible risks must be weighed against benefit in refractory human cases in whom suicide is possible. Borg and Krijnen4Borg P.A. Krijnen H.J. Long-term intrathecal administration of midazolam and clonidine.Clin J Pain. 1996; 12: 63-68Crossref PubMed Scopus (73) Google Scholar observed no side effects (e.g., hypotension), nor any clinical signs of neurotoxicity in their patients. The longest infusion ran for 2.5 years without any sign of tolerance or accumulation. Midazolam and clonidine serum and CSF levels were measured before and after discharge to determine any possible accumulation. Doses were up to 6.5 mg/day and 600 μg/day of midazolam and clonidine, respectively. Midazolam serum levels were <50 and <25 μg/L; CSF levels were 924 and 360 μg/L, respectively; clonidine levels were below the threshold for detection (<1 μg/L). Rainov et al.5Rainov N.G. Heidecke V. Burkert W. Long-term intrathecal infusion of drug combinations for chronic back and leg pain.J Pain Symptom Manage. 2001; 22: 862-871Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar reported that midazolam and clonidine concentrations in CSF and blood (measured in one patient only) showed no signs of accumulation. Titration in the first three patients was carried out by adding 1 mg midazolam and 50 μg clonidine each time until reaching a self-imposed limit of 6 mg midazolam daily, in which case further titration consisted of increasing the clonidine dose by 10 μg each time (increases of 0.1 mg midazolam and 5 μg clonidine were indicated in another). In our large experience with acute boluses of IT midazolam (2.5 mg) for neurogenic pain control (unpublished observations), we never have seen any sign of toxicity of any kind, nor side effects of any kind (e.g., nausea or drowsiness) with doses up to 6 mg. We have used preservative-free midazolam (Ipnovel). The drug mixture has remained analgetically active in the pump reservoir for the time period between refills (weeks to months). Due to the low concentrations of drugs available to us, pump refills must be repeated relatively often (usual doses: 70 mg of midazolam and approximately 600 μg of clonidine). The median number of refills per year is 12 to 15. In conclusion, IT midazolam appears safe in the long term. No sign of acute or chronic neurotoxicity has emerged up to now. More consideration should be given to this drug in the IT management of chronic pain.