• This paper presents an overview of agents administered intrathecally for the relief of pain including opioids, a-2 adrenergic agonists, N-type voltage-sensitive calcium channel blockers, somatostatin, N-methyl-D-aspartate receptor antagonists, adensosine, cholinergic agonists and GABA agonists. Opioids are the most commonly used drugs for long-term analgesic spinal drug therapy and morphine is the only opioid with FDA approval for long-term intrathecal delivery.
  
  
• Combined therapy using agents with different mechanisms of action may have positive therapeutic effects in pain states involving different mechanisms and as a result of lack of cross-tolerance or positive synergistic effects of certain agents.
  
  
• Chronic pain management is very difficult to study because of its multifactorial nature, and it is difficult to draw conclusions about the efficacy of long-term intraspinal drug therapy from the clinical studies available. There is a need for well-designed prospective randomised trials.
  
  
• These authors reported problems associated with the development of tolerance to intraspinal drug therapy. (However, in the Journal of Pain and Symptom Management, 1997, Portenoy RK reported minimal likelihood of opioid addiction or tolerance.) Several factors may contribute to the development of tolerance including changes in pharmacokinetics, pain intensity, pain mechanisms, pharmacodynamics and psychosocial status.
  
  
• The risks and costs of the different types of intraspinal drug delivery techniques must be taken into account when deciding which approach to use:
  • Externalised systems are used to deliver drugs to the perispinal space and are designed for short-term use. As the systems breach the cutaneous barrier, there is a risk of infection. Infection may be minimized by using a catheter with a Dacron cuff, or by tunneling the catheter under the skin if it is to be left in place for more than 7-10 days. The Medtronic Algoline Catheter is currently the only FDA-approved external intrathecal delivery system.
  • In partially externalized systems, the catheter and the access port are placed beneath the skin and drugs are injected directly into the access port. These systems permit freedom of movement and reduce the risk of catheter removal. Clinical studies suggest that the subcutaneous port systems are superior to Externalised systems if long-term-use is anticipated. There are no currently approved partially Externalised systems for intrathecal drug delivery.
  • Totally implanted systems involve the complete implantation of the catheter and delivery system. These systems reduce the risk of infection allow more patient independence, but are more expensive and require more surgical intervention than Externalised and partially Externalised systems. Accordingly patients should be carefully selected. Three devices are currently approved by the FDA; the Synchromed Infusion System, the Arrow Constant Flow Implantable System and the Infusaid System.
    
    
• Pump systems may be classified according to their mechanism of delivery, programmability, delivery rates and reservoir volumes. An additional consideration is the drug for which the pump is approved. The Synchromed Infusion System is the only programmable implantable pump and offers the advantage of permitting a variety of infusion modes. The SynchroMed is approved for intrathecal morphine and baclofen. The Arrow Continuous Flow Infusion System (ACFIS) is the newest pump on the market, but is only approved for intrathecal morphine, is not programmable and has a factory pre-set flow rate. The Medtronic constant flow rate infusion system, the IsoMed Infusion System, is currently undergoing clinical trials.

• Morbidity from intraspinal drug delivery is uncommon, but the risk of infection is more common with external systems. The most serious morbidity is related to the catheter system and includes neurological complications, infection, fibrosis and inflammation.
  
  
• The decision to embark upon long-term intraspinal drug administration should not be undertaken lightly and patients should undergo careful evaluation and psychological screening. Potential candidates should then undertake a screening trial before pump placement. Treatment is usually started with morphine and if this fails, other agents may be used, based on the physician's judgement as no studies support the use of one agent over another.
  
  
• The cost of therapy is a highly controversial topic and is complicated by the differing viewpoints of the patient, the provider and the payer.
  
  
• Cost-effectiveness studies show that base-costs for intrathecal drug therapy using an implanted system are less than the costs of medical management over a period of a few years; implanted intrathecal delivery was the least expensive, followed by oral and transdermal delivery and then by epidural and subcutaneous delivery through an external infusion pump. It is yet to be determined if there are cost savings over the life of the patient, although the studies suggest savings. However, only a portion of the picture is captured with these studies and there may be far-reaching significance to the patient, payer, provider and society as a whole.

• The paper reviews the new agents that are being evaluated for possible intrathecal delivery in the treatment of chronic pain, and the development of new local drug-delivery systems to treat neurological diseases.
  
  
• The a2-adrenergic agonists, clonidine, epinephrine and norepinephrine have all been shown to have analgesic effects, and to act synergistically with the opioids. Intrathecal allografts of adrenal tissue, or encapsulated chromaffin cells can be used to provide continuous delivery of the catecholamines. Encapsulation technology may provide new means to administer drugs from natural or genetically modified cells.
  
  
• Intrathecal delivery of local anaesthetics is now commonly used to treat chronic pain. The anaesthetics are usually combined with opioids to improve pain control.
  
  
• Somatostatin itself has been shown to be neurotoxic in animal models. The analogue, octreotide does not appear to be neurotoxic in humans, and has been given intrathecally for the relief of cancer and non-cancer pain.
  
  
• Preliminary findings suggest that the selective N-type calcium channel blockers, particularly SNX-111, are a promising new class of analgesics for intrathecal delivery. However, the adverse effects of SNX-111 require further investigation.
  
  
• Preliminary data with neostigmine suggest it is a relatively safe drug for intrathecal delivery, and that further investigation of its efficacy and safety are warranted.
  
  
• The NMDA antagonist, ketamine, has been used as an intrathecal and epidural therapy to relieve perioperative and chronic pain. However, a major disadvantage of this class of drugs is their PCP-like adverse effects, and the fact that some of these agents may be neurotoxic.
  
  
• The GABAB agonist, baclofen, and GABAA agonist midazolam have both been used in the treatment of pain, and the combination of midazolam and clonidine appears promising in treating non-cancer pain. Further research is needed into the risk:benefit ratio for these drugs in the long-term.
  
  
• New disease states that may be targeted by intrathecal therapy include neurodegenerative disorders that may benefit from the use of neurotrophic agents, for example, amyotrophic lateral sclerosis, Parkinson's disease and Alzheimer's disease.
  
  
• New drug-delivery systems under development include those for intratumoral administration of interleukin-4 toxin to glioblastomas. Alternative approaches are also being explored to increase the delivery of antiviral therapies to the brain in HIV-related neurological conditions.

• Use of non-opioid agents for intrathecal infusion to manage chronic pain is at the forefront of current pain research. Compounds currently under investigation include clonidine, bupivacaine, ocreotide and SNX-111, which are reviewed here. Other treatments that may prove promising include neostigmine, NMDA antagonists, GABA agonists, midazolam and cells in matrix.
  
  
• Clonidine is an a2-adrenergic agonist, shown to be effective in treating both cancer and neuropathic pain. There is extensive information on the epidural administration of clonidine but less data on its intrathecal use. However, data indicate that intraspinal delivery of a2-adrenergic agonists may be more effective than opioids in the treatment of neuropathic pain. Opioids and a2-adrenergic agonists also act synergistically in the spinal cord, so they can be used as admixtures to treat chronic pain.
  
  
• Use of the long-acting local anaesthetic bupivacaine has been reported in several clinical trials and case studies. It is effective in treating both neuropathic and nociceptive pain. However, its use is limited by side-effects, such as sensory and motor blockade, and haemodynamic instability. Liposomal encapsulation of local anaesthetics reduces toxic side-effects, so is a promising development.
  
  
• Somatostatin produced pain relief in a small study, but was also associated with histopathological changes on autopsy. The somatostatin analogue, octreotide, has shown promising results in patients with cancer pain. Data indicate that this drug may be a useful alternative to opioids for intrathecal delivery.
  
  
• Selective antagonists of the N-type calcium channel, such as the synthetic molecule SNX-111, have been shown to reduce both nociceptive and neuropathic pain. This class of drugs therefore shows promise as intrathecal therapy for chronic pain.

• This paper reviews the adverse events and complications of intrathecal opioid delivery, and their management.
  
  
• The pharmacological side-effects of systemic opioids have been well characterised. The acute effects include pruritus, nausea and vomiting, and urinary retention. Constipation, sexual dysfunction and endocrine changes can be long-term side-effects. Intrathecal administration produces similar side-effects, but less sedation and constipation.
  
  
• Therefore, switching from systemic to intrathecal therapy may be associated with a reduction in side-effects. Switching from morphine to other opioids, such as fentanyl, sufentanil or buprenorphine, or the addition of a non-opioid agent, may also reduce side-effects.
  
  
• Many side-effects can be avoided by slow dose titration, adequate prophylactic co-medication and thorough patient screening. Management of the side-effects involves reducing opioid dose, administration of adjuvant medications or, in severe cases, administration of an opioid antagonist.
  
  
• Implantation of a pump for intraspinal opioid delivery must be done with the highest level of care to minimise surgical complications, which may include bleeding, infection, tissue damage and CSF leaks.
  
  
• Most of the delivery-system complications involve the catheter, not the pump. Catheter complications can include dislodgement, kinks, breakage, occlusion or migration. Resulting problems may include drug underinfusion (resulting in loss of analgesia, inconsistent analgesia or withdrawal symptoms); delivery of drug into the pump pocket or subcutaneous area; free-floating catheter fragments in the CSF; CSF leakage; or spinal-cord damage.
  
  
• Complications associated with programmable pumps can include: overfilling of the pump, battery failure, pump failure and pump torsion. Pump failure is always associated with a loss of analgesia. However, this also may have several other aetiologies.
  
  
• The development of improved catheters will help to prevent many delivery-system complications.
  

• The purpose of the paper is to review studies that have evaluated multiple outcome measures to determine the efficacy of intrathecal opioids for the treatment of cancer and non-cancer pain.
• In a study of 26 patients with non-cancer pain followed up for an average of 23 months, patients showed an average of 39.2% reduction in pain at 6 months and 44.9% reduction at 12 months. Most patients (77%) rated their satisfaction with treatment as good to excellent.
• A study of 120 patients with non-cancer pain showed reductions in pain at the first follow-up point of 6 months. Overall, 74.2% of patients benefited from continuous intrathecal opioid delivery over a period of 0.5-5.7 years; tolerance was observed in only 7 patients and psychological dependence was not observed.
• A study of 32 patients with non-cancer pain followed up for at least 6 months showed that patients experienced a significant reduction in global pain and an improvement in well-being.
• A large study sent out questionnaires to physicians using intrathecal pumps, and received replies from 35 physicians with data from 429 patients, some of whom had cancer pain and some non-cancer pain. Most patients reported at least some increase in activities of daily living, and almost all patients reported some relief of pain.
• In a study of 54 patients with cancer pain, followed up for between 2 months and 2 years, 91% of patients reported a satisfactory level of analgesia.

• This joint consensus statement was developed by the two societies in response to state-level policy developments on the use of opioids in pain treatment. It is designed to encourage dialogue with regulators about the appropriate relationship between the law and practice of pain medicine.

• Pain is often inadequately treated, and common reasons for withholding effective therapy are often based on unfounded fears.
  
  
• Development of addiction to opioids used for pain relief is low.
  
  
• Respiratory distress and other side-effects are usually short-lived.
  
  
• Tolerance is not generally a limitation of long-term opioid use.
  
  
• The risk of diverting prescribed opioids can be reduced by careful monitoring.
  
  
• The guidelines for prescribing opioids should be an extension of good clinical practice.
  
  
• Patients should be fully evaluated for their pain history, diagnosis, previous treatments etc.
  
  
• Treatment plans should be tailored to both the individual patient and the presenting problem, and made in consultation with relevant experts.
  
  
• Patients and/or care partners should be made fully aware of the risks and benefits of opioid use, and the conditions under which they will be prescribed.
  
  
• Treatment efficacy should be reviewed periodically to assess its risks/benefits and the need for continued therapy. ¨

Patients and treatments should be fully documented to ensure effective monitoring.

• Scrupulous patient selection is a critical factor in the prediction of successful clinical outcomes for intrathecal pain therapy.
  
  
• The nature of the chronic pain needs to be taken in to consideration. As pain is such a multifactorial experience, with significant psychological and cognitive contributions, the nature and degree of these influences must be assessed to evaluate the likely risks and benefits of intraspinal therapy.
  
  
• A detailed history and complete medical examination will detect the presence of absolute and relative contraindications to intrathecal therapy, and provide important information about the patient's pain and previous treatment history.
  
  
• It is important that the patient's pain intensity level is systematically assessed, so the efficacy of the intervention can be measured. The type of pain the patient is experiencing should also be identified.
  
  
• Psychological factors to be taken into account include those that may be contributing to the pain itself, those that might influence treatment outcomes, and the patients' expectations of the therapy. Social factors can also strongly influence treatment outcome.
  
  
• Intraspinal therapy should only be used in patients who have failed on oral and systemic therapies. An adequate trial of intraspinal therapy should then be conducted to assess treatment efficacy and control for a possible placebo effect. A positive response on such a trial is generally defined as a minimum 50% reduction in pain with no intolerable side-effects.
  

• A retrospective study reviewed data from 429 patients of 35 physicians using intrathecal therapy. Physician reports of global pain-relief scores were excellent or good in a total of 95.2% of patients. The mean percentage pain relief was 61%.
  
  
• In a second retrospective study of intrathecal therapy in patients with non-cancer pain, the average pain intensity decreased from 93.6 to 30.5 after 6 months. This was accompanied by improvements in social functioning and quality of life.
  
  
• An unpublished prospective study of this therapy showed improvements or trends towards improvements that persisted throughout the 24-month follow-up.
  
  
• The short-term side-effects of intrathecal therapy are those typically related to opioid administration, including pruritus, nausea and vomiting, urinary retention, and constipation. In general, all are amenable to symptomatic therapy and all except constipation typically disappear within a few days.
  
  
• Care must also be taken during the implantation of the pump to minimise bleeding, the possibility of infection, tissue damage, and leaks of cerebrospinal fluid.
  
  
• Mechanical catheter complications can include breaking, kinking, disconnections, catheter-tip obstruction and dislodgement, and accidental withdrawal. Complications of programmable pump delivery systems may include overfilling, battery failure, pump failure and pump torsion. All these problems are likely to be indicated by a loss of analgesic effect.
  

• The aim of the study was to investigate current practice in the use of implantable, programmable, drug-administration pumps in patients with severe pain.
  
  
• Physicians who had implanted more than five SynchroMed pumps were identified. They were asked to fill out questionnaires based on their practice and on each patient they had treated for more than 8 months.

• A total of 35 physicians responded, reporting excellent or good pain relief scores in 95% of 429 patients treated for a mean of 14.6 months (range 8-94 months).
  
  
• 57% of patients experienced a moderate or great increase in ADL, 82% some increase and 28 patients returned to work.
  
  
• Mean pain relief was 61%; patients with somatic pain tended to have greater relief than those with other types of pain.
  
  
• The most common side-effects were nausea and vomiting, pruritus and diaphoresis. Delivery-system complications occurred in 82 patients (21.6%). The most common were catheter-related (28 patients).

• In the absence of well controlled, prospective trials, this retrospective survey provides preliminary evidence for the efficacy of intrathecal morphine to treat severe pain.
  
  
• However, many questions remain, and can only be answered by well designed, prospective, controlled trials comparing intraspinal opioid therapy with systemic administration.

• Surgically implanting a drug-delivery system is invasive and has risks. However, the delivery system is not permanent and does not preclude other treatment approaches. In addition, motor function is not compromised by the procedure.
  
  
• The initial costs of fitting an implantable delivery system seem high, but in the long term, trials have indicated that such systems are more cost-effective than oral or systemic therapies.
  
  
• The ethics of treating chronically or terminally ill patients with high-tech therapies remains controversial, and some physicians prefer more conservative therapies. Patient selection is critical, and all the positive and negative issues involved in these therapies should be considered.
  
  
• In the future, drug-pump technology may well be developed that allows greater flexibility. For example, medication doses may be initiated or changed by telemetry via telephone.
  Future applications of the technology include delivery of large molecules to the CNS, for example giving neurotrophic agents in amyotrophic lateral sclerosis.

• When opioids are administered systemically or intraspinally, they bind to opioid receptors presynaptically on the primary afferent neurons, inhibiting neurotransmitter release and therefore transmission of the pain message.
  
  
• The rationale for intraspinal administration of opioids is that the delivery of drugs directly to their site of action would allow the use of smaller doses of drugs, resulting in fewer side-effects than systemic therapy. Most patients receiving intrathecal opioids experience less sedation and constipation than those taking systemic drugs.
  
  
• The criteria for choosing intrathecal opioid therapy include the presence of intractable pain (not controlled despite aggressive use of systemic opioids and adjuvant drugs), or the development of severe side-effects from systemic drugs that do not respond to appropriate treatment.
  
  
• The location of the pain (originating from nerve roots below the cervical spine) and appropriate psychosocial factors should be assessed before choosing this therapy.
  
  
• A trial of the drug to be used in long-term therapy is necessary before an intraspinal delivery system is implanted. However, the type of screening used varies. Continuous epidural infusions most closely emulate the final therapy, but prolongs hospitalisation and therefore increases cost. Bolus administration requires a shorter screening period, but may produce pronounced side-effects and greater therapeutic effects than the continuous intraspinal delivery.
  
  
• During the immediate postoperative period, the medications injected into the implanted pump can be infused immediately, with titration starting soon afterwards. Because the necessary doses are difficult to predict, frequent titration is usually needed. Once the patient is discharged from hospital, titration is dependent of their reports of their condition. Often titration is needed soon after discharge, as the patient becomes more active.
  
  
• In the postoperative period, patients will also require supplemental opioids to treat postoperative pain. Supplemental therapy is essential to treat breakthrough cancer pain. The chronic use of supplemental opioids in patients with non-cancer pain is controversial among some physicians.
  
  
• Patient education is needed to ensure good therapeutic outcomes.

• This study aimed to address the following issues:
  (i) How chronic infusion of intrathecal morphine should be used in the overall management of cancer pain; (ii) what role it may have in treating pain of benign origin; and (iii) whether the incidence of tolerance with long-term infusion is significant.
  
  
• The study population consisted of 43 patients with intractable pain (35 with pain due to cancer and 8 with pain of benign origin).
  
  
• Following a successful trial with epidural morphine, patients received an Infusaid constant-flow pump or a Medtronic programmable device implant to deliver intrathecal morphine.
  
  
• Patients were followed for at least 6 months. Subjective pain reports, objective changes in daily living, and information regarding oral narcotic use, were used to determine patient response to chronic intrathecal morphine.

• Eighty percent of patients with cancer pain, and all patients with pain of benign origin had a good or excellent response to intrathecal morphine infusion, characterised by a reduction in the use of oral narcotics, a decrease in pain ratings and improvement in activities of daily living.
  
  
• Many patients with malignant pain required increasing doses of intraspinal morphine. Increasing the dose did not result in any central side effects.

• The implanted Medtronic programmable devices all functioned properly. In contrast, two of the Infusaid devices failed.
  

• In this study, Medtronic programmable devices proved more reliable than the Infusaid constant flow pumps.
  
  
• The effective, sustained relief from cancer pain achieved with chronic intrathecal morphine infusion warrants its earlier application in patients with malignant pain. It should not be used as a last resort in patients who are likely to die soon - external pumps are more appropriate in such patients.
  
  
• Patients with severe pain of benign origin are also candidates for chronic intrathecal morphine infusion. However, whether the gains are sufficient to employ this continuous technique remains an issue. Larger, long-term studies are required.
  
  
• Increasing the dose of intraspinal morphine does not result in any central side effects.
  
  
• The issue of tolerance does not seem to be a major complication of chronic intrathecal morphine infusion.