Comparative efficacy of epidural, subarachnoid, and intracerebroventricular opioids in patients with pain due to cancer

  • Review
  • Intervention




Since the 1970s, when endogenous opioids and opioid receptors were first isolated in the central nervous system, attempts have been made to optimize opioid therapy by delivering the medication centrally rather than systemically. Although the vast majority of cancer patients obtain satisfactory pain relief from individualized systemic treatment, there remain the few whose pain is refractory to systemic treatments. These patients may obtain relief from neuraxial opioid therapy: intracerebroventricular, epidural or subarachnoid.


To compare intracerebroventricular therapy with other neuraxial treatments and to determine whether intracerebroventricular (ICV) has anything to offer over epidural (EPI) and subarachnoid (SA) catheters in terms of efficacy, adverse effects, and complications.

Search methods

A number of electronic databases were searched to retrieve information for inclusion in this review up to January 2003. Non-English language reports are awaiting assessment. Unpublished data were not sought.

Selection criteria

Randomised studies of intracerebroventricular therapy for patients with intractable cancer pain were sought. However, this level of evidence was not available so data from uncontrolled trials, retrospective case series and uncontrolled prospective cohort studies were assessed.

Data collection and analysis

Our search did not retrieve any controlled trials. We therefore used data from uncontrolled studies to compare incidences of analgesic efficacy, adverse effects, and complications. We found 72 uncontrolled trials assessing ICV (13 trials, 337 participants), EPI (31 trials, 1343 participants), and SA (28 trials, 722 participants) in cancer patients. From these we extracted data on analgesic efficacy, common pharmacologic adverse effects, and complications.

Main results

Data from uncontrolled studies reported excellent pain relief among 73% of ICV patients compared with 72% EPI and 62% SA. Unsatisfactory pain relief was low in all treatment groups. Persistent nausea, persistent and transient urinary retention, transient pruritus, and constipation occurred more frequently with EPI and SA. Respiratory depression, sedation and confusion were most common with ICV. The incidence of major infection when pumps were used with EPI and SA was zero. There was a lower incidence of other complications with ICV therapy than with EPI or SA.

Authors' conclusions

Neuraxial opioid therapy is often effective for treating cancer pain that has not been adequately controlled by systemic treatment. However, long-term use of neuraxial therapy can be complicated by problems associated with the catheters. The data from uncontrolled studies suggests that ICV is at least as effective against pain as other neuraxial treatments and may be a successful treatment for patients whose cancer pain is resistant to other treatments.

Plain language summary

Effectiverness of epidural, subarachnoid and intracerebroventricular opioid drugs to treat patients with pain due to cancer

Cancer patients who do not obtain pain relief from treatment with opioids administered by mouth, rectally or by injection may do so if the drugs are administered in other ways. This review compared three alternative routes: intracerebroventricular (where opioids are injected through a small hole bored into the skull); epidural (where opioids are introduced into the epidural space in the spine using a catheter); and subarachnoid (where opioids are introduced into the subarachnoid space in the spine using a catheter). The evidence from uncontrolled studies showed that giving opioids intracerebroventricularly was more effective for pain relief than either epidural or subarachnoid administration. Adverse effects and complications were reported for all three procedures.