Article Text

Download PDFPDF
Comparative Efficacy of Epidural, Subarachnoid, and Intracerebroventricular Opioids in Patients With Pain Due to Cancer
  1. Jane C. Ballantyne, M.B., B.S., F.R.C.A*,
  2. Daniel B. Carr, M.D,
  3. Catherine S. Berkey, D.Sc,
  4. Thomas C. Chalmers, M.D§ and
  5. Frederick Mosteller, Ph.D
  1. *Department of Anesthesiology, MGH Pain Center, Massachusetts General Hospital, Boston, the
  2. Departments of Anesthesiology and Medicine, Tufts University School of Medicine and New England Medical Center, Boston, the
  3. Technology Assessment Group, Harvard School of Public Health, Boston, and the
  4. §Division of Clinical Care Research Department of Medicine Tufts University School of Medicine and New England Medical Center, Boston and Metaworks, Inc., Boston, Massachusetts
  1. Reprint requests: Jane C. Ballantyne, M.B., B.S., F.R.C.A., Department of Anesthesia, Massachusetts General Hospital, Fruit Street, Boston, MA 02114.


Background and Objectives Although rarely used, intracerebroventricular opioid therapy (ICV) is an option for the control of intractable pain due to cancer when systemic treatments have failed. The aim of the present study is to use available data from published trials to compare ICV with the more common epidural (EPI) and subarachnoid (SA) opioid treatments in an attempt to establish the utility and safety of ICV.

Methods Because there are no published controlled trials comparing these routes of administration, the combined data from multiple uncontrolled trials were used, with differences between the treatments analyzed statistically. Trials assessing ICV (13 trials, 268 patients), EPI (29 trials, 909 patients) and SA (21 trials, 410 patients) in cancer patients were identified; data on analgesic efficacy, common pharmacologic side effects, and complications were then extracted and the accumulated incidence data analyzed.

Results The findings (weighted means) indicated ICV to be at least as effective against pain as other neuraxial treatments, with 75% of ICV-treated patients obtaining excellent pain relief as compared with 72% of EPI- and 58% of SA-treated patients (not significant). The failure rate of both spinal treatments tended to be greater than that of ICV and was significantly higher in the case of EPI (P = .045). In general, persistent side effects appeared to be more of a problem with the spinal treatments, while transient symptoms occur more often with ICV. Persistent nausea, urinary retention, and pruritus all were more frequent with the two spinal treatments than with ICV, but transient nausea and respiratory depression occurred more often with ICV. Sedation and confusion appeared to occur more often with ICV than with spinal therapy, while constipation and headache were rarely encountered with ICV. There were no real differences in infectious complication rates among the three treatments (except for a lower rate of infection when an implanted pump was used), but technical problems such as catheter blockage, misplacement, or leakage tended to occur less often with ICV.

Conclusions Intracerebroventricular therapy appears to be at least as effective against pain as other neuraxial treatments. The ICV technique is the only fixed system that is associated with fewer technical problems than the use of simple percutaneous epidural catheters (difference 9%, standard error of the difference 3.4). The present state of evidence indicates that ICV is a successful treatment for patients with intractable cancer pain and compares well with spinal opioid treatments in terms of efficacy, side effects, and complications.

  • epidural opioid
  • subarachnoid opioid
  • intracerebroventricular opioid
  • cancer pain

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • Thomas C. Chalmers is deceased.