HYPERBARIC OXYGEN FOR MULTIPLE SCLEROSIS. A CRITICAL APPRAISAL BY META-ANALYSIS.

Michael H. Bennett , Robert Heard

Dept. of Diving and Hyperbaric Medicine, Prince of Wales Hospital, Australia

introduction: Despite a number of RCTs investigating the efficacy of HBOT for MS, controversy still surrounds this indication. Although not appearing in the UHMS, ANZHMG or EUBS lists, a network of chambers in the UK have reported over one million treatments and other centres continue to consider such patients. The aim of this quantitative review is to provide guidance for clinicians asked to accept such patients for treatment.

methods: A standardised search was instituted to reveal all RCTs involving hyperbaric oxygen as therapy for MS. Sources included: Cochrane Controlled Trials Register, databases of the National Library of Medicine (PubMed) and Randomised Controlled Trials in Hyperbaric Medicine, targeted journals and proceedings of major meetings in Hyperbaric Medicine and references of articles found above. Sensitivity analyses were by study quality and oxygen dosage. All analyses were made using Revman 4.0.3 software employing a Peto fixed-effects odds ratio.

results: 10 studies met the inclusion criteria. Outcomes analysed include Kurtzke Extended Disability Status Score (EDSS), Functional Status Score and relapse rate. The OR for EDSS improvement on completion of HBO course was 2.93 (95%CI 0.71-12.17), while for bladder/bowel sphincter function improvement the OR was 1.40 (95%CI 0.80-2.43). Relapse occurred within one year in 20/80 HBO patients versus 28/77 control patients. Only two trials measured this outcome, making meta-analysis unhelpful. Treatment side-effects were minimal.

  Fig 1. Summary of EDSS and Sphincter Improvement  

Favours control  (OR) Favours HBO      Favours control (OR) Favours HBO  

      EDSS Improvement              Bladder/bowel Improvement

conclusion: Meta-analysis failed to identify a significant benefit of HBO in MS. There is a trend to better outcome, but this remains non- significant despite 10 randomised trials and is unlikely to be large. There is very little evidence to test the claim of long-term benefit from regular maintenance treatment. There is a case for mounting larger randomised, preferably multi-centre, trials, but insufficient evidence to warrant routine treatment of such patients outside experimental protocols.

Fax: 61 2 9382 3882

Email: m.bennett@unsw.edu.au