Surgery for disc disease

BMJ 2007; 335 doi: https://doi.org/10.1136/bmj.39315.580637.BE (Published 08 November 2007) Cite this as: BMJ 2007;335:949
  1. J N Alastair Gibson, consultant spinal surgeon
  1. Royal Infirmary of Edinburgh, Edinburgh EH16 4SU
  1. j.n.a.gibson{at}blueyonder.co.uk

    New evidence supports its use in selected patients

    Spinal surgeons have striven to underwrite their surgical practice by sound evidence from clinical trials, yet data of adequate quality have not always been available. This has led to clinical dilemmas with respect to the simplest of questions, such as when should surgery be recommended for acute disc prolapse and, in degenerative disc disease, whether surgery is more effective than extended non-operative treatments? Recent trials answer these questions.

    The authors of the 2007 Cochrane review of surgical interventions for lumbar disc prolapse1 conclude that surgical discectomy for carefully selected people with sciatica provides faster relief from the acute attack than conservative management. However, it was unclear whether surgery had any positive or negative effects on the natural history of the underlying disc disease. This conclusion was based primarily on one unblinded study published in 1983,2 in which around a quarter of people treated conservatively crossed over to surgery (although there was an intention to treat analysis). Patient and observer ratings showed that discectomy produced significantly better relief of low back pain and sciatica than conservative treatment at one year, although these differences were not maintained at four and 10 years. Importantly, the trial also showed that postponing surgery to further assess clinical progress delayed recovery but did not cause long term harm. Discectomy was highly cost effective, at around $29 000 (£14 600; €21 500) per quality adjusted life year gained.3 However, neither this trial, nor two other trials,4 5 considered the optimal timing of surgery.

    This matter is dealt with in a recently published randomised trial of 283 people who had had severe sciatica for six to 12 weeks. This trial compared early surgery (mean 2.2 weeks) with prolonged conservative treatment and surgery if needed (mean 18.7 weeks).6 Outcomes at one year were similar in both groups, but pain relief and perceived recovery were faster for people who had early surgery. This suggests that there is a case for avoiding delay in surgery, although people with a mild motor deficit who choose not to have surgery will probably not develop a progressive deficit.7

    Two year data from the multicentre US spine patients outcomes research trial (SPORT) are also now available.8 This trial was designed to assess the relative efficacy and cost effectiveness of surgical and non-surgical approaches for treating the three most common conditions for which spinal surgery is performed—disc herniation, degenerative spondylolisthesis (vertebral slip forwards of one lumbar vertebra on another with an intact neural arch), and spinal stenosis (degenerative narrowing of the spinal canal). Data from 501 people with disc herniation were reported.8 Although high rates of crossover (surgery was performed in only half of people assigned to it and in just under a third of the non-surgical group within three months) led to inconclusive results in the intention to treat analysis, the as treated analysis showed strong, statistically significant advantages of surgery on all outcomes for up to two years of follow-up. Back pain improved in both non-operative and surgical groups, but the improvement was significantly greater in people who had surgery. The result was consistent for all herniation locations and morphologies.

    In total, 607 people were enrolled in the degenerative spondylolisthesis and spinal stenosis cohort.9 One year crossover rates were high in the randomised group (about 40% in each direction), and an intention to treat analysis found no statistically significant effects for primary outcomes. However, as treated analysis for both cohorts combined showed significant advantages for surgery at one year, which diminished only slightly by two years. Symptoms improved quickly (as early as six weeks) in patients who had surgery. Greater improvements in spinal function and patient satisfaction were also seen at two years in people treated surgically included in the third arm of the SPORT trial (spinal stenosis without degenerative spondylolisthesis versus non-operative treatment).10

    These results are encouraging and generally in line with those from a randomised trial published earlier this year that included people with and without degenerative spondylolisthesis.11 What the studies do not adequately look at is the extent of spinal fusion, if any, that should accompany decompression of the spine to relieve symptoms from spinal stenosis and nerve root compression. Recent data from a randomised controlled trial in Edinburgh suggest that people with foraminal stenosis and single level degenerative disc disease may be better treated by decompression alone, with the proviso that further surgery may be needed at a later date.12 If the surgeon does elect to proceed to fusion then stabilisation of the spine by pedicular instrumentation has been shown to promote a higher fusion rate, but with marginal benefit in terms of clinical outcome.13

    Spinal surgeons have been particularly proactive in the past five years in publishing randomised trials. These latest trials support the common practice of early referral by the primary care doctor of people with acute sciatica who are “failing” non-operative treatment. A surgical option should also be explored for people with progressive lumbar spinal stenosis. The ongoing challenge for surgeons is to produce evidence that supports or discredits new technical advancements in minimal intervention surgery, disc arthroplasty, and procedures to correct spinal deformity. These technologies are generally driven by commercial interests and at a premium cost to health providers. They should be introduced to spinal centres, where they can be tested rigorously against other available treatments. Producing this evidence within a reasonable time frame will require multicentre collaboration as used in the SPORT trial.


    • Competing interests: None declared.

    • Provenance and peer review: Commissioned; not externally peer reviewed.


    View Abstract

    Log in

    Log in through your institution


    * For online subscription