Lumbar Disc Replacement

The disc between two vertebrae acts as a cushion and allows a small amount of movement. When a disc becomes worn and is judged to be the source of persistent low back pain, one long-standing surgical option is fusion, which stops movement at that segment. The concern with fusion is that, by stopping movement at one level, it may place extra load on the discs above and below — so-called adjacent-segment disease — potentially leading to further problems years later. Total disc replacement was developed on the theory that keeping the segment moving, rather than fusing it, might relieve pain while protecting the neighbouring discs. It is a logical idea, and it is the central argument made in its favour. Whether it actually delivers that protection in the long term is precisely where the debate lies.

Several manufacturer-sponsored randomised trials, including the FDA investigational-device trials of the earlier Charité and ProDisc devices, reported that disc replacement was at least as good as fusion at two years for carefully selected patients with single-level disc degeneration, and some five-year follow-ups reported that these results were maintained. Meta-analyses of these trials have broadly concluded non-inferiority to fusion in the short-to-medium term. That is the evidence most often cited in favour. However, the more sceptical reading — and it is a legitimate one — is that much of the strongest evidence comes from industry-sponsored trials, that follow-up is often shorter than the decades over which adjacent-segment problems would actually emerge, that loss of patients to follow-up limits certainty, and that at least one independent randomised trial found only marginal or no clear benefit over fusion. Genuinely long-term data are scarce. The fair summary is that disc replacement can produce good results in well-chosen patients, but that the central promise — long-term protection of the adjacent discs — has not been convincingly proven, and the evidence base leans heavily on trials with commercial involvement.

Both supporters and critics agree on one thing: the results depend enormously on choosing the right patient. Disc replacement is generally considered only for a narrow group — typically younger patients, with pain from a single degenerated disc, good bone quality, no significant arthritis of the small facet joints at the back of the spine, no instability, and no deformity. A great many patients with back pain fall outside these criteria, and using the device in the wrong patient is a recognised route to a poor outcome. This is one reason the operation is offered far less commonly than fusion, and it is part of why appropriateness has been debated by health-technology bodies.

A specific and important concern with any artificial disc in the lower back is what happens if it fails or needs to be removed. Because the lumbar disc is approached from the front of the abdomen, revising or removing a device means operating again through scar tissue near the large blood vessels that lie directly in front of the spine — a technically demanding procedure with real risks. This is a meaningful consideration that is sometimes underplayed in enthusiastic accounts of the technology, and it is one of the reasons caution is warranted.

I do not perform lumbar disc replacement. My reasons are partly about my own practice and partly about my reading of the evidence: I am not persuaded that the long-term advantage over well-performed alternatives has been established, and I believe patients are best served by surgeons who perform a given operation regularly and are fully committed to it. I would not want anyone to take this page as a blanket judgement that disc replacement is wrong — for the right patient, in the right hands, it clearly can work. But it is a contested technology, not a settled one, and anyone considering it deserves a frank account of the uncertainty as well as a balanced second opinion.

• —Blumenthal S, et al. Spine. 2005. Charité FDA IDE RCT vs fusion, Part I: non-inferiority at 2 years. PMID 16025024
• —McAfee PC, et al. Spine. 2005. Charité FDA IDE RCT, Part II: outcomes depend on placement accuracy. PMID 16025025
• —Zigler J, et al. Spine. 2007. ProDisc-L FDA IDE RCT vs fusion: non-inferiority at 2 years. PMID 17495770
• —Guyer RD, et al. Spine J. 2009. Charité 5-year RCT follow-up. PMID 18805066
• —Putzier M, et al. Bone Joint J. 2025. ALIF vs disc arthroplasty RCT, mean 14-year follow-up. PMID 40449550
• —Sköld C, et al. Eur Spine J. 2013. Independent Swedish RCT: marginal/no clear benefit over fusion at 5 years. PMID 23893083
• —Radcliff K, et al. Clin Spine Surg. 2018. 5-year reoperation rates, 2-level disc replacement vs fusion. PMID 28005616
• —Zigler JE, et al. Spine. 2018. Progression of adjacent-level degeneration after disc replacement. PMID 29570121
• —Zigler J, et al. Global Spine J. 2018. Meta-analysis of 5-year RCT outcomes vs fusion. PMID 29977727
• —Lang SAJ, et al. Eur Spine J. 2021. Meta-analysis: disc replacement vs fusion. PMID 33837832
• —Thavaneswaran P, et al. ANZ J Surg. 2014. Systematic review: limited evidence, appropriateness questioned. PMID 23869616
• —Wagner WH, et al. J Vasc Surg. 2006. Vascular access for revision/removal of a lumbar artificial disc. PMID 17145428