Lumbar Disc Arthroplasty (Total Disc Replacement)

Lumbar total disc replacement (TDR) is a motion-preserving alternative to spinal fusion for degenerative disc disease (DDD). The aim is to relieve discogenic axial low back pain while keeping segmental motion at the affected level, theoretically reducing the adjacent segment disease (ASD) that commonly follows fusion.

History. The concept dates to the 1960s and Fernström's stainless steel ball-and-socket designs. Modern TDR evolved in the 1980s and 1990s with better biomechanical understanding and materials science. The Charité artificial disc (SB Charité III, DePuy Spine) was the first FDA-approved device, in 2004, followed by ProDisc-L (Synthes) in 2006; several other designs hold European CE marking but not FDA approval.

The degenerate disc as a pain generator. A healthy disc maintains an organised annulus, a hydrated nucleus and intact endplate exchange; degeneration brings matrix loss, annular disorganisation and impaired nutrient transport. Annular and endplate disruption combine with inflammatory signalling, neovascularisation and nociceptive ingrowth, which explains why a morphologically abnormal disc can become a pain generator.

View larger

View larger

View larger

The case against fusion. Adjacent segment disease is well documented after fusion: radiographic change develops in 8-30% of patients at 10 years, and clinical ASD requiring reoperation in 2-4% per year. The proposed mechanisms:

• Increased intradiscal pressure at the adjacent levels, up to 45%
• Altered kinematics, with hypermobility compensating
• Facet joint overload from loss of normal load sharing
• Accelerated disc degeneration from abnormal stress concentration

What TDR is meant to do. It aims to maintain physiological motion, 5-15° of flexion-extension per level, preserve the normal distribution of intradiscal pressure and reduce stress transfer to the adjacent segments. Whether it truly prevents ASD compared with fusion remains controversial; the benefit is still theoretical, and data beyond 10 years are still accumulating.

What the trials show. Multiple Level I randomised trials have compared TDR with fusion, and the FDA Investigational Device Exemption (IDE) trials for Charité and ProDisc-L established it as a viable alternative for highly selected patients. Across them:

• Clinical outcomes were non-inferior at 2, 5 and 10 years
• Pain relief and functional improvement were equivalent
• Motion was preserved at the index level, averaging 7-10°
• Overall complication rates were similar, but the complication profiles differed
• Adjacent-level reoperation was not significantly lower (1.9% vs 4.0%, p = 0.68); only the radiographic degeneration score differed

The designs divide by their bearing: a free polyethylene core, a fixed polyethylene inlay, or metal on metal.

View larger

The approved indication. Symptomatic degenerative disc disease at L4-5 or L5-S1. Charité and ProDisc-L were approved for a single level, and ProDisc-L was later approved for two.

The criteria. The patient who qualifies has:

• Failed a minimum of 6 months of non-operative management, including physiotherapy, medication and injections
• Discogenic pain, shown by concordant pain provocation on discography or a high-intensity zone (HIZ) on T2-weighted MRI
• Maintained disc height of at least 4mm (some implants require at least 5mm)
• An age typically of 18-60 years, the upper limit varying by study and surgeon preference
• No facet arthropathy: less than Grade 2 facet degeneration on imaging
• No posterior element disease: an intact pars and no spondylolisthesis
• Skeletal maturity, with closed physes confirmed

View larger

Absolute contraindications. Each of these excludes the patient from TDR:

1. Facet joint arthropathy, moderate to severe (Fujiwara Grade 2 or higher). TDR does not address facetogenic pain.
2. Spondylolisthesis. Motion preservation is inappropriate with baseline instability. Isthmic spondylolisthesis at any grade excludes TDR; for degenerative slips the threshold quoted ranges from any grade to greater than Grade I.
3. Spinal stenosis: central canal stenosis requiring decompression, lateral recess or foraminal stenosis. TDR does not decompress.
4. Previous surgery at the target level: discectomy, laminectomy or fusion. Scar tissue and altered anatomy preclude safe placement.
5. Osteoporosis or severe osteopenia: a T-score below −1.5 on DEXA, or bone too poor for implant fixation, with a high risk of endplate subsidence and implant migration.
6. Active infection: discitis, osteomyelitis or systemic infection, which contraindicate a metal implant.
7. Severe disc collapse, with less than 4mm of disc height remaining: too little space for the implant, and altered biomechanics.
8. Documented allergy to cobalt, chromium, molybdenum or titanium.

Relative contraindications. These raise the risk and call for an individual decision:

• Obesity, BMI over 35 kg/m², for approach difficulty and increased mechanical loading; morbid obesity, BMI over 40, brings vascular access challenges
• Active smoking, which impairs wound healing and may affect outcomes
• More than two levels of symptomatic DDD, which is not FDA approved
• Age over 60, leaving limited life expectancy for the benefit of motion preservation
• Workers' compensation, with worse outcomes in some studies, and ongoing injury litigation, associated with poor outcomes
• Psychological factors: untreated depression, chronic pain syndrome
• Pregnancy, current or planned within 2 years

View larger

View larger

The ideal candidate. Beyond the criteria, the ideal candidate is aged 30-50, with single-level disease, pain confirmed by provocative discography and facets of Grade 0-1, and also has:

• Normal sagittal alignment
• Normal bone density, a T-score above −1.0
• No smoking, and a BMI under 30
• No previous lumbar surgery
• Psychological health and realistic expectations
• No involvement in litigation or workers' compensation

Few patients meet every ideal criterion, and clinical judgement decides which deviations are acceptable.

Before offering TDR you must be confident the pain is discogenic and isolated. The commonest reason for TDR failure is an unrecognised or mislabelled pain generator, and the table triages the patient presenting with chronic low back pain.

Phenotypes. Vertebrogenic, annulogenic and mixed phenotypes have different anatomical pain generators. Separating them prevents selection of disc arthroplasty for pain dominated by endplate or neural pathology.

View larger

Vertebrogenic pain. The basivertebral nerve enters the posterior vertebral body and branches through the endplate region, an anatomical basis for vertebrogenic pain that can mimic discogenic pain. On MRI, Modic endplate change may indicate a vertebrogenic generator:

• Type 1: low signal on T1 and high signal on T2, reflecting active inflammatory endplate change that must be interpreted alongside the clinical pain phenotype
• Type 2: hyperintense on both T1 and T2, indicating fatty marrow replacement and a possible vertebrogenic rather than purely discogenic pain generator

View larger

View larger

View larger

View larger

Intraoperative

Vascular injury (1-3%). The most feared complication of the anterior approach. The iliac vessels are at highest risk, the vein more than the artery, and the left common iliac vein is particularly vulnerable at L4-5. Prevention is gentle retraction, good lighting and an available vascular surgeon; an injury needs immediate vascular surgery consultation, direct repair and possibly blood transfusion.

Visceral injury. Bowel perforation is rare if the peritoneum stays intact, and is recognised and repaired primarily. The ureter is identified, protected and retracted with the peritoneum. Bladder injury is rare, and more common with low transverse incisions.

Neurological injury. Nerve root injury is rare, a risk of posterior rather than anterior approaches. Sympathetic plexus injury causes retrograde ejaculation (below).

Malposition. Recognised on intraoperative fluoroscopy, and corrected by removing and repositioning the implant if it is significant. Checking with trial implants before the final one makes this easier.

Early (First 6 Weeks)

Retrograde ejaculation. Occurs in 1-5% at L5-S1 and in less than 1% at L4-5, higher with inexperienced surgeons or excessive dissection. Stretch or thermal injury to the hypogastric plexus leaves orgasm intact but ejaculation dry. It is usually permanent, so counsel the patient before surgery and refer to urology for fertility concerns; prevention is the careful handling of the plexus described in the approach.

Wound complications. Infection occurs in 1-2%, and prophylactic antibiotics are standard. Haematoma is rare because the retroperitoneal approach leaves less dead space, and dehiscence is rare with proper fascial closure.

Ileus. Occurs in 5-10%, from bowel manipulation during retraction. Management is conservative (nil by mouth, with a nasogastric tube if needed), and it usually resolves within 3-5 days.

Early subsidence. Subsidence in the first 6 weeks suggests osteoporosis or endplate violation. Serial radiographs monitor it, and more than 3mm is concerning. Asymptomatic subsidence is usually observed; progressive or symptomatic subsidence is treated by fusion.

Late (After 6 Weeks)

Heterotopic ossification (HO). Reported in 5-60% depending on the grading system, and clinically significant in 5-15%. It is ectopic bone formation, lying only in the surrounding soft tissue at Grade I and within or across the disc space at higher grades, and it is graded by McAfee (Grade 0-IV, below) or Brooker (Grade 1-4). Severe HO (Grade 3-4) may limit motion; Grade 1-2 is usually asymptomatic. Prevention is a postoperative NSAID (indomethacin 75mg daily for 6 weeks), with low-dose radiation controversial; established HO is observed if asymptomatic, and excision is rarely indicated.

Why HO grading matters. HO is the specific way a motion-preserving disc loses its motion and starts to behave like a fusion, the process behind the Putzier auto-fusion. Grading tells you whether the prosthesis is still moving. General HO biology and prophylaxis are covered in the heterotopic ossification topic; the McAfee scale, adapted from Ryu's cervical grading, is the disc-arthroplasty-specific grade, set by where the HO lies and what it does to prosthesis motion:

• Grade 0: no heterotopic ossification
• Grade I: HO present, but only in the surrounding soft tissue and not within the disc space
• Grade II: HO within the disc space, but not bridging it and not limiting prosthesis motion
• Grade III: HO bridging the disc space, with some motion still preserved on flexion-extension (a restricted range)
• Grade IV: bony ankylosis with no motion; the prosthesis has spontaneously fused

Grades 0-II preserve useful motion, and Grades III-IV, the "clinically significant" HO, progressively abolish it. High-grade HO negates the motion-preservation rationale, which is why postoperative NSAID prophylaxis is used and why long-term auto-fusion undermines the case for TDR.

Facet degeneration. Facet arthropathy may progress despite motion preservation. That suggests poor patient selection, with facet disease unrecognised before surgery, and symptomatic progression may require fusion.

Wear and loosening. Polyethylene wear is a long-term concern, beyond 10 years, and osteolysis from wear debris is rare but reported. Progressive endplate failure causes late subsidence, with osteoporosis a risk factor. Migration is very rare with modern designs, which use keel and teeth fixation.

Core dislocation (Charité). Subluxation or dislocation of the mobile core occurs in 1-3%, related to malposition at surgery (non-midline placement). It may cause recurrent pain or a clicking sensation, and lateral radiographs show the core position.

Adjacent segment disease. Despite the theoretical benefit, ASD still occurs after TDR, at a lower rate than after fusion in some studies. Radiographic ASD is seen in 10-20% at 5 years, and symptomatic ASD requiring surgery in 2-4% per year, similar to fusion.

Persistent or recurrent pain. Occurs in 10-30%, similar to fusion. The causes are facetogenic pain missed at selection, persistent discogenic pain and psychosocial factors. Work-up rules out implant-related problems (subsidence, malposition, HO), and management is conservative, with fusion if the TDR has clearly failed.

Revision

When to revise. Revision is indicated for:

• Implant malposition causing symptoms
• Progressive subsidence with instability
• Infection
• Intractable pain despite appropriate patient selection

The options.

1. Anterior revision TDR, removing and replacing the implant; technically demanding
2. Anterior implant removal and fusion, converting to an ALIF with cage and screws
3. Circumferential fusion, adding posterior instrumentation; the most common

Why revision is hard. The surgeon meets severe scar from the previous anterior approach, vascular adhesions to the vertebral bodies, and keel and teeth fixation that makes removal difficult. The complication rate is higher than for primary surgery, with 5-10% major complications. Revision to fusion gives 60-80% pain relief; data for revision TDR are limited and its outcomes less predictable.

View larger

View larger

0-6 weeks. The anterior approach spares the posterior musculature and causes less soft tissue trauma than posterior approaches, so patients mobilise on day 1 and typically go home within 1-3 days. With no arthrodesis to protect, rigid bracing is not required and a soft lumbar support is optional. The priority is early walking, avoiding heavy lifting, repetitive bending and end-range extension while the annulus and approach heal.

6-12 weeks. Progressive core and lumbar-stabilisation physiotherapy, with a graded increase in activity. Light or sedentary work often resumes by 2-6 weeks and manual work by around 3 months, guided by symptoms and radiographic stability (checking for subsidence or migration).

Sport. Low-impact activity (walking, cycling, swimming) is encouraged early. Higher-impact loading and contact sport are deferred to around 3-6 months, once motion and stability are confirmed.

How it differs from fusion. With no fusion mass to protect, rehabilitation is generally faster and less restrictive than after lumbar fusion, and motion is actively encouraged. The device relies on bony on-growth to the endplates, however, and early endplate loading before that fixation matures risks subsidence, so heavy loading is still staged.

Global Epidemiology and Utilisation

• Chronic low back pain is the leading cause of years lived with disability worldwide; isolated single-level discogenic pain in a young patient is the narrow niche TDR targets.
• TDR represents a small minority of lumbar surgery globally. Uptake peaked in the late 2000s after the FDA approvals and has since declined in most high-income systems as fusion techniques matured and long-term motion-preservation data disappointed.
• Highest sustained use is in parts of continental Europe (where multiple devices hold CE marking); use is comparatively low in the UK, North America, and most of Asia.

Side-by-Side Guideline Positions

All major bodies converge on one point: patient selection and surgeon experience are decisive, and TDR is never a substitute for fusion when instability, facet disease, stenosis, or deformity is present.

Registry Evidence

• Unlike hip and knee arthroplasty, lumbar TDR has no large dedicated international implant registry; long-term survivorship data therefore rest on the IDE trial cohorts and single-centre series (e.g. the 17-year Charité data above).
• The Swedish Spine Register (Swespine) and similar national spine registries capture some TDR cases and consistently show good short- to mid-term patient-reported outcomes but high heterogeneity, reinforcing the selection-dependent nature of results.
• The absence of arthroplasty-grade registry surveillance is itself a recognised gap and a reason for ongoing caution.

High- vs Limited-Resource Practice Variation

• Well-resourced specialist centres: access to an approach (vascular/access) surgeon, intraoperative fluoroscopy, a range of implant sizes/lordotic angles, and MRI/CT selection workup — the setting in which TDR outcomes are best.
• Limited-resource settings: device cost, lack of an access surgeon, and limited revision capacity make TDR impractical; instrumented fusion (TLIF/PLIF/ALIF) is the pragmatic and far more commonly available motion-segment treatment.
• Globally, fusion remains the default for surgical discogenic pain; TDR is reserved for the rare ideal candidate where the infrastructure to do it safely exists.

Antibiotic Prophylaxis (Implant Surgery — Global Principle)

• Single-dose IV cephalosporin (e.g. cefazolin) at induction; substitute a glycopeptide where MRSA colonisation or beta-lactam allergy applies — consistent with WHO and major national surgical-site-infection prevention guidance.
• Reported surgical-site infection rates are under 2% with appropriate prophylaxis; deep implant infection is managed along arthroplasty/prosthetic-infection lines.

Does TDR actually prevent ASD? This is the central selling point. The 5-year ProDisc-L trial data show that radiographic adjacent-level degeneration is genuinely lower with TDR, yet adjacent-level reoperation rates are statistically similar to fusion. The radiographic benefit has not yet translated into a hard clinical endpoint: plausible, and not proven.

Auto-fusion of the prosthesis. If a motion-preserving device fuses itself, the rationale collapses, and long-term Charité data show 60% spontaneous ankylosis at 17 years, with heterotopic ossification the driver. Whether modern fixed-bearing designs behave better over 15-20 years is unknown.

Is the clinical advantage real? The Cochrane review found ODI and VAS differences favouring TDR that are statistically significant but below the threshold for clinical relevance, on low-quality, industry-sponsored evidence. NICE and several payers therefore do not endorse routine TDR.

The revision burden. Revising a TDR means a repeat anterior approach through scarred retroperitoneum and vascular adhesions, with a higher major-complication rate than primary surgery. Many surgeons argue this future morbidity outweighs an unproven ASD benefit, and favour primary fusion in borderline cases.

Discography as a selection tool. Provocative discography is used to confirm a discogenic source, but it is invasive and operator-dependent, has false positives, and may itself accelerate disc degeneration. Reliance on it for TDR selection is contested, and many centres prefer MRI HIZ or Modic changes plus clinical correlation.

Where TDR sits today. Use has declined globally as minimally invasive and lateral fusion techniques matured and long-term concerns about TDR emerged. It survives as a niche option for the rare, strictly selected young patient with single-level discogenic pain, not as a mainstream alternative to fusion.

FDA Investigational Device Exemption Trials

Long-Term Outcomes (Beyond 10 Years)

The Putzier series highlights that the long-term behaviour of a mobile-bearing prosthesis may converge on that of a fusion (auto-ankylosis), questioning whether the theoretical adjacent-segment benefit is realised in practice. Contemporary fixed-bearing devices (ProDisc-L) maintain measurable motion at 5 years, but comparable independent 15-20 year data remain scarce.

Comparison to Fusion Outcomes

Predictors of Success

Positive Predictors:

• Age under 50 years
• Single-level disease
• Normal BMI (less than 30)
• No facet arthropathy
• No previous surgery
• Non-smoker
• No litigation/workers' compensation
• Realistic expectations

Negative Predictors:

• Multi-level disease
• Facet arthropathy (even mild)
• BMI over 35
• Active smoking
• Previous surgery at level
• Compensation/litigation
• Psychological comorbidities