Adjacent Segment Disease

Adjacent segment disease (ASD) is the development of degenerative change at the spinal levels immediately adjacent to a previous fusion.

Define the terms. The distinction that runs through the whole topic is between what the imaging shows and what the patient feels:

• Radiographic ASD, used interchangeably with adjacent segment degeneration (ASDeg): imaging evidence of degeneration at an adjacent level (disc space narrowing, osteophytes, facet hypertrophy)
• Symptomatic ASD (ASDis): pain or neurological symptoms arising from adjacent level degeneration
• Adjacent segment pathology: symptomatic disease requiring treatment

View larger

How common radiographic change is. Pooled radiographic degeneration is 26.6% in the lumbar spine and 32.8% in the cervical spine (Hashimoto 2018 systematic review), up to 30% at 10 years. Only about one-quarter to one-third of that radiographic degeneration becomes symptomatic; radiographic change may not progress, and not every change on imaging requires intervention.

How common symptomatic disease is. Symptomatic ASD develops at roughly 2.5-3% per year. In the cervical spine the rate is 2.9% per year (Hilibrand 1999). In the lumbar spine, disease-free survival is 83.5% at 5 years and 63.9% at 10 years, with symptomatic degeneration warranting surgery predicted at 16.5% by 5 years and up to about 36% by 10 years (Ghiselli 2004).

Convert the annual rate before quoting it. A constant 2.9% per year is not a small number over the life of a fusion. Kaplan-Meier survivorship in the same cervical cohort (374 patients, 409 arthrodeses, followed up to 21 years) predicted that 25.6%, roughly one in four, would develop new adjacent-level disease within 10 years (95% CI 20-32%), and two-thirds of those affected, about 17% of all patients, ultimately required further surgery. An annual rate sounds reassuring and a decade of it does not; consent should use the compounded figure, because that is the one that matters to a patient.

Where it happens. L4-5 is the most common lumbar level: it is the most mobile lumbar segment, and it is the level left next to a fusion that ends at L5 rather than S1. In the cervical spine C5-6 and C6-7 are the most common levels.

Pre-existing degeneration at the adjacent level is the strongest non-modifiable predictor of ASD. PI-LL mismatch with loss of lordosis, and facet joint violation, are the most important modifiable, technique-dependent factors, and they are the ones the surgeon controls.

Sagittal alignment. Loss of lumbar lordosis and PI-LL mismatch are the strongest modifiable risk (Lau 2021). A PI-LL mismatch of 10 degrees or more and an SVA of 50 mm or more are associated with ASD even after a single-level PLIF, as is high pelvic tilt, and a fusion in kyphosis accelerates ASD. Restoring sagittal balance at the index operation is protective: lumbar lordosis should match pelvic incidence, with a PI-LL mismatch under 10 degrees.

Facet joint violation. Of all the risk factors, this is the one most directly in the surgeon's hands, and examiners reward candidates who can explain it. When the cranial-most pedicle screw is placed, a starting point that is too superior or lateral, or a trajectory that is too cephalad, breaches the facet joint of the segment above the construct, the UIV+1 mobile level, destabilising and degenerating the very level the construct was meant to protect. The risk is higher with freehand or open placement and a cephalad-angled cranial screw, and it is reported in a meaningful fraction of cranial screws.

Avoid it with a more caudal, medial starting point and a non-cephalad trajectory for the UIV screw, keep instruments and the screw head off the unfused facet capsule, and confirm on fluoroscopy or navigation that the cranial screw has not entered the joint above.

View larger

Decompression at the next level. An adjacent-level laminectomy, or a wide decompression with extensive facet removal, destabilises the segment next to the construct.

Prevention at the index operation. The levers the surgeon holds follow directly from the modifiable factors:

• Restore sagittal alignment: match lumbar lordosis to pelvic incidence, with adequate lordosis
• Preserve the facet joints: avoid unnecessary facet removal and keep the cranial screw out of the UIV+1 facet
• Preserve the posterior musculoligamentous complex, and do not extend a decompression into the adjacent facet or lamina
• Choose a sensible construct endpoint: end at a stable level rather than a mobile segment
• Fuse no more than necessary
• Treat the comorbidities: BMI, smoking, bone health

Prophylactic hardware. Surgeons have also tried to prevent ASD at the index operation with motion-sparing hardware, the so-called "topping-off" construct: a semi-rigid or dynamic rod, an interspinous process device or a pedicle-based dynamic stabiliser at the level immediately above the fusion to soften the stress transition, or a hybrid of fusion with an adjacent disc replacement. The evidence has been disappointing. Randomised and comparative studies of dynamic stabilisation, interspinous spacers and topping-off constructs have not reliably shown a reduction in clinically important ASD, and the devices bring their own problems: screw loosening, device failure, interspinous fracture and cost.

That dovetails with the natural-history argument: if much of ASD is the patient's own progressive spondylosis, a gadget at one extra level cannot prevent it. Cervical disc arthroplasty is the one motion-preserving strategy with the best, though still debated, ASD-reduction data, and that is a different setting. Topping-off and hybrid constructs remain investigational and should not be presented as standard care; the prevention that matters is technique, not implants.

Load transfer. Fusion eliminates motion at the treated segments, so load and motion transfer to the adjacent levels. Stress on the adjacent disc and facets rises, and degeneration at those levels accelerates. That is the load transfer hypothesis, and it is the biomechanical basis of ASD.

View larger

View larger

What the laboratory shows. Cadaveric studies show increased intradiscal pressure at the adjacent levels after fusion, increased facet joint loading above the fusion, and greater motion at the adjacent segments to compensate for the fused levels. Finite-element models show the same, and show that longer constructs raise stress at the transition zones.

What changes the load. The stress an adjacent level sees depends on the construct and on the level itself:

• Fusion length: a longer fusion places greater stress on the adjacent levels
• Fusion position: sagittal balance affects how the load is distributed
• Adjacent segment quality: a segment with pre-existing degeneration is the vulnerable one
• Construct rigidity: stiff pedicle screw constructs transfer more load than cables or wires

View larger

View larger

The natural history argument. Patients who need a fusion already have degenerative disease, and their adjacent segments may have subclinical degeneration at the time of surgery. Some ASD is therefore natural progression of that disease, and age-matched controls develop degeneration too. The high rate of imaging change set against the much lower rate of clinical disease supports natural history playing a major role.

By location. ASD is described by where it sits relative to the construct.

By severity. Radiographic change at the adjacent segment is graded by what the imaging shows.

History. The pattern to elicit is a symptom-free interval after the index surgery, months to years, followed by new pain in a location or pattern different from the original complaint: axial back pain, radicular symptoms, or both. It may mimic the original presentation. Ask about functional impact (walking distance, activities of daily living) and the details of the previous surgery: the levels fused, the approach and any complications.

Presentation patterns. The pathology at the adjacent level decides the symptoms:

• Stenosis: neurogenic claudication, radiculopathy
• Disc herniation: radicular pain, dermatomal sensory changes
• Instability: mechanical back pain, worse with motion
• Spondylolisthesis: back pain, radicular symptoms, gait changes

Examination. The findings point towards the pathology and towards the operation it will need.

Differential diagnosis. New or recurrent symptoms after a fusion are not automatically ASD. Exclude the mimics systematically before attributing symptoms to the adjacent level; several require entirely different management.

View larger

View larger

Radiographs. Standing views are the first-line study and are essential, because they reveal functional alignment and dynamic change:

• AP and lateral lumbar spine
• Flexion-extension views for dynamic instability
• Long-cassette scoliosis views if there is deformity

Look for disc space narrowing at the adjacent levels, osteophyte formation, endplate sclerosis, a new spondylolisthesis, loss of disc height compared with the preoperative films, and change in sagittal alignment. Dynamic instability is translation greater than 4 mm or angulation greater than 10 degrees on flexion-extension.

View larger

View larger

MRI. The study for soft tissue and neural assessment, indicated for radicular symptoms, neurogenic claudication or progressive symptoms, and essential for assessing neural compression and planning surgery. It shows disc degeneration or herniation, central or foraminal stenosis, ligamentum flavum hypertrophy, facet joint effusion or cyst, and nerve root compression. Metal artefact from the hardware is the limitation, and titanium-compatible sequences may be needed.

View larger

View larger

View larger

CT. The study for bone detail and fusion status: whether the fusion is solid or a pseudarthrosis, bone quality, the detail of facet arthropathy, and preoperative planning. CT myelography is the alternative when MRI is contraindicated or inadequate, and it allows dynamic assessment.

Diagnostic injections. Selective nerve root blocks, facet joint injections, medial branch blocks and adjacent-level discography (controversial) localise the pain source. They confirm the pain generator, help distinguish ASD from pseudarthrosis and guide surgical planning, and they earn their place when the clinical and imaging findings are discordant.

Conservative first. Symptomatic ASD is treated non-operatively for 3-6 months before surgery is considered, and approximately 50% of patients may improve with conservative treatment. The programme:

• Physiotherapy: core strengthening, flexibility exercises, posture optimisation, activity modification
• Medication: NSAIDs, neuropathic agents (gabapentin, pregabalin), short-term oral steroids for an acute exacerbation, muscle relaxants
• Injections: epidural steroid injections, facet joint injections, medial branch blocks, selective nerve root blocks
• Bracing: a limited role in the lumbar spine; it may provide short-term relief

When conservative treatment fails. The choice of operation depends on the pathology, the stability of the segment, the alignment and the patient.

Decompression alone. For isolated stenosis without instability, with preserved disc height and no significant spondylolisthesis. It preserves motion; the risk is that it destabilises the segment and a fusion is needed later.

Extension of fusion. For instability, spondylolisthesis, or sagittal imbalance requiring correction. It is the most common surgical approach, and most symptomatic ASD that comes to surgery is treated this way. The disadvantage is further load transfer to the next level.

Motion preservation and hybrid constructs. Total disc replacement adjacent to the fusion, or dynamic stabilisation, may reduce load transfer, but both are investigational with limited long-term data. A hybrid construct, fusion at the index level with motion preservation at the adjacent level, has the theoretical advantage of reduced ASD and is technically demanding.

Fusion extension. Planning covers how many levels to add, whether sagittal balance needs correcting, where the upper instrumented vertebra (UIV) will be, and the quality of the bone. Extending to the sacrum and pelvis may be needed if fusing below L4.

View larger

View larger

Adjacent-level decompression. For the patient who meets the criteria for decompression alone and has focal neural compression. A laminotomy or laminectomy at the adjacent level, preserving as much facet as possible (greater than 50%), with a foraminotomy for radicular symptoms, taking care not to destabilise the segment. It carries less morbidity than a fusion and preserves motion at that level; against that is a 10-20% risk of subsequent instability, which may need a fusion later. Patient selection is critical.

View larger

Motion preservation. The options are total disc replacement at the adjacent level, dynamic stabilisation devices and interspinous process devices. Total disc replacement replaces the disc at the level adjacent to the fusion and theoretically reduces load transfer; it requires intact posterior elements and has limited application for extension to a fusion. Dynamic stabilisation uses pedicle screw-based systems with a flexible connection that allow some motion while providing stability, with mixed outcomes in the literature.

Long-term data are limited, patient selection is critical, and motion preservation may reduce but not eliminate ASD. It remains investigational for treating or preventing ASD.

Revision surgery. Revision surgery for ASD has higher complication rates than primary surgery. The complications, by timing:

• Intraoperative: dural tear (more common in revision), nerve root injury, vascular injury with an anterior approach, hardware malposition
• Early postoperative: wound infection (higher in revision surgery), hardware failure, adjacent-level injury from retraction, medical complications
• Late: pseudarthrosis at the new fusion, hardware loosening, recurrent ASD at the next level, persistent symptoms

The cascade. Each extension creates a new transition zone and puts the next adjacent level at risk; after a fusion extension, symptomatic ASD continues at 2.5% per year. The end result may be multiple revisions and a long fusion. The possibility of a cascade is what makes prevention at the index operation, and attention to the modifiable risk factors, matter.

After fusion extension. Recovery follows the phases below, with bone healing assessed at 3-6 months on standing radiographs.

Protecting the next level. Prevention of ASD at the newly adjacent levels is an important component of care, and much of it is the patient's: smoking cessation, weight management, bone health optimisation, activity modification, core strengthening and regular follow-up surveillance.

Surgical treatment of ASD is generally effective but carries the risk of further ASD at the next levels.

Global epidemiology

• Symptomatic ASD requiring surgery is a leading cause of lumbar fusion revision worldwide; pooled radiographic degeneration is ~26.6% (lumbar) and ~32.8% (cervical) (Hashimoto 2018).
• A systematic review reported a higher prevalence of lumbar ASDeg in Western than in Eastern populations (Cannizzaro 2022), likely reflecting differences in surgical indications, BMI distribution and follow-up imaging practice rather than a single biological cause.
• As global life expectancy and fusion volumes rise, the absolute burden of ASD is increasing, making prevention at the index operation a worldwide priority.

Side-by-side guidance (recommendations converge more than they differ)

All major societies agree on three points: (1) confirm the diagnosis and exclude pseudarthrosis/infection before surgery, (2) treat conservatively first, and (3) restore sagittal alignment and avoid iatrogenic destabilization at the index operation. Genuine differences are minor and largely about thresholds rather than principles.

Registry evidence

• Large national spine registries (e.g. the Swedish Spine Register (Swespine), the British Spine Registry, and Norwegian (NORspine) datasets) capture reoperation as a proxy for symptomatic ASD and consistently identify revision for adjacent-level pathology as a major late reoperation cause.
• Registry reoperation data underpin patient counselling on long-term revision risk and are increasingly used to benchmark alignment-restoring techniques.

High- vs limited-resource practice variation

• Well-resourced settings: routine long-cassette standing radiographs for spinopelvic parameters, MRI with metal-artifact-reduction sequences, navigation/robotics to reduce facet violation, and access to revision deformity surgery and arthroplasty.
• Limited-resource settings: sagittal alignment may be judged on standard radiographs without full spinopelvic measurement; metal-artifact MRI and revision deformity capacity may be scarce, so prevention at the index operation and prolonged non-operative management carry even greater weight.
• Bone-health optimization (vitamin D, osteoporosis treatment) and smoking cessation are universally relevant, low-cost, modifiable measures that improve both prevention and revision outcomes regardless of setting.

ASD is one of the most debated topics in spine surgery, and examiners use it to test whether a candidate can reason with uncertain evidence rather than recite dogma.

Does motion preservation help? Total disc replacement reduces overall reoperation in some trials but has not consistently reduced adjacent-level surgery versus fusion at mid-term (Radcliff lumbar RCT; Harrod/Hilibrand cervical review). The counter-signal is the 7-year two-level Mobi-C IDE, with adjacent-level secondary surgery of 4.4% against 11.3% with ACDF, an industry-sponsored trial in highly selected patients. Whether arthroplasty truly prevents ASD, rather than avoiding fusion-specific reoperations, remains unproven.

Prophylactic extension. Should a mildly degenerate adjacent level be included in the index fusion? Hilibrand argued that symptomatic degenerate segments should be fused, but prophylactically extending onto an asymptomatic level lengthens the construct, creates a new transition zone and is not evidence-supported. Most surgeons fuse only symptomatic levels.