Pseudarthrosis of the Spine

Smoking. The single most important modifiable risk factor. Smoking roughly doubles the risk of pseudarthrosis (meta-analysis RR 2.2), and with multi-level fusion the risk factors compound, with reported nonunion rates well above single-level baselines. Glassman's nonunion figures, and what they show about when to stop, are set out under Prevention.

• Even second-hand smoke exposure increases risk
• Nicotine patches and replacement therapy also impair fusion, because nicotine is the problem
• Risk reduction requires complete abstinence, not just cutting down
• Even with revision, smoking impairs outcomes

Diabetes mellitus. Diabetes impairs bone healing through several mechanisms. Microangiopathy reduces the blood supply to the fusion bed, advanced glycation end-products (AGEs) impair bone quality, and the increased infection risk compromises healing further.

Obesity. A BMI over 30 is associated with increased pseudarthrosis, through greater mechanical stress on the fusion construct and the metabolic syndrome (insulin resistance, inflammation). Greater soft-tissue dissection increases the infection risk, and operative times are longer.

Nutrition. Vitamin D deficiency impairs calcium absorption and bone formation, protein malnutrition leaves insufficient substrate for collagen synthesis, and calcium deficiency leaves inadequate mineral for bone matrix. Preoperative nutritional optimisation is recommended.

Age and bone density. Older age is associated with higher nonunion: osteoblast activity and bone-healing capacity fall, medical comorbidities accumulate, and osteoporosis is common. Low bone density gives poor-quality bone for graft incorporation, and hardware loosening is more common.

Number of levels. Multi-level fusion increases the pseudarthrosis risk cumulatively, because each additional level increases mechanical stress and healing demand. Above the single-level baseline, reported rates are:

• Two levels: 15-20%
• Three or more levels: 20-30%
• Long constructs (more than 5 levels): 30-40%

Fusion technique. Posterolateral fusion alone has the highest nonunion rate, 10-15%, against 5-8% for PLIF or TLIF and 3-5% for ALIF; the combined anterior-posterior figure is given above. Interbody fusion is more successful for four reasons:

• Load-bearing support of the anterior column
• Compressive forces across the graft, which promote fusion
• A larger surface area for fusion
• A better blood supply from the vertebral body endplates

Instrumentation. Non-instrumented fusion has a pseudarthrosis rate of 20-30%, instrumented fusion 5-15%. Pedicle screws provide rigid fixation that reduces micromotion.

Graft type. Autograft, allograft and bone substitutes differ in what they bring (biologic augmentation is covered under Prevention):

• Iliac crest autograft - the gold standard; osteogenic, osteoconductive and osteoinductive
• Local bone from laminectomy - good if the volume is sufficient
• Allograft alone - higher nonunion (15-20%), with no viable cells

Technique errors. Poor technique increases the pseudarthrosis risk:

• Inadequate decortication - the fusion bed must be bleeding bone
• Insufficient graft volume - a critical mass of bone is needed
• Damage to the graft - excessive cautery kills osteoblasts
• Excessive distraction - increases tension and impairs fusion
• Hardware malposition - eccentric loading and instability

Previous fusion attempt. A revision starts from scarring, poor biology and a fusion bed made avascular by the prior surgery. Each revision decreases the success rate; the figures are under Management.

NSAIDs. The effect is dose-dependent, and blanket avoidance overstates the evidence. The concern originates in animal studies showing impaired osteogenesis. In human data the signal separates by dose: pooling five comparative studies (1,403 patients, none exposed beyond 14 days), high-dose ketorolac raised nonunion risk (RR 2.87, 95% CI 1.53-5.38), while normal-dose NSAIDs (ketorolac, diclofenac, celecoxib or rofecoxib) did not (RR 1.39, 95% CI 0.74-2.61) (Li, PMID 20489674).

What to do with that. The defensible position is to avoid high-dose ketorolac perioperatively rather than deny all patients effective analgesia; short-course normal-dose NSAIDs appear safe on current evidence. The common practice of avoiding NSAIDs for 3-6 months after fusion is a convention. No study has tested exposure beyond 14 days, so the long-term question is genuinely unanswered rather than answered in the negative.

COX-2 inhibitors. They are often assumed to spare bone healing, but the meta-analysis grouped them with the other normal-dose agents and did not demonstrate a class advantage.

Corticosteroids. Chronic steroid use impairs bone formation and delays fracture healing and fusion. If possible, minimise the dose perioperatively.

Chemotherapy and radiation. Cytotoxic agents impair cell division and healing, and radiation damages local vascularity. A history of radiation to the fusion site increases nonunion risk.

Systemic disease. Three conditions are relevant:

• Rheumatoid arthritis and autoimmune disease - the disease itself and the immunosuppressive medications impair healing, and the infection risk is higher
• Chronic kidney disease - renal osteodystrophy affects bone quality, with parathyroid dysfunction
• Hypothyroidism - impaired bone turnover; it should be corrected before surgery

Infection. Postoperative infection dramatically increases pseudarthrosis. Infection prevents bone formation, biofilm on hardware prevents fusion, and treatment may require hardware removal with loss of stability. It must be eradicated before a revision fusion is attempted.

The best test, not the reference standard. CT with thin cuts (1-2 mm slices) and multiplanar reconstructions is the best imaging test for assessing spinal fusion, and it is what you should order. Call it the gold standard only loosely: the actual reference standard is surgical exploration.

How it performs against exploration. The result was sobering. Across five spine surgeons reading 69 levels with metallic interbody cages, interobserver agreement on "fused or not" was kappa 0.25, only fair, with sensitivity 70-97% but specificity just 28-85% (Carreon, PMID 18280214). The error runs in one direction: raters consistently over-read fusion. The sentinel sign was little better (kappa 0.34, overall accuracy 61%; posterior sentinel sign 74%).

What that changes. A CT reported as "fused" in a patient with persistent, positional, load-related pain and no other explanation does not exclude pseudarthrosis, because the test's weakness is precisely its tendency to call a nonunion solid. Weigh the clinical picture and hardware behaviour (screw haloing, rod fracture) alongside the scan, and be more sceptical of a fused report around metallic cages than around a posterolateral graft mass.

Corroborating an equivocal scan. Correlate it with dynamic flexion-extension films, where motion means nonunion, and with hardware integrity: a broken rod or screw means nonunion regardless of the CT read.

Technique. Thin-cut axial images through the fusion levels, sagittal and coronal reconstructions, and bone windows for optimal visualisation.

Signs of solid fusion. Continuous bridging trabecular bone through the graft on all planes, graft incorporation (remodelling and trabecular integration with host bone), and no radiolucent line at the graft-host interface.

Signs of pseudarthrosis. On CT, look for:

• Lucency at the graft-host junction - a radiolucent line indicates fibrous tissue
• Discontinuous bone - a gap in the expected bone bridge
• Sclerotic margins - dense bone at the edges without bridging suggests nonunion
• Resorption - graft material absorbed without new bone formation
• Hardware loosening - lucency around the screws (halo sign)
• Hardware breakage - rod or screw fracture indicates motion and nonunion

Role. Plain films are less sensitive than CT but useful for initial screening, for serial follow-up once CT has confirmed fusion, and for identifying hardware complications. Overlapping structures obscure the fusion mass, and anteriorly located bone (interbody fusion) cannot be assessed.

Views. The standard views:

• AP and lateral lumbar spine
• Oblique views if posterolateral fusion is being assessed
• Dynamic flexion-extension views to assess motion

Signs suggesting fusion. Continuous bony trabeculae across the fusion site, incorporation of the graft with host bone, and hardware position that stays stable over time.

Signs suggesting pseudarthrosis. On plain films:

• Hardware breakage - rod or screw fracture, highly specific for nonunion
• Hardware loosening - progressive lucency around screws over serial films; a halo of lucency around a pedicle screw of more than 2mm suggests loosening
• Motion on dynamic films - greater than 3-4 degrees angulation or 3mm translation
• Persistent radiolucent line at the graft-host interface

SPECT/CT. Single-photon emission computed tomography combined with CT shows metabolic activity, identifying a biologically active pseudarthrosis. It distinguishes active from quiescent nonunion: bright uptake suggests a symptomatic nonunion, and active uptake may predict who benefits from revision. Against that, it is more expensive than CT alone, the radiation exposure is higher, it is not always available, and uptake can be positive for other reasons (infection, hardware reaction).

When SPECT/CT helps. Three situations:

• Equivocal CT findings
• Clinical-radiographic discordance - pain, but CT shows fusion
• Planning revision surgery, to identify the active site

MRI. Not the gold standard for fusion assessment: metal artefact from the hardware degrades the images, trabecular bone detail is difficult to see, and it is less accurate than CT for bridging bone. It has a role for what lies around the fusion:

• Infection - abscess, discitis, osteomyelitis
• Adjacent segment disease - disc degeneration, stenosis
• Nerve compression - hardware-related radiculopathy
• Soft tissue - haematoma, seroma

Laboratory tests. If infection is suspected, check the ESR and CRP (elevated in infection), the white blood cell count (may be elevated), and blood cultures if there is systemic sepsis. Bone turnover markers, P1NP for formation and CTX for resorption, may predict fusion but belong to the research setting and are not used clinically.

What to do. No surgery is indicated. Observe clinically, educate the patient about the diagnosis and reassure them that no treatment is needed. Serial imaging is not necessary while the patient is asymptomatic; they return if symptoms develop.

Why. Surgery should be for symptoms, not for a radiographic finding. Revision has risks (infection, neurological injury, dural tear), and in a pain-free patient the fibrous union may provide adequate stability.

Who. Conservative management is the initial approach for symptomatic pseudarthrosis. A trial of non-operative treatment suits:

• Mild to moderate symptoms
• A patient who wishes to avoid surgery
• Medical comorbidities that make surgery high-risk
• A recent diagnosis (within 12-18 months), which may stabilise

The options. Treatment draws on six measures:

• Activity modification - avoid aggravating activities, weight management, ergonomic adjustments
• Bracing (controversial) - a TLSO may reduce motion and pain; evidence for efficacy is limited, though some patients report symptomatic improvement
• Physiotherapy - core strengthening to support the spine; not curative but may reduce pain; avoid aggressive manipulation
• Medication - paracetamol for pain; avoid NSAIDs, which may impair fusion; neuropathic pain medication if there is a nerve component; minimise opioids (chronic pain, addiction risk)
• Interventional pain management - epidural steroid injections (temporary relief, not curative), facet injections for adjacent segment pain, radiofrequency ablation (palliative)
• Risk factor modification - smoking cessation, which may allow late fusion even in an established pseudarthrosis; diabetes control, nutrition and bone health (vitamin D, calcium)

What to expect. Most symptomatic pseudarthrosis does not resolve without surgery. Conservative measures provide temporary relief, and definitive treatment requires revision fusion if symptoms persist.

Indications. The case for revision rests on:

• Symptomatic pseudarthrosis causing significant pain and disability, with axial pain at the fusion site beyond 1 year
• Failed conservative management (a trial of at least 3-6 months)
• CT-confirmed nonunion
• Compromised stability - hardware breakage or loosening
• A patient medically fit and willing
• Smoking cessation achieved - mandatory, for a minimum of 6 weeks

Contraindications. Smoking appears on both lists: cessation is a prerequisite above, and active smoking is a relative contraindication below.

• Active smoking - a relative contraindication, with a very high failure risk
• Active infection - it must be eradicated first
• Medical comorbidities precluding surgery
• Asymptomatic pseudarthrosis

The operation. Six steps, which the GRIFT mnemonic below compresses:

1. Debride - expose the pseudarthrosis, remove all fibrous tissue and scar, excise the sclerotic bone margins and create fresh bleeding bone surfaces
2. Decorticate - the transverse processes for PLF and the vertebral endplates for interbody fusion, to punctate bleeding and the maximum surface area for fusion
3. Graft - fresh iliac crest autograft is the gold standard, taken from the posterior crest if available (less morbidity than anterior); local bone from the revision laminectomy if the volume is sufficient; morselised graft packed into the fusion bed, and structural graft for interbody (a cage with autograft)
4. Instrument - revise existing hardware if loose or malpositioned, extend the construct proximally and/or distally for stability, consider larger-diameter rods for greater stiffness, and ensure solid purchase in bone (rescue screws if needed)
5. Add interbody support - posterior (PLIF), transforaminal (TLIF) or anterior (ALIF) lumbar interbody fusion; the cage provides compressive loading and a larger fusion surface, and it is worth considering if the initial fusion was PLF alone
6. Augment - rhBMP-2 improves fusion in the revision setting and is expensive but effective in high-risk revisions; BMAC is an adjunct, and demineralised bone matrix (DBM) an extender

Approach. Three options:

• Posterior - re-explores the original posterior fusion site, gives access to the posterolateral fusion bed and allows PLIF or TLIF; scar tissue and epidural fibrosis make the dissection challenging
• Anterior (adding ALIF) - fresh tissue planes, excellent visualisation of the disc space and room for a large structural graft, at the cost of a separate incision and longer total surgery; consider it if the posterior anatomy is too scarred
• Combined anterior-posterior (360-degree) - the most stable construct, with the highest fusion rates in the revision setting, at the morbidity of two approaches; consider it in high-risk or multiply failed fusions

Outcomes. Primary fusion succeeds in 90-95%. A first revision fuses in 70-80% if risk factors are controlled, leaving 20-30% with persistent pseudarthrosis, some of whom may require multiple revision attempts; a second revision fuses in 50-60%. BMP may improve rates by 10-15%, and smoking cessation is essential for acceptable success.