Pseudarthrosis (Failed Spinal Fusion)

A spinal fusion is, in essence, a controlled attempt to make the body weld two or more vertebrae into one. Bone graft is laid across the levels to be joined, and screws, rods, plates or cages are placed to hold everything still while that graft matures. Over the following months, in a successful fusion, the graft is gradually replaced by living, load-bearing bone that bridges the gap, and the operated segment becomes a single solid unit. Pseudarthrosis is what happens when this welding process does not complete: the bridge of bone fails to form, and a false joint — a zone of fibrous tissue or persistent micromotion — remains in its place.

The defining feature is that it is a failure of bone biology and bone healing, not primarily a failure of metal. This matters because the conditions it is most easily confused with are different in kind. Implant or hardware failure means breakage of a rod, loosening of a screw, or subsidence of a cage; metalwork is only ever intended to be a temporary internal splint, and it usually fails because the bone never fused — so broken metal is often the visible signature of an underlying non-union rather than an independent problem. Adjacent segment disease is wear at a mobile level next to a fusion that has, in fact, healed perfectly well — a problem of altered mechanics at a neighbouring level, not a failure of the original fusion to unite. Infection is a microbial problem that can itself prevent bone from healing and loosen metalwork, but it is distinct from a simple mechanical or biological non-union. Holding these four ideas apart — non-union, implant failure, adjacent segment disease and infection — is the most useful framework for making sense of a fusion that is not behaving as hoped.

For a fusion to succeed, bone must heal across the gap, and bone healing has well-recognised requirements. It needs living cells capable of making bone, a scaffold for them to grow along, the biological signals that prompt bone formation, a good blood supply, and a mechanically stable environment in which to work. The causes of pseudarthrosis divide broadly into problems of biology and problems of mechanics.

On the biological side, the most consistently reported and most important modifiable factor is smoking. Nicotine narrows small blood vessels, reduces the blood flow that delivers oxygen and nutrients to the healing graft, and directly hampers the cells that build bone. Anti-inflammatory painkillers of the NSAID type, taken in the early period after surgery, have been associated with higher rates of non-union in several studies. Diabetes (particularly when poorly controlled), osteoporosis, poor nutrition, obesity, long-term steroid use, inflammatory conditions, kidney disease, previous radiotherapy and older age all contribute.

On the mechanical side, the recurring theme is that bone will not bridge a gap that moves too much. Multilevel fusions carry a higher risk than single-level ones, simply because there are more surfaces that must all heal and a longer lever arm acting on the construct. Long constructs, fusions extending down to the sacrum and pelvis, and large deformity corrections all place high mechanical demands on the healing bone. Inadequate fixation prevents union directly. The choice and quantity of bone graft also matters: graft taken from the patient's own body contains living bone-forming cells, whereas donor graft does not, and in some settings donor graft is associated with a higher non-union rate. Revision surgery in a previously operated, scarred bed is biologically less favourable than surgery in untouched tissue.

Pseudarthrosis is rarely a matter of biology or mechanics alone. A fusion fails to unite because the body's capacity to heal has been impaired, because the mechanical environment was too unstable, or, most often, because of some combination of the two.

Reported frequency varies enormously according to the type of operation, the number of levels and the population studied. In adolescent idiopathic scoliosis surgery, non-union is uncommon, on the order of 1–2%. In neuromuscular scoliosis it is somewhat higher, around 2%, and in adult spinal deformity rates of around 6% and above are reported. In the lower back, posterolateral fusion has been reported across a wide band from roughly 3% to 25%. In the neck, fusion is achieved in the great majority of single-level operations, with success rates around 96%, falling somewhat for two-level and three-or-more-level procedures. The consistent message is that the more levels are fused and the greater the mechanical and biological challenge, the higher the chance that union fails.

A radiographic non-union and a symptomatic non-union are not the same thing. A substantial proportion of pseudarthroses — by some accounts up to around half — are completely without symptoms, discovered only because a scan happens to show that the bone has not fully bridged, or because a piece of metalwork is seen to have loosened or broken, with no accompanying pain or disability. Some non-unions identified at one year after surgery go on to consolidate by themselves over the following year. In the early period after surgery, a thin line of lucency around a graft or cage is common and, in the majority of cases, eventually fills in with solid bone.

The practical consequence is that imaging findings are interpreted in the light of how the operated person actually is. A non-union that is stable and causing no trouble is regarded very differently from one associated with persistent or returning pain, progressive deformity or broken metalwork. Management hinges far more on symptoms than on the scan appearance alone.

When pseudarthrosis does cause symptoms, the commonest is pain, typically felt in the region of the operated spine, often described as a deep, mechanical, activity-related ache that may have eased for a time after surgery and then returned, or that never fully settled. Because the operated level retains an element of abnormal movement, the pain is frequently worse with loading, with twisting and bending, and toward the end of the day, and may be relieved by rest. In some cases the first sign is not pain at all but a change seen on imaging, such as a broken rod or a loosening screw found at routine follow-up. Where a non-union is associated with progressive collapse or angulation, there may be a gradual change in posture or alignment. Less commonly, shifting bone or hardware encroaches on nearby nerves, producing renewed leg or arm symptoms.

The metalwork placed at a fusion — screws, rods, plates and cages — is best thought of as an internal splint. Its job is to hold the spine still and share the load while the bone heals. In a fusion that unites successfully, the load is gradually transferred from the metal to the newly formed bone, and once solid union is achieved the metalwork is largely unloaded and protected.

When a fusion fails to unite, this protective transfer never happens. The metalwork is left to bear repeated, cyclical loading at a level that keeps moving, day after day, step after step. No metal can withstand this indefinitely, and over time it fatigues and fails. The result is a fractured rod, a broken or loosening screw, a halo of lucency around a screw, or a cage that has subsided or migrated. A broken rod in particular is regarded as a reliable pointer to an underlying non-union. In this common scenario, the broken metal is the consequence and the visible signature of the non-union, not the disease itself.

Plain radiographs are insensitive for detecting whether a fusion has truly united. Metalwork casts shadows that obscure the very graft one is trying to see, and a film that looks acceptable does not confirm that solid bone has bridged the gap. Plain films can suggest a problem — by showing broken hardware or a lack of bridging bone — but a normal-looking film cannot be relied upon to prove union.

Dynamic, or flexion-extension, X-rays are taken to look for movement at the operated level. An absence of motion is reassuring and suggests union, but the presence of some motion does not reliably prove a non-union, and the metalwork tends to mask whatever movement is present.

The reference standard among the scans is the fine-cut CT, acquired in thin slices and reconstructed in several planes. It can show whether trabecular bone has genuinely bridged across the graft or interbody space, and it reveals lucency or haloing around screws, broken hardware, graft resorption and subsidence. Grading systems describe what the CT shows, ranging from clear non-union, through an indeterminate 'locked' appearance, to solid fusion. The absolute gold standard remains direct surgical exploration but this is not used as a routine test. Where infection is a possibility, blood tests and other investigations address that separate question.

The evidence base for managing pseudarthrosis is built largely on retrospective series, registry data, biomechanical experiments and reliability studies, rather than on large randomised trials. The following describes what the literature reports, not guidance for any individual.

A stable, asymptomatic non-union is, in the great majority of cases, observed rather than operated upon. It is monitored with imaging over time, some go on to consolidate, and the modifiable factors that influence bone healing — stopping smoking, control of diabetes, attention to bone health and nutrition — are addressed.

Surgery, in the form of revision fusion, is reserved for the symptomatic non-union — one causing persistent or recurrent pain, progressive deformity, or hardware failure that is itself causing problems. The governing principle is to improve both the biology and the mechanics. On the mechanical side, this typically means removing or revising the failed metalwork and applying more rigid, often longer or circumferential, fixation, which is reported to achieve higher union rates than a single-sided approach. On the biological side, it means cleaning out the false joint, freshening the bone surfaces and adding new bone graft, often taken from the patient's own body, sometimes supplemented by biological bone-stimulating agents such as bone morphogenetic protein in selected cases. Throughout, the importance of optimising the host, and of excluding low-grade infection, is consistently emphasised.

For the large group whose non-union is stable and causes no symptoms, the outlook is generally good, and the situation is often one of monitoring rather than treatment. For those with symptomatic non-union undergoing revision surgery, the published outcomes are broadly encouraging, particularly where the revision improves both the biology and the mechanics. Reported union rates are high after well-planned revision, although a small proportion of revisions themselves fail to unite. The literature consistently notes that addressing the underlying contributors — above all stopping smoking — gives the best chance of a sound result.

This page is general medical education and is not a substitute for individual advice. New or worsening neurological symptoms after spinal surgery require urgent assessment.

• —Wheeless CR. Pseudarthrosis: diagnosis and management. Wheeless' Textbook of Orthopaedics, Section 18, Chapter 4.
• —How JNL, Street JT, Dvorak MF, et al. Pseudarthrosis in adult and pediatric spinal deformity surgery: a systematic review and report of risk factors. Neurosurg Rev. 2019;42(2):319–336.
• —Pseudarthrosis risk factors in lumbar fusion: a systematic review and meta-analysis. PMC11149252.
• —Hardesty CK, Poe-Kochert C, Son-Hing JP, et al. Rate and risk factors for pediatric cervical spine fusion pseudarthrosis. Spine Deform. 2023;11(3):665–671.
• —Risk factors for pseudarthrosis in MIS-TLIF. PMC6147882.
• —Kanna RM, Shetty AP, Rajasekaran S, et al. Diagnostic accuracy of imaging modalities to detect pseudarthrosis after spinal fusion: systematic review and meta-analysis. Skeletal Radiol. 2019;48(10):1559–1571.
• —Kaiser MG, Mummaneni PV, Matz PG, et al. Guideline update for fusion procedures for degenerative lumbar spine: radiographic assessment of fusion status. J Neurosurg Spine. 2014;21(1):23–30.
• —Radiological assessment of lumbar fusion status: which imaging modality is best? PMC12014583.
• —Hilibrand AS, Carlson GD, Palumbo MA, et al. The effect of smoking on spinal fusion. PMC5779238.
• —Adverse impact of smoking on spine fusion and patient-reported outcomes: a systematic review and meta-analysis. PMC12510999.
• —Do NSAIDs affect bone healing rate, delay union, or cause non-union: updated systematic review and meta-analysis. Front Endocrinol. 2024;15:1428240.
• —Daniels AH, Ritterman SA, Rubin LE. Failure in lumbar spinal fusion and current management modalities. PMC8110346.
• —Murphey MD, et al. Spine fixation hardware: an update. AJR. DOI 10.2214/AJR.20.22810.
• —Prevalence and modes of posterior hardware failure. Int J Spine Surg. 2022;16(3):481.
• —Smith JS, Shaffrey CI, Ames CP, et al. Instrumentation failure following pedicle subtraction osteotomy. Eur Spine J. 2017;26(8):2128–2137.
• —Prognosis of symptomatic pseudarthrosis observed at 1 year after lateral lumbar interbody fusion. Spine. PMID 33534522.
• —Suchomel P, Barsa P, Buchvald P, et al. Pseudarthrosis of the cervical spine: risk factors, diagnosis and management. PMC4995265.
• —Allograft versus autograft in ACDF: a propensity-matched analysis. PMC8956488.
• —Pseudarthrosis rate following ACDF using an allograft cellular bone matrix. Neurosurg Focus. 2021;50(6):E6.
• —Steinhaus ME, York PJ, Bronheim RS, et al. Outcomes of revision surgery for pseudarthrosis after anterior cervical fusion: case series and systematic review. Global Spine J. 2020;10(8):1056–1066.
• —Carreon LY, Glassman SD, Campbell MJ. Outcomes analysis of treatment of cervical pseudarthrosis with posterior fusion. Spine. 2006;31(5):542–548.
• —Meta-analysis of clinical and fusion results following treatment of symptomatic cervical pseudarthrosis. PMC4369200.
• —Fogel GR, Toohey JS, Neidre A, Brantigan JW. Fusion assessment of PLIF using radiolucent cages: radiographs and helical CT compared with surgical exploration. Spine J. 2008;8(4):570–577.
• —CT evaluation of lumbar interbody fusion: principle-based interpretation framework. Diagnostics. 2025.
• —Glassman SD, Howard JM, Sweet A, Carreon LY. Complications and concerns with osteobiologics for spine fusion: rhBMP-2 in posterolateral lumbar fusion. PMC3621610.
• —SPECT-CT assessment of pseudarthrosis after spinal fusion: diagnostic pitfall due to a broken screw. PMC3789298.

This information is for general patient education and is not a substitute for an individual consultation. Always seek personalised advice from a qualified spinal specialist.