Adult Spinal Deformity

Adult spinal deformity (ASD) encompasses a spectrum of conditions characterised by abnormal spinal curvature in adults, with increasing recognition of the importance of sagittal plane alignment. Unlike adolescent idiopathic scoliosis, ASD is predominantly a sagittal plane problem.

The types. Deformity in the adult arises in one of the following ways:

• De novo degenerative scoliosis - develops in adulthood from asymmetric disc degeneration and facet arthropathy
• Progressive idiopathic scoliosis - adolescent scoliosis that progresses in adulthood
• Iatrogenic deformity - after prior spinal surgery (flatback syndrome, adjacent segment disease)
• Secondary deformity - metabolic bone disease, trauma or infection

Why the sagittal plane matters. Sagittal imbalance correlates strongly with pain and disability, and the health-related quality of life (HRQOL) measures (ODI, SF-36) correlate with the sagittal parameters. An SVA over 50mm is associated with significant disability, and PI-LL mismatch is the strongest predictor of poor HRQOL.

The chain. The spine functions as a chain of segments that must keep the centre of gravity over the pelvis and lower extremities. Loss of lumbar lordosis is the primary driver of sagittal imbalance.

Normal sagittal alignment. The values against which a deformity is judged:

• Cervical lordosis 20-40 degrees
• Thoracic kyphosis 20-50 degrees (T4-T12)
• Lumbar lordosis 40-60 degrees (L1-S1)
• Sacral slope 30-50 degrees
• Pelvic tilt low, under 20 degrees, which is the SRS-Schwab grade boundary; the ideal rises with age (Lafage, PMID 26689395; figures in Controversies)

Pelvic incidence (PI). The angle between the perpendicular to the sacral endplate and the line to the femoral head centre, normal range 40-65 degrees. It is a fixed anatomical parameter that does not change after skeletal maturity, and it determines the amount of lumbar lordosis required for balance. The fundamental equation is PI = PT + SS.

Pelvic tilt (PT). The angle between the vertical and the line from the S1 midpoint to the femoral head centre. It is positional, changing with posture, and it increases with pelvic retroversion as the pelvis compensates; a PT over 30 degrees indicates exhausted compensation.

Sacral slope (SS). The angle between the sacral endplate and the horizontal. It decreases with pelvic retroversion, and it follows from the equation: SS = PI - PT.

The compensation cascade. When lumbar lordosis is insufficient for a given PI, a predictable cascade of compensation occurs:

When compensation is exhausted, the C7 plumb line falls anterior to the sacrum (positive SVA) and the patient becomes symptomatic.

History. The symptoms that bring the patient in:

• Back pain (axial, often positional)
• Radiculopathy or neurogenic claudication
• Difficulty standing upright
• Need to support the trunk with the hands on the thighs
• Decreased walking tolerance
• Progressive postural change

The elements of the history that shape the plan are the duration and progression of symptoms, prior spinal surgery (fusion levels, approach), walking capacity in blocks or time, pain location and character, neurological symptoms (weakness, numbness, bowel or bladder), and medical comorbidities and bone health.

Observation. Standing posture in the sagittal and coronal planes, forward trunk lean, a hip and knee flexion posture, and shoulder balance.

The spine. Flexibility on the forward bend test, sagittal balance by plumb line, coronal balance, and the chin-brow vertical angle in a fixed deformity.

Neurology. Motor strength through the L2-S1 myotomes, sensation, reflexes, straight leg raise and gait.

Hips, knees and function. Hip range of motion for a flexion contracture, examination of the knees, and overall mobility and function.

Patient-reported outcomes. The measures used to assess surgical outcome:

• Oswestry Disability Index (ODI)
• SF-36 (physical and mental component scores)
• Scoliosis Research Society-22 (SRS-22)
• Visual Analogue Scale (VAS) for pain
• EQ-5D

Standing full-length radiographs. The gold standard for alignment assessment: standing AP and lateral on a 36-inch cassette, including C2 to the femoral heads, with the arms positioned (hands on the clavicles or held in front).

View larger

Supine and bending radiographs. These assess the flexibility of the curves and help plan the fusion levels and the need for an osteotomy. A supine lateral over a bolster assesses sagittal flexibility.

MRI of the whole spine. For neural compression, disc degeneration, the cord and the cauda equina, and to rule out intraspinal pathology.

CT, when indicated. For bone quality, a previous fusion mass, existing hardware, and osteotomy planning.

The measurements. Each parameter against its normal range:

Bone density. DEXA of the hip and spine, and consider CT-based bone density (Hounsfield units). Bone quality shapes surgical planning and the fixation strategy, and osteoporosis significantly increases complication risk.

Additional studies. Ordered when the question demands them:

• CT myelogram if MRI is contraindicated
• Flexion-extension radiographs for instability
• Standing hip-to-ankle films for limb length and hip assessment
• Pulmonary function tests for severe deformity

ASD surgery has significant complication rates. Understanding and communicating these risks is essential, but quote figures that come from a named population, because the ranges in circulation ("25-80%") span populations that are not comparable.

The gradient. Across 578 thoracolumbar fixed sagittal-deformity cases in the SRS Morbidity and Mortality database the overall complication rate was 29.4%, and it rose stepwise with the osteotomy performed (the figures are in the table below). Performing any osteotomy roughly doubled the adjusted odds of a complication (OR 2.07). A VCR buys 45-70 degrees at roughly three and a half times the complication rate of operating without an osteotomy, so the osteotomy grade should be dictated by the correction the deformity actually requires, never by what is technically available. This is a large multicentre registry (Level III) of a fixed-deformity population, so these are the rates for the harder end of the spectrum.

These figures are not a head-to-head comparison and no randomised comparison exists. Frames of reference differ (a whole-cohort registry, a single-institution three-column series, a pooled PVCR review), and the patients differ from the constructs: a VCR is chosen for the more rigid, more severe, more often revision deformity, so part of every gradient is the deformity and not the operation.

Early complications. The incidence and the response to each:

Proximal junctional kyphosis (PJK). The most common mechanical complication. It is measured as the sagittal Cobb angle from the inferior endplate of the UIV to the superior endplate of UIV+2, two levels above, not one, and the Glattes criterion requires a proximal junctional angle of 10° or more that is also at least 10° greater than the preoperative value (PMID 16025035). No universal definition exists: some groups measure to UIV+1 and some compare against the early postoperative rather than the preoperative film, which is a large part of why reported incidence ranges from 10% to 48% (Glattes' own incidence was 26%).

Risk factors for PJK. Later series report older age, over-correction, osteoporosis, an upper-thoracic UIV and posterior-only surgery. Glattes himself identified none, and found SRS-24 scores were not worse in PJK patients, so radiographic PJK is not automatically a clinical problem. It may require extension of the fusion.

Proximal junctional failure (PJF) is the structural form: fracture of the UIV or UIV+1, instrumentation pull-out, sagittal subluxation, or myelopathy, not simply "PJK that gets revised".

Rod fracture. Occurs in 5-20% of cases, with the higher risk at the osteotomy site. It may be asymptomatic if the fusion is solid; revise if painful or progressing.

Pseudarthrosis. Failure of fusion. The risk factors are smoking, diabetes, multilevel surgery and a previous failed fusion; treatment is revision with bone grafting and possibly an osteotomy.

Adjacent segment disease. Degeneration above or below the fusion, more common with long rigid constructs. It may require extension of the fusion.

Risk factors for complications. The patient and operative factors that raise the rate:

• Advanced age (over 70)
• Higher ASA grade
• Obesity
• Osteoporosis
• Smoking
• Three-column osteotomies
• Revision surgery
• Long operative time

The first fortnight. The sequence from theatre to discharge:

Rehabilitation. Activity advances in phases:

Bracing. Practice varies with surgeon preference. Some protocols use a TLSO for 6-12 weeks; it may provide comfort and remind the patient of precautions.

Follow-up. Clinical and radiographic review at fixed points:

Radiographic outcomes. Correction of the SVA and the PI-LL mismatch to the alignment targets, restoration of coronal balance, and the fusion rate.

View larger

Clinical outcomes. The measures and their MCIDs:

• ODI improvement (MCID: 12-15 points)
• VAS pain reduction (MCID: 2 points)
• SRS-22 improvement
• SF-36 improvement

What to expect. Approximate rates after successful surgery:

• SVA correction achieved: 70-85%
• Fusion rate: 85-95%
• Significant pain improvement: 60-75%
• Patient satisfaction: 70-80%

What shapes the result. The most consistent predictor of a good outcome is achieving the appropriate PI-LL match (under 10 degrees mismatch): under-correction is associated with persistent symptoms, and over-correction increases the risk of PJK. Beyond alignment, the result depends on the baseline severity of the deformity, whether a major complication occurs, the patient's age, comorbidities and bone quality, whether they keep smoking, and whether the surgery is a revision rather than a primary procedure.

Global Epidemiology

Adult spinal deformity is common and strongly age-related. In a healthy elderly volunteer cohort, 68% met the radiographic definition of scoliosis (Cobb greater than 10 degrees), against earlier general-population estimates of up to 32% (Schwab et al. Spine 2005, PMID 15864163). Disability tracks with sagittal malalignment rather than coronal Cobb angle: positive sagittal balance is the radiographic parameter most strongly correlated with poor health status, and severity worsens linearly with C7 plumb-line deviation (Glassman et al. Spine 2005, PMID 16166889). As populations age worldwide, demand for both non-operative and operative ASD care is rising in every region.

Classification and Alignment Frameworks (side by side)

The SRS-Schwab system and its sagittal modifiers are accepted across SRS (international), AOSpine / AO Foundation educational frameworks, EuroSpine / European deformity societies, and North American (AANS/CNS, NASS) practice. There is broad international consensus that sagittal realignment goals (PI-LL within roughly 10 degrees, SVA under 50mm, PT under 20-25 degrees) drive outcomes, with a clear modern shift toward individualised (age-adjusted and GAP-proportioned) rather than fixed population targets.

Registry and Database Evidence

Unlike arthroplasty, ASD has no single dominant implant registry; the strongest comparative evidence comes from large prospective multicentre deformity databases (the International Spine Study Group in North America and the European Spine Study Group). These have generated the landmark complication and alignment data summarised above: the stepwise osteotomy complication gradient (SRS Morbidity and Mortality database; Smith et al. Spine 2011, PMID 21192289) and the persistent mechanical-failure risk even in well-aligned patients (Haddad/Yilgor et al. Spine J 2024, PMID 39332683).

Global Practice Variation

In high-resource settings, ASD correction is concentrated in tertiary deformity units with full-length EOS/stereoradiography, intraoperative neuromonitoring, cell salvage and high-dependency/ICU support; staged or multidisciplinary "deformity MDT" pathways and bone-health optimisation (DEXA, anti-resorptive/anabolic therapy for osteoporosis) are increasingly standard. In limited-resource settings, access to long-cassette imaging, neuromonitoring and revision capacity is constrained, so thresholds for major three-column osteotomy are often higher and non-operative management or shorter constructs may predominate. Across all settings, the high complication burden mandates careful patient selection, shared decision-making and referral to experienced deformity surgeons.

Differential Diagnosis

The single PI-LL target under 10 degrees is the classic teaching, and the age-adjusted evidence says it is wrong for older patients. Across 773 patients, the ideal spinopelvic values shifted markedly with age: under 35 years the targets are approximately PT 11 degrees, PI-LL minus 10 degrees, SVA 4mm, while over 75 years they are approximately PT 28 degrees, PI-LL plus 17 degrees, SVA 78mm. Applying a fixed population target to a 78-year-old therefore does not merely miss the mark — it over-corrects them, and over-correction is exactly what raises PJK risk, which the Outcomes section above already identifies as the failure mode.

Both positions are defensible and a good answer holds them together: PI-LL under 10 degrees remains the right instinct in the younger patient and the right thing to quote as the classic Schwab modifier, while realignment goals in the elderly should be relaxed deliberately rather than by accident. State the classic target, then state that it is age-dependent.

Proportional alignment refines the target further, and its gradient is the argument for it. The GAP score grades the postoperative spinopelvic state as proportioned, moderately disproportioned or severely disproportioned, with mechanical complication rates of 6%, 47% and 95% respectively — a spread wide enough that the score is about planning, not description.

But alignment alone does not abolish mechanical failure, and this is the honest limit. In patients over 55 who achieved a proportioned GAP state, the mechanical complication rate was still 40% at four years, with 18% requiring revision specifically for mechanical failure. Independent risk factors were the residual coronal lumbosacral curve, the number of instrumented levels and relative spinopelvic alignment, alongside higher body weight and frailty.

And "proportioned" is not one thing. Within that supposedly well-aligned group the failure rate still climbed with every GAP point: 31% at GAP 0, 54% at GAP 1, 75% at GAP 2 — so a score of 2 is not the same reassurance as a score of 0, and the useful operative thresholds derived from the same data are a residual lumbosacral curve of 4 degrees or less and a relative spinopelvic alignment of 3 degrees or less. The lesson is that radiographic perfection is necessary but not sufficient — the coronal plane and the host both matter, and a patient should not be promised that a well-planned correction removes the risk of reoperation. Note also what this cohort did not find: three-column osteotomy, postoperative bracing and an added anterior approach made no difference to mechanical failure here, which with only 83 patients is far more likely to be absent power than absent effect.