Adult spinal deformity (ASD) is a complex disease state that pathologically alters standing upright posture and is associated with substantial pain and disability. This article provides an overview of classification systems for spinal deformity, clarifies the need to differentiate between pediatric and adult classifications, and provides an explanation on the use of the Scoliosis Research Society-Schwab Adult Spinal Deformity Classification (SRS-Schwab ASD Classification). This information allows surgeons, researchers, and health care providers to (1) identify sources of pain and disability in patients with ASD and (2) accurately use the SRSeSchwab ASD Classification to evaluate patients with ASD.
Key points
- •
Classification systems should describe important features of disease states and provide clinical information regarding the classified disease state.
- •
Previous classifications for adult spinal deformity (ASD) neglected evaluation of sagittal spinopelvic parameters.
- •
Pelvic incidence/lumbar lordosis mismatch (PI-LL), increased sagittal vertical axis (SVA), and increased pelvic tilt (PT) correlate strongly with pain and disability in patients with ASD.
- •
The Scoliosis Research Society (SRS)–Schwab ASD Classification describes the location of the scoliotic curve and uses 3 sagittal modifiers (PI-LL, SVA, and PT) to evaluate the sagittal plane and correlate spinal deformity with patient pain and disability.
- •
Greater spinal deformity grade on the SRS-Schwab ASD Classification predicts patient disability.
Introduction
Classification systems are created to provide organization to pathologic conditions and provide treatment options for disease states that share a common theme. A classification ideally provides a cohesive approach to the disease state that (1) identifies different severities of the disease state (often in a hierarchical manner), (2) facilitates communication between health care providers and researchers to assure accuracy and reproducibility in describing the disease state, (3) allows for comparison of different treatment methods and, as a consequence, (4) allows for creation of accurate treatment recommendation guidelines. From a statistical standpoint, a classification system should have high construct validity (the extent to which classification accurately measures the disease state) and high reliability as shown by high intrarater reliability (consistent grading by 1 rater at different time points) and inter-rater reliability (consistent grading by different raters). The classification should also have high reproducibility (the degree of agreement between measurements on replicate specimens in different locations by different observers). This article provides an overview of existing classification systems for spinal deformity and highlights the challenges of creating an effective classification for adult spinal deformity (ASD). This article then focuses on the Scoliosis Research Society (SRS)–Schwab ASD Classification, including the rationale behind the development of the SRS-Schwab ASD Classification, guidelines for use of the SRS-Schwab ASD Classification, and initial data on use of the classification.
Introduction
Classification systems are created to provide organization to pathologic conditions and provide treatment options for disease states that share a common theme. A classification ideally provides a cohesive approach to the disease state that (1) identifies different severities of the disease state (often in a hierarchical manner), (2) facilitates communication between health care providers and researchers to assure accuracy and reproducibility in describing the disease state, (3) allows for comparison of different treatment methods and, as a consequence, (4) allows for creation of accurate treatment recommendation guidelines. From a statistical standpoint, a classification system should have high construct validity (the extent to which classification accurately measures the disease state) and high reliability as shown by high intrarater reliability (consistent grading by 1 rater at different time points) and inter-rater reliability (consistent grading by different raters). The classification should also have high reproducibility (the degree of agreement between measurements on replicate specimens in different locations by different observers). This article provides an overview of existing classification systems for spinal deformity and highlights the challenges of creating an effective classification for adult spinal deformity (ASD). This article then focuses on the Scoliosis Research Society (SRS)–Schwab ASD Classification, including the rationale behind the development of the SRS-Schwab ASD Classification, guidelines for use of the SRS-Schwab ASD Classification, and initial data on use of the classification.
Background on spinal deformity classification systems
Most classifications traditionally used to describe spinal deformity have been oriented toward pediatric spinal deformities. In the past, the King-Moe Classification for Adolescent Idiopathic Scoliosis (AIS) has been the standard to describe scoliosis. The principle benefit of the King-Moe classification was that it provided a treatment algorithm based on curve type that allowed surgeons to determine the appropriate curves and vertebral levels for spinal fusion. Five types of curves were described in detail including type I, a double major curve, in which the thoracic and lumbar curves are considered structural and both curves should be included in the fusion; type II, a single major curve, in which the thoracic and lumbar curves cross the midline (center sacral vertical line [CSVL]), but the thoracic curve is larger and more rigid than the lumbar curve and therefore the thoracic curve is structural and only the thoracic curve should be included in the fusion; type III, in which only the thoracic curve crosses the midline, and only the thoracic curve should be included in the fusion; type IV, a long, sweeping thoracic curve in which L4 is titled toward the thoracic curve and L5 is centered over the sacrum, distal fusion level recommended to be the first vertebra bisected by the CSVL; type V, a double thoracic curve pattern in which the proximal and main thoracic curves are considered structural and are included in the fusion ( Fig. 1 ).
The King-Moe system remained the principal classification for AIS for more than 20 years, guiding evaluation and treatment; however, several limitations of the King-Moe system have been highlighted. First, the King-Moe system is not comprehensive, because isolated thoracolumbar and triple major curves were not described. All patients in the King and colleagues series received Harrington rod instrumentation that solely corrected deformity in the coronal plane via distraction, therefore the deformities were evaluated only in the coronal plane, failing to recognize scoliosis as a three-dimensional deformity and the need to assess the coronal, sagittal, and axial planes. The King-More system has also shown fair to poor interobserver and intraobserver validity, reliability, and reproducibility by 2 separate studies. In response to these shortcomings, Lenke and colleagues developed a classification system for the operative treatment of AIS. This classification was designed to be (1) comprehensive, to include all AIS curve types; (2) provide two-dimensional analysis with increased emphasis on evaluation of the sagittal plane; (3) treatment based, advocating selective arthrodesis only of the structural curves; and (4) provide objective criteria to differentiate individual curve types and provide guidelines for fusion. The Lenke classification describes 3 curve regions (proximal thoracic [PT], apex at T3, T4, or T5; main thoracic [MT], apex between T6 and the T11–T12 disc; and thoracolumbar/lumbar [TL/L], apex at T12 or L1 for thoracolumbar curves, and between the L1–L2 disc and L4 for lumbar curves) and 2 curve types (major curve, the largest measured curve; minor curve, the smaller curves). The minor curves are then established as structural or nonstructural by evaluating curve flexibility and sagittal alignment. Structural curves show coronal plane rigidity (do not reduce less than 25° on side bending radiographs) and/or are focally kyphotic in the sagittal plane (focal kyphosis >20°). Focal kyphosis for the described curve regions is measured in the following areas: PT, T2 to T5; MT, T10 to L2; and TL/L, T10 to L2. Using the Lenke classification, 6 curve types can be assigned according to the identified major and minor structural curves, and are named according to the identified structural curves: type 1, MT curve (structural); type 2, double thoracic (PT and MT curves are structural); type 3, double major (MT and TL/L curves are structural, MT is larger on standing radiographs); type 4, triple major (PT, MT, and TL/L curves are structural); type 5, thoracolumbar/lumbar (TL/L curve structural); and type 6 (MT and TL/L are structural, TL/L is larger on standing radiographs). Treatment guidelines recommend selective fusion, advocating fusion only of those curves that are structural as per the classification guidelines ( Fig. 2 ).
The Lenke classification has been widely adopted for evaluation and treatment of AIS, with excellent success. However, despite the initial usefulness of the King-Moe and, subsequently, the Lenke classifications to describe AIS, there have been attempts to apply many of these same classification guidelines to ASD. The error in applying AIS and/or pediatric spinal deformity assessment guidelines to creating an ASD classification lies in heterogeneity of the clinical and radiographic presentation of ASD. Although cosmetic deformity and coronal malalignment are the most common reasons for presentation and treatment of pediatric patients with spinal deformity, especially AIS, pain is the primary complaint for patients with ASD. Effective AIS classifications have therefore focused on strategies to evaluate and treat scoliosis and coronal malalignment; however, classifications for ASD must quantify the major predictors of pain in the ASD population. Early attempts at developing ASD classifications by Aebi and by the SRS failed to integrate this clinical component of pain in the ASD population, and therefore, despite being descriptive of the cause and observed radiographic parameters, these classifications lacked clinical relevance. Scoliosis and coronal plane deformities are common for both AIS and ASD; however, it has been repeatedly shown that sagittal malalignment is a fundamental component of ASD, and that sagittal malalignment is a primary determinant of pain and disability in the ASD population. One advantage of the Lenke classification is that it calls for greater attention to the sagittal plane, although it does so within the confines of defining the structural behavior of a coronal deformity (idiopathic scoliosis) rather than purely quantifying the amount of sagittal plane deformity. Based on these concepts, initial work by Schwab and colleagues established a foundation for a clinical impact classification for ASD that integrated radiographic parameters correlating with poor health-related quality of life (HRQOL) parameters. This initial classification by Schwab and colleagues described 5 types of scoliosis based on the apical level of the curve: type I, thoracic only; type II, upper thoracic major (apex T4–T8); type III, lower thoracic major (apex T9–T10); type IV, thoracolumbar major curve (apex T11–L1); type V, lumbar major curve (apex L2–L4). Two radiographic parameters were then added as modifiers to the curve type: lumbar lordosis and intervertebral subluxation. Loss of lumbar lordosis and increased intervertebral subluxation correlated with poor HRQOL scores, so it is these modifiers that denoted the clinical impact of this classification. Soon after publication, it was recognized that global sagittal malalignment, measured by sagittal vertical axis (SVA; distance from the C7 plumb line to the posterior, superior corner of S1) is an equally important predictor of poor HRQOL scores as regional sagittal malalignment (loss of lumbar lordosis). Therefore the classification added a third modifier, termed the global balance modifier, as a final component to describe the radiographic deformity and predict the associated disability ( Box 1 ).

Full access? Get Clinical Tree

