5.3 Deformity severity measures: 5.3.1 Spine Classifications and Severity Measures
1 Classifications systems and outcome measures for adolescent idiopathic scoliosis
The history of spinal deformity classification started with the work from Schultess in 1905 [1]. He classified idiopathic scoliosis presenting in five different types: cervicothoracic, thoracic, thoracolumbar, lumbar, and combined. Due to its simplicity this simple classification is still the most commonly used classification. In 1954 James enlarged the classification from 5 to 9 types [2]. In 1983 King et al published their classification of thoracic curves in five different distinct patterns. To the present date, this classification system is still widely used for thoracic curve patterns [3]. In 1990 Winter and Lonstein described seven idiopathic curve patterns [4]. Upon reviewing 2,000 patients with idiopathic scoliosis in 1998 Coonrad and coworkers presented a new classification system that consisted of eleven categories that take into account the mirror image of each pattern [5]. This classification included the equally important sagittal variable as the next step in treatment planning. In 2001 Lenke in his classification system described six different curve patterns and added lumbar and sagittal modifiers to take into account the deviation of the lumbar curve from the midline and the sagittal contour of the patient. With such a lumbar modifier system Lenke described a total of 14 different curve pattern as identified in the coronal plane [6]. Taking into account the sagittal plane with his three sagittal modifiers (a, b, c) the amount of possible curves specified in this system rises to 42 subtypes. Very recently, in an attempt to have a more simplistic classification system, the Peking Union Medical College (PUMC) proposed a new classification system according to the number of apex vertebrae [7]. The term ‘apex of the curve’ utilizes the Scoliosis Research Society (SRS) definition. In this classification the deformity is characterized in single curve (type 1), double curve (type 2) and triple curve (type 3). Then each category is subdivided according to the different subtypes (a, b, c) depending on the location of the curve. For instance a type I is a single curve and a subtype Ia is a single curve thoracic, a subtype Ib is a single curve thoracolumbar, while a Ic curve presents a lumbar single apex scoliotic deformity. Sadly this simple system does not reflect sagittal balance.
2 Goal of classification systems in idiopathic scoliosis
The goal of any classification system is to have a comprehensive system where all curves can be integrated into the system
The classification system should be easily remembered by the spinal deformity surgeon
The system should be reliable with excellent interobserver and intraobserver reliability
The classification system should be useful for treatment recommendations and for research.
If we look at all the previously mentioned classification systems none fit all these criteria.
Are any of them comprehensive?
Each classification system claims to be comprehensive. However, experience clearly demonstrates that until now all classifications systems have failed to characterize all possible curves.
Can we remember them and can we teach them easily?
So many classification systems are now available that it is a real challenge for the spinal deformity surgeon to remember them all as their degree of complexity parallels the number of categories. Even for an experienced spine surgeon a diagram of the Lenke classification attached to the wall may be necessary to distinguish all the different groups in the different classifications available.
Are any of these classifications reliable?
Interestingly most new classifications are now published with their inter- and intraobserver reproducibility supporting their advantages. Invariably, these initial inter- and intraobserver ratings are usually reported to be excellent initially in the index publication by the originating authors. However when such classifications are tested by other investigators, such ratings tend to drop dramatically. For instance Coonrad reported a rate of 98.2% agreement between raters. This number dropped to 38% agreement when the classification was tested by other investigators [8]. Likewise the King classification has only 55% agreement between observers and 65% agreement on test-retest by the same observer when this psychometric was examined by other investigators [9]. Lenke reported an initial interobserver and intraobserver kappa values for the curve type to be 0.92 and 0.83, respectively for the five developers of the system and 0.74 and 0.89 for the independent group of seven scoliosis surgeons. However, Ogon in a more recent papers testing the Lenke classification found a mean kappa value of 0.62 for inter-observer reliability, and a mean kappa of 0.73 for intraobserver reliability [10]. Likewise Richards found the reliability of the Lenke classification not as good as initially reported [11]. As to the PUMC classification, this has shown excellent intraobserver and interobserver reliability [7], but was not tested by other investigators yet. While it is easy to blame these differing lower reproducibility results at least in part on lack of knowledge, understanding or experience of later authors, complexity of systems with indistinct parameters for a real-life situations undoubtedly have to share part of the blame.
Are any of these classifications useful for treatment guidelines?
King et al were the first group to determine treatment guidelines in their classification system [3], however subject to the specifications of a Harrington nonsegmental instrumentation. Similarly, the other classification systems claim to offer treatment guidelines aimed at the extent of fusion, however, often contradicting other classification systems or even previous publications from the same authors. As outcomes criteria and an associated grading system has yet to be published formal treatment guidelines are difficult to give.
What parameters/classifications are then useful in adolescent idiopathic scoliosis?
The SRS has come up with a glossary of definitions that have been widely accepted [12]?.
The location of the apex determines the type of curve; a thoracic curve has the apex located in the thoracic spine, a thoracolumbar curve has the apex between T12 and L1 and the lumbar curve has the apex below L1. With such definitions different curve patterns can at least be described with some clarity. For example, simple thoracic, double thoracic, double major, thoracolumbar and lumbar curve can thus be differentiated. This identification of the apex as defining for the basic curve distribution has been shown to be highly reliable over the years [6].
Radiographic determination of the end vertebrae (EV), neutral vertebrae (NV), and stable vertebrae (SV) have demonstrated good to excellent intraobserver, but poor interobserver reliability [12]. However, interobserver agreement increased significantly when concurrence within one vertebral level was assessed, with 91, 73, and 76% agreement for identifying the EV, NV, and SV [13], If given a small margin range these parameters become reliable.
The Cobb angle is a reflection of the curve severity. It has superseded the Ferguson method of measurement of the curve, and has been shown to be extremely reliable if one accepts a 5 degree margin of error.
Other parameters like apical vertebral translation, apical vertebral rotation, stiffness of the curves as assessed on traction films or on side bending and sagittal contour of the deformity are essential and can be considered like the Cobb angle as severity parameters. It is obvious that a thoracic curve whose apex is deviated from the midline by 70 mm (apical vertebral translation of 70 mm) will be more severe than if it is deviated by only 30 mm.
Overall the evaluation of one curve can include 7 different parameters in the coronal plane (apex, 2 end vertebrae, Cobb angle, apical vertebral translation, apical rotation, stable vertebra) and at least 4 in the sagittal plane (apex, 2 end vertebrae, Cobb)
The major advantage of these geometric parameters is that they are very reproducible and show little variation in the intraobserver and interobserver reliability tests, especially if one attributes a small range into these parameters [13, 14].
3 Integrated growth parameters to predict the progression of the curve
The Risser test, the state of the tri-radiate cartilage, evaluation of bone age and the Tanner scale can be predictors of the progression of the curve in skeletally immature patients. All these parameters or classifications have shown good reliability in terms of predicting curve progression relative to skeletal maturity. Recently the peak velocity height was found to be on of the best predictor of the “crankshaft phenomenon”, which describes progressive torque like deformation found in adolescent idiopathic scoliosis.
4 Outcome tools
Outcomes tools have always been essential for all parties involved in scoliosis care, patients, treating physicians, third party payers and for research purposes. However actual measurement of outcome in spinal deformity patients is a complex undertaking. For instance, the outcome of a spine fusion in a skeletally mature right thoracic curve of 50 degrees in a skeletally mature adolescent idiopathic scoliosis female patient with clinically barely detectable deformity will be quite difficult to determine. Corrective surgery will not impact basic health parameters, such as pulmonary function, for a long time. The main indication for this initially quite painful surgery is to prevent curve progression and therefore prevent adverse cardiopulmonary effects associated with curve deterioration decades from now. This anticipated health impact is difficult if not impossible to quantify. In contrast, the goal of operating on a 45 degrees thoracolumbar curve with marked waistline asymmetry in a slender female teenager is quite different. This is purely a cosmetic operation. How can we judge the cosmetic result of such surgery? Should patient perception which is notorious to change over time be prioritized over an operating surgeon’s impression or could an attempt at objective scoring, for instance with standardized clinical photographs pre and post operatively be developed? In contrast spinal deformity in an infantile population with severe congenital deformity can take on a profound role in extending survival time and decrease care needs in children unable to articulate pain or relay concrete direct feedback on their well being to their examiners. These few cases demonstrate the difficulty in evaluating surgery in deformity surgery with goals varying dramatically from one patient to the next. Despite these more skeptical remarks there have been increasing attempts to create meaningful outcomes instruments that are more disease and treatment specific.
5 Function, pain, and cosmesis
Assessment of pain, cosmesis, self-perception of the disease, and function are essential foundations for patients with idiopathic scoliosis. A variety of outcome measurements have been published, with the Scoliosis Research Society (SRS) Outcomes Instrument, the SRS 30 questionnaire, having become the most utilized one. Derived from the SRS 22 and 24 questionnaire and translated and validated in three different languages (English, Spanish, and Turkish) its five domains gives an evaluation of pain, self-image/appearance, mental health, function/activity, and satisfaction with management [15]. The SRS Mental Health and Pain domain scores can be accurately calculated from correlating SF-36 domain questionnaire. SRS Function scores can be fairly well predicted from correlated SF-36 domain scores. Self-image and management satisfaction/dissatisfaction domain scores, however, cannot be approximated from SF-36 domain scores [16]. Sadly actual results of SRS instrument based studies have been disappointing due to indistinct results found for a majority of patients. The SRS instrument for instance does not appear to be able to discriminate between different treatments. For example, the outcomes of total SRS scores were not significantly different among three groups that underwent different treatments (brace only surgery or brace plus surgery) in a recent study done by Weigert [17]. Likewise the SRS questionnaire does not discriminate among patients with single, double, or triple curves [18], and cannot differentiate bad and poor outcomes.
Assessment of cosmesis has been done with many different tools and scales, the latest one being the Walter Reed Visual Assessment Scale (WRVAS) [19]. The WRVAS scores correlate significantly with curve magnitude and treatment. Parents and patients have similar scores, lending construct validity to this type of assessment tool. Interestingly when comparing the patient perception of cosmesis with the spine surgeon’s perception, there is little if any correlation [20]. Other concerns include the well described variation over time in patient satisfaction regarding the cosmesis domain. For example, the perception of a 13 years old teenager who just had her surgery and who is happy to have recovered from it may vary from her perception of her cosmesis 10 years later when she has forgotten the way she used to look. This is one of the reasons that surgery for adolescent idiopathic scoliosis preferably is carried out with clinical photographs of the patient pre and post surgically in support of the change. Similar to a plastic surgeon that takes clinical photographs before his reconstructive surgery the spinal deformity surgeon should do the same so the patient can remember and the surgeon can have some objective reference to judge his/her “art-” work.
6 Function, pain, and cosmesis—radiographic outcome tools
Radiographic outcome tools do not exist as such, but rely on a number of different parameters. Aside from attempting to establish radiographic fusion or the presence of hardware integrity or loosening, spine surgery has not emphasized many outcomes criteria. For a long time the only radiographic parameter of a good result in deformity surgery was thought to be the amount of Cobb angle correction, with the concept being the straighter the instrumented spine looked, the better the surgery and the better the surgeon. It has now been recognized that the post operative value of the Cobb angle correction is of little value when evaluating the x-rays of a scoliosis correction. For instance the concept of ‘straighter is better’ in terms of Cobb angle correction has not been supported. It has further been recognized that overcorrecting of certain curves could lead to shoulder or trunk imbalance or even decompensation. Far more important than the amount of Cobb angle correction are the overall balance of the patient, the numbers of fused vertebrae, the “harmonious ” contour of the spine and the fate of the unfused lumbar spine below the instrumented area. The spine deformity surgeon looking at his/her own work should not be a Cobb angle mathematician. Rather, the surgeon should step back and look at his/her work in a more artistic fashion and appreciate the contour of the corrected spine, especially the unfused part.
7 Function, pain, and cosmesis—the art of spinal deformity surgery
The art of spinal deformity surgery can be compared with the building of a bridge or a tower, or the drawing of a painting. The artist will look at the overall balance of the patient, the harmonious shape of the spine both in the sagittal and coronal plane, the shoulder balance, the rotation of the scapulae, the residual thoracic hump, and the waist line symmetry. The architect will judge the appropriateness of the instrumentation in terms of its extent and proper positioning, the soundness of the instrumentation with respect to whether it is biomechanically solid, or whether it is too long or too short. Unquestionably these are highly subjective assessment criteria.
8 Function, pain, and cosmesis—the future
Shall we need any classification systems? According to Lenke, “The ultimate goal of any classification system is to allow organization of similar curve patterns to provide comparisons of various treatment methods to provide optimal treatment for each spinal deformity surgical patient”. In the world of electronic database and fast Internet technology shall we need any classification system in the near future? Patients x-rays from the Hospital PACS systems and digital photographs can be entered very easily in any database as well as outcome tools like the SRS 30 questionnaire. More recently an online registry named Scolisoft has been created to simplify comprehensive data gathering in a systematically structured database. Translated in 5 different languages the Scolisoft registry database is characterized by its demographics, its anatomical radiographic parameters, the clinical photographs of the patient, and the outcome instruments of the SRS [21]. Morbidity and mortality of the surgical procedure are also recorded following the SRS morbidity and mortality database. In difference to other databases this system allows visualization of all the different parameters registered. Therefore it is possible to query the software with an instant answer. Such query will try to identify identical patients or a range of similar patients in terms of demographics and curve characteristics, and observe the variation in treatment and outcome. With such software the need of true classification will become far less important as the user can visualize online the result of his/her query based on variables entered.
Is there a need for better outcomes measurement tools for spinal deformity surgery? To date existing instruments have shown limitations in demonstrating the effect of any form of treatment intervention. Ideally a panel of experts could attempt to address these shortcomings by consensus agreement on necessary modifications of outcome instruments. Ultimately a scoring system is needed that would enable us to grade the outcome of our treatment. What are the surgeon’s accepted parameters of good outcomes, for instance, in the surgical treatment of adolescent idiopathic scoliosis? Experienced spinal deformity surgeons agree that it is not the number of screws or hooks in the spine, but rather the number of saved lumbar segments, the ultimate fate of the unfused lumbar spine, and overall balance of the spine. A comprehensive spinal deformity database which includes clinical images as well as radiographic parameters of correction will hopefully provide a foundation for a future scoring system. Such a system will demonstrate comparable overall results for a similar group of patients and curve patterns despite inter-individual variances within a highly functional category. Such a system would also be expected to identify most lumbar curves lower scores than their thoracic counterpart due to greater fusion-related activity restrictions. Larger cohort groups assembled in such a database will limit distortion caused by homogeneity of an observed population. Cosmetic parameters will need to be identified and receive a scoring system as well. Such cosmetic scoring system must take into account the patient perspective and some objective clinical parameters such as scapula rotation, trunk shift or shoulder elevation. By quantifying the three entities of uneven shoulder, waist lines asymmetry or rib prominence we can hopefully finally assess how much these deformities impact a patient. Again, a multimodal patient assessment hopefully could untether these three deformity components since a patient with a solitary rib prominence will have fewer complaints associated with shoulder symmetry compared with somebody who has the three components of the spinal deformity. These questions can realistically only be obtained with an integrated approach of comprehensive data gathering from the point of first contact onwards. Such a futuristic scoring system would also render dependence on classification systems, which seem to go in- and out of fashion with some regularity more unnecessary.
9 Function, pain, and cosmesis—conclusions
Classification systems such as the Lenke classification constitute tremendous improvements over the last 10 years as the sagittal contour of the spine and the lumbar modifiers have allowed us to break down spinal deformity curves into clinically more relevant and distinct categories. However the advent of electronic database that can allow for vastly superior data comparison may render current classification systems obsolete. Developments such as shape recognition software can greatly enhance our ability to objectively quantify spinal deformities in an age independent and three-dimensional fashion.
Outcome measures have been developed recently and the SRS questionnaire has been shown to be a valid instrument to assess patients with adolescent idiopathic scoliosis. However its ability to discriminate between different treatments has not yet been demonstrated. The need for a scoring system to evaluate the results of our deformity surgery should be our main objective for the future. Such a scoring system should be based on a patient questionnaire and on objective radiographic and cosmetic parameters. An electronic library that incorporates all these outcome parameters is essential to determine the best treatments for our patient.
10 References
1. Scheier H, Dvorak J, Sandler A (1994) Wilhelm Schulthess. Spine; 19:2239-2243. 2. James J (1954) Idiopathic scoliosis. The prognosis, diagnosis and operative indications related to curve patterns and age of onset. J Bone Joint Surg; 36-B:36–49. 3. King HA, Moe JH, Bradford DS, et al (1983) The selection of fusion levels in thoracic idiopathic scoliosis. J Bone Joint Surg Am; 65:1302–1313. 4. Winter RB, Lonstein JE (1992) Idiopathic scoliosis. In Rothman-Simeone The Spine, Rothman R, Simeone F (eds). Philadelphia: W.B. Saunders, 373–430. 5. Coonrad RW, Murrell GA, Motley G, et al (1998) A logical coronal pattern classification of 2,000 consecutive idiopathic scoliosis cases based on the scoliosis research society-defined apical vertebra. Spine; 23:1380–1391. 6. Lenke LG, Betz RR, Harms J, et al (2001) Adolescent idiopathic scoliosis: a new classification to determine extent of spinal arthrodesis. J Bone Joint Surg Am; 83-A:1169–1181. 7. Qiu G, Zhang J, Wang Y, et al (2005) A new operative classification of idiopathic scoliosis: a peking union medical college method. Spine; 30:1419–1426. 8. Behensky H, Giesinger K, Ogon M, et al (2002) Multisurgeon assessment of coronal pattern classification systems for adolescent idiopathic scoliosis: reliability and error analysis. Spine; 27:762–767. 9. Cummings RJ, Loveless EA, Campbell J, et al (1998) Interobserver reliability and intraobserver reproducibility of the system of King et al. for the classification of adolescent idiopathic scoliosis. J Bone Joint Surg Am; 80:1107–1111. 10. Ogon M, Giesinger K, Behensky H, et al (2002) Interobserver and intraobserver reliability of Lenke’s new scoliosis classification system. Spine; 27:858–862. 11. Richards BS, Sucato DJ, Konigsberg DE, et al (2003) Comparison of reliability between the Lenke and King classification systems for adolescent idiopathic scoliosis using x-rays that were not premeasured. Spine; 28:1148-56; discussion 1156–1157. 12. Dewald R, Arlet V, Carl A, et al (2002) Spinal deformities: the comprehensive text. Thieme: New York. 13. Potter BK, Rosner MK, Lehman RA, Jr., et al (2005) Reliability of end, neutral, and stable vertebrae identification in adolescent idiopathic scoliosis. Spine; 30:1658–1663. 14. Kuklo TR, Potter BK, O’Brien MF, et al (2005) Reliability analysis for digital adolescent idiopathic scoliosis measurements. J Spinal Disord Tech; 18:152–159. 15. Climent JM, Reig A, Sanchez J, et al (1995) Construction and validation of a specific quality of life instrument for adolescents with spine deformities. Spine; 20:2006–2011. 16. Lai SM, Asher M, Burton D (2006) Estimating SRS-22 quality of life measures with SF-36: application in idiopathic scoliosis. Spine; 31:473–478. 17. Weigert KP, Nygaard LM, Christensen FB, et al (2006) Outcome in adolescent idiopathic scoliosis after brace treatment and surgery assessed by means of the Scoliosis Research Society Instrument 24. Eur Spine J; 15:1108–1117. 18. Asher M, Min Lai S, Burton D, et al (2003) Discrimination validity of the scoliosis research society-22 patient questionnaire: relationship to idiopathic scoliosis curve pattern and curve size. Spine; 28:74–78. 19. Sanders JO, Polly DW, Jr., Cats-Baril W, et al (2003) Analysis of patient and parent assessment of deformity in idiopathic scoliosis using the Walter Reed Visual Assessment Scale. Spine; 28:2158–2163. 20. Buchanan R, Birch JG, Morton AA, et al (2003) Do you see what I see? Looking at scoliosis surgical outcomes through orthopedists’ eyes. Spine; 28:2700–2704. 21. Scolisoft spine registry at www.scolisoft.com.Related posts:
5.1 Spine Classifications and Severity Measures
5.3 Deformity severity measures: 5.3.2 Spine Classifications and Severity Measures
5.3 Deformity severity measures: 5.3.4 Deformity severity measures: spondylolisthesis
5.5 Osteoporosis severity measures
5.4 Degenerative measures: 5.4.1 Degenerative measures: cervical myelopathy
5.7 Tumor severity measures
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
