5 Scheuermann’s Disease
Scheuermann’s disease is a rigid angular kyphosis usually occurring after puberty. The apical vertebral bodies are wedge shaped in the median sagittal plane with a shorter anterior height. The prevalence rate of thoracic Scheuermann’s kyphosis is of the order of 8%, not dissimilar to idiopathic scoliosis, and there is also a familial trend (▶ Fig. 5.1). 1 However, the condition is bimodal as regards site, with two-thirds of cases apical at T8, similar to idiopathic thoracic scoliosis, and the remainder at the thoracolumbar junction or upper lumbar region. 2 The clinical features of the kyphosis at these sites differ such that there is merit in considering two types: type I, apical at T8 (▶ Fig. 2.27 and ▶ Fig. 5.2a), and type II, apical lower down (▶ Fig. 5.2b), as implied by Scheuermann. Type I appears to be an idiopathic deformity and would merit being called idiopathic kyphosis. While an epidemiological survey has not been performed, boys are affected more commonly than girls. 1, 2 The preponderance of females in Bradford’s study of those presenting clinically suggests that girls may be more concerned about their appearance and hence present more frequently than boys, rather than having a truly higher prevalence rate. 3 In contradistinction, type II Scheuermann’s kyphosis appears to be related to strenuous physical activity, affects almost solely males, and presents around maturity usually with pain rather than deformity. 1, 4– 8 Accordingly it is referred to as “apprentice’s spine.”
Fig. 5.1 Brother and sister standing sideways, both showing mild Scheuermann’s disease. They were unhappy about being round backed (Ponte’s disease).
Fig. 5.2 (a) Lateral radiograph showing type I Scheuermann’s disease with apical wedging of three consecutive vertebrae by 5 degrees. (b) Lateral radiograph of type II thoracolumbar Scheuermann’s disease with two consecutive thoracolumbar vertebrae significantly wedged anteriorly.
The most important factor governing the behavior of the deformity is the axis of spinal column rotation which, in Scheuermann’s, or any other kyphosis, lies well in front of the apex, thus protecting the spine from rotation (▶ Fig. 3.4). The deformity is therefore not rotationally unstable and so remains in the median plane 9 unlike the opposite condition of idiopathic thoracic scoliosis.
In all respects type I Scheuermann’s disease appears to be precisely the opposite deformity to idiopathic scoliosis in the median sagittal plane (kyphosis vs. lordosis) (▶ Fig. 5.3). This is primarily another example of median sagittal plane spinal asymmetry with anterior vertebral wedging and end plate irregularity, the reverse of idiopathic scoliosis. Both conditions show the same site of predilection at the lower two-thirds of the thoracic spine, both start from a normal spine in normal children during growth, both are familial, and both have a strong tendency to stop progressing at spinal maturity. Whatever the fundamental mechanisms in play, the final common pathway is a disturbance of growth between the front and back of the spine, the former having relatively decreased growth in Scheuermann’s disease.
Fig. 5.3 (a) True lateral radiograph of an idiopathic thoracic curve showing more posterior Schmorl’s node formation in the lower vertebrae which are lordotically shaped. (b) Lateral X-ray of a type I Scheuermann’s idiopathic thoracic hyperkyphosis. The Schmorl’s node formation is clearly seen anteriorly. Idiopathic thoracic scoliosis and type I Scheuermann’s disease are opposite deformities in the sagittal plane. (Reproduced with permission from Newton P, O’Brien M, Shufflebarger H, et al. Idiopathic Scoliosis: The Harms Study Group Treatment Guide. Stuttgart/New York: Thieme; 2010: 36.)
Scheuermann’s original concept that this was an apophysitis, avascular necrosis of the cartilage ring apophysis, 10 was subsequently invalidated by the observations that the ring apophysis has nothing whatsoever to do with spinal growth, which is the responsibility of the end plate epiphysis. 11– 15 Furthermore, the common radiographic finding of apophyseal irregularity is not more prevalent in patients with Scheuermann’s kyphosis than straight-backed counterparts. Meanwhile Schmorl felt that the growth asymmetry between the front and back of the growth plate was attributable to intravertebral protrusions of disk material more anteriorly so that anterior growth was relatively suppressed. 16 That central Schmorl’s node formation is a common finding on a lateral radiograph of the normal-shaped spine does not invalidate this hypothesis and may well reflect that during the phase of spinal growth we are all vulnerable to the consequences of vertical loading, with Scheuermann’s kyphosis being one severe end of the spectrum and idiopathic scoliosis the other. Indeed, in Scheuermann’s kyphosis the Schmorl node formation is clearly anterior 17, 18 (▶ Fig. 5.3), and in idiopathic scoliosis clearly posterior, 9 whereas in normal spines this node formation is randomly scattered and more central. 16
There is certainly no clear evidence that Scheuermann’s kyphosis is associated with any upset of endocrine balance or calcium homeostasis. 3, 19 Anteroposterior (AP) views of the kyphotic region often show bilateral paravertebral shadows, suggesting that thickening of the anterior longitudinal ligament and periosteum are also associated with the end plate pathology, although very probably secondarily. Certainly at surgery on the front of a Scheuermann’s kyphosis these structures do appear thickened and rather tight. The common factor among all these different pathological findings is a disturbance of end plate growth anteriorly, such that anterior vertebral body height progressively lags behind posterior around the apex of the curve. This could occur as a result of a series of repetitive minor traumatic events which may have the effect of producing a type V epiphyseal plate injury 20 with growth suppression anteriorly, while precisely the opposite would occur with idiopathic scoliosis, where the reduction in vertebral height is posterior. Histological examination of vertebral body material removed through anterior surgery shows nothing more than would be compatible with this theory. 19, 21– 23
Ascani has considered that growth hormone secretion might be excessive 24 and although Murray 25 has shown normal heights in patients with Scheuermann’s disease, these heights were uncoiled. Meanwhile when heights were titrated against the height measures of the normal population in Leeds heights did seem to be increased in both males and females. 26 That Scheuermann’s disease might be a form of juvenile osteoporosis 19 has been disproven by the work of Gilsanz who has reported normal vertebral bone density. 27
Very interestingly, both Willner in Sweden 28 and Oxborrow with the Leeds Group 29 have shown that the thoracic kyphosis changes markedly in shape with growth in normal children and that in both boys and girls the thoracic kyphosis flattens from around the age of 10 to 12 years (▶ Fig. 3.11) but increases again at the age of 14 when boys are growing fastest. Thus, the shape of the spine in the sagittal plane is under very subtle control and while the child goes through these changes without succumbing to a progressive deformity; the range would also include unfortunate individuals at each end, one group with an increased kyphosis and the other with a lordosis and subsequent idiopathic scoliosis (▶ Fig. 3.14 and ▶ Fig. 3.15). The latter is far more important as this group is vulnerable to the consequences of rotational progression because of the more posteriorly situated axis of spinal column rotation. 9
Delmas 30 and Stagnara 31 have both studied the lateral profile in normal children and have observed a wide range with flat backs on the one hand and increased round backs on the other. This, like all other aspects of growth and shape, is genetically determined 32 (flat-backed children come from flat-backed families and round-backed children, from round-backed families) and therefore it is quite reasonable to expect Scheuermann’s kyphosis and idiopathic scoliosis to have a similar prevalence rate and familial trend. 1 The similarity of the median plane changes in vertebral body shape in the two conditions has also been remarked upon. 33, 34 However, the more obvious change in shape of the apical Scheuermann’s vertebral bodies is due to this progressive deformity staying in the median plane (▶ Fig. 5.4). With idiopathic scoliosis the degree of pathological change in the median plane (the lordosis) is very much less because the instability of the lordotic configuration leads to progressive rotation away from the midline and a gradual reduction of the compression on the back of the end plate. 34 Accordingly while a severe Scheuermann’s kyphosis may measure in excess of 100 degrees, the lordosis in severe idiopathic thoracic scoliosis is only of the order of 10 to 15 degrees, 35 the rest of the deformity being expressed as Cobb angle in the frontal plane (▶ Fig. 3.1 and ▶ Fig. 3.16).
Fig. 5.4 (a) The center of gravity of the body lies just in front of the lumbar spine, and with hyperkyphosis the thoracic spine is therefore progressively behind the axis of spinal column rotation. (b) Consequently, the deformity progresses solely in the sagittal plane, with no buckling potential. (Reproduced with permission from Newton P, O’Brien M, Shufflebarger H, et al. Idiopathic Scoliosis: The Harms Study Group Treatment Guide. Stuttgart/New York: Thieme; 2010: 37.)
5.2 Clinical Diagnosis
The clinical presentation and features of Scheuermann’s kyphosis vary according to whether the deformity is type I (thoracic) or type II (thoracolumbar or lumbar). Type I (thoracic) deformities present as a result of the deformity (▶ Fig. 5.5), with relatively few patients complaining of pain. It is the appearance of the back which is the problem and this is combined with poor general posture with the shoulders drooping forward. In less severe cases it is not usually the patient who complains of the specific deformity but the parents who are aware of the bad shoulder posture. On examination while standing the angular kyphosis is visible apical at T8/9 and is more apparent on forward spinal flexion (▶ Fig. 5.5b). In the erect position with shoulder retraction the deformity is less obvious but the kyphosis itself remains unchanged. In the prone position pressure over the apex of the kyphosis demonstrates it to be rigid, and this important clinical feature is what allowed Scheuermann himself to differentiate the condition from postural roundback deformity. 10 There is a compensatory increase in the normal lumbar lordosis. Straight-leg raising is commonly reduced by hamstring tightness which is considered to be secondary to the tilted pelvis position rather than a primary deforming mechanism as Lambrinudi postulated. 36 In all but the most severe degrees of Scheuermann’s kyphosis the neurological examination is normal. However, rarely the spinal cord can be bow-strung over the back of the kyphotic apex, in which case there may be objective neurological signs of impending paraplegia with clonus, hyperreflexia, and up-going plantar responses 37– 39 (▶ Fig. 5.6).
Fig. 5.5 (a) Severe angular thoracic kyphosis of type 1 Scheuermann’s disease is often better seen from the back where it can also be seen that there is a mild coexistent scoliosis. (b) On forward bending the type I Scheuermann’s deformity has an angular appearance over the apex.
Fig. 5.6 Lateral myelogram of a severe type I Scheuermann’s thoracic hyperkyphosis. There is no specific area of cord compression but the cord is thinned over several levels over the curve apex. This was sufficient to produce upper motor neurone symptoms.
Pain is an unusual presenting complaint with type I but the prevalence rate would appear to rise with the duration of the deformity, 40 and is not uncommon in the older patient. The site of the pain is generally at the apex of the deformity but can spread upward and downward. It would be imagined that pain would be worse the bigger the deformity but that correlation doesn’t seem to be particularly good. 24
It was conjectured that about one-third of patients with thoracic Scheuermann’s kyphosis had a mild, generally nonprogressive scoliosis. 1, 41 This is not so. A careful analysis of 50 cases of thoracic Scheuermann’s disease demonstrated that a mild lateral curvature of the spine was present in 85%. 34 In the minority this was present in the region of the kyphosis (▶ Fig. 5.5a) and was strictly an asymmetric kyphosis in the nature of a lateral curvature of the spine but with the spinous processes not rotated or rotated toward the curve convexity (discordant rotation). Where the majority of scolioses in association with Scheuermann’s kyphosis occur is four or five vertebral segments below the apex of the kyphosis (▶ Fig. 2.27). This is in the region of the compensatory hyperlordosis below which has rotated to the side in precisely the same way as in the production of the idiopathic scoliotic deformity, nicely confirming the lordotic etiology of idiopathic scoliosis. We found that the average curve size was 14 degrees and the spinous processes were rotated toward the curve concavity (concordant rotation). 34 However, progression was minimized by the older age of the adolescent with Scheuermann’s kyphosis and an axis of spinal rotation placed more anteriorly because of the presence of the kyphosis.
In type II (thoracolumbar or lumbar) kyphosis pain is the presenting complaint in the majority of patients. 7, 8 This tends to be in association with increased physical activity, and is therefore generally relieved by rest. There is a definitely increased prevalence rate of spondylolysis and spondylolisthesis in patients with type II Scheuermann’s kyphosis, 7 and the spondylolyses may indeed be multiple. It may be that symptomatic pain from these derangements is the mode of presentation in some. Because the vertebral wedging process occurs lower down the deformity is less marked, and is therefore not usually the cause of clinical presentation.
Indeed, on clinical examination mild degrees of type II Scheuermann’s kyphosis are difficult to detect. On forward flexion, however, the kyphotic area becomes more obvious and does not correct by manual pressure or spinal hyperextension. An associated lateral curvature of the spine is much less common in type II Scheuermann’s disease. 7
5.3 Radiological Diagnosis
The “normal” thoracic kyphosis extends from T3 to T9 or T10 and is caused by the shape of the thoracic vertebral bodies in the median sagittal plane (▶ Fig. 5.2). 42 Disk height does not contribute to the kyphosis. Therefore, the thoracic vertebral bodies are normally wedge-shaped, having a shorter anterior than posterior height. There is a wide range of the normal thoracic kyphosis and this range is generally regarded to be between 20 and 40 degrees. 3, 43 However, there have been several reports about so-called “normal” thoracic kyphosis either using empirical measurement points (e.g., T2 to T12), or using end-vertebrae, and these have been means of normal children of varying ages—27 degrees in one study, 43 33 in another, 44 and for instance 36 degrees in another. 45 Mixing boys and girls and different ages altogether to derive a mean is not particularly clever because as Willner and Oxborrow clearly showed there is considerable change in thoracic kyphosis with growth 28, 29 (▶ Fig. 3.12). This is really important in the pathogenesis of both idiopathic scoliosis and Scheuermann’s disease because when girls go through their peak adolescent growth velocity their thoracic kyphosis is at a minimum and if they overgrow then they can easily develop a lordosis. However, boys don’t mature until of the order of 3 years later when they are regaining a maximum thoracic kyphosis and of course if they overgrow then they are liable to develop Scheuermann’s disease. This accounts for the marked gender discrepancy with idiopathic scoliosis being much more prevalent in females and Scheuermann’s disease being much more prevalent in males. 31, 42, 44, 45 Sorensen felt that an overall thoracic kyphosis of more than 40 degrees should be one radiographic index of the presence of Scheuermann’s disease, 1 but this is clearly much less important than the shape of the apical vertebral bodies themselves. Furthermore, measurements of the overall thoracic kyphosis also include the compensatory lordoses above and below which have the effect of masking what is going on in the middle. Therefore, wedging of the apical three vertebrae by 5 degrees or more each is the best diagnostic guide (▶ Fig. 5.2). 1 In addition, vertebral wedging is the only constant radiological finding. Such end plate irregularity and Schmorl node formation as is present is situated anteriorly (▶ Fig. 5.3). However, the minimum of three consecutive apical vertebrae wedged by 5 degrees or more as the essential criterion seldom appears to be adhered to. Ponte’s phenomenal series of more than 3,000 cases of so-called Scheuermann’s disease 46 was only based upon the apical vertebra alone being wedged. While this might be better called Ponte’s disease, it certainly is not Scheuermann’s disease.
Interestingly, Knutson observed that appositional growth of the apical vertebral bodies in Scheuermann’s kyphosis occurs only anteriorly, such that any Schmorl’s node tends to move relatively toward the middle of the vertebrae in the median plane with time. 17 He also pointed out that there is an increased sagittal dimension of the vertebral bodies in Scheuermann’s kyphosis. Persistence of the anterior vascular groove in the thoracic vertebral bodies is an index of general immaturity but has been suggested as a means by which anterior collapse could occur. 47 As the phase of spinal maturity approaches so any Schmorl’s node formation or end plate irregularity tends to heal.
AP views of the thoracic spine in Scheuermann’s kyphosis reveal the associated lateral spine deformity in 85% of patients, with an apex four or five segments below the kyphosis (▶ Fig. 2.27). 34 Oblique or lateral projections of the lumbar spine reveal a defect in the pars interarticularis in one-third of type II cases, 7 and AP and lateral projections of the lumbar spine show an increased lumbar lordosis in type I cases with a tendency toward early so-called degenerative change. Similar projections of the Scheuermann’s kyphosis with increasing age also show changes with anterior bone spurring, the deformity occasionally progressing to apparent bony union anteriorly (▶ Fig. 5.7). 1, 2, 40 One has to be very careful to suggest that these appearances truly represent degenerative change. 40 What in fact is being seen are anterior bony spurs which have been erroneously called either osteophytes or syndesmophytes. Osteophytes are projections of new bone around effete hyaline cartilage cells in osteoarthritis while syndesmophytes are bony spurs projecting from the margins of syndesmoses. It should be remembered that intervertebral body joints are neither synovial joints nor syndesmoses (bone-fibrocartilage-bone). Intervertebral body joints are symphyses (bone-hyaline cartilage [cartilaginous end plate]–fibrocartilage [intervertebral disk]-hyaline cartilage-bone) 48 (▶ Fig. 5.8). However, what occurs with intervertebral joints on the concave side are marginal spurs of new bone formation simply because they are under compression as per Wolf’s law (see ▶ 3). Such appearances should not be misconstrued as showing evidence of degenerative joint disease.
Fig. 5.7 Lateral CT scan of a patient who underwent unsuccessful posterior instrumentation for a type I Scheuermann’s disease. Note the multilevel anterior spurring with possible fusion in some places.
Fig. 5.8 Intervertebral body joints are symphyses – bone (vertebral body) – highline cartilage (endplate cartilage) intervertebral disk (fibrocartilage) – highline cartilage – bone. (Reproduced with permission from Gray’s anatomy. Williams PL, Warwick R, Dyson M, Bannister LH (eds). Churchill Livingstone, Edinburgh 1989: Fig 4.1b).
The characteristic three level apical vertebral wedging helps to differentiate Scheuermann’s kyphosis from other forms of local kyphosis, in particular congenital kyphosis, neurofibromatosis kyphosis, the kyphosis associated with bullet-shaped vertebrae in mesenchymal disorders (see ▶ 10), and the kyphosis associated with infection and tumors. 49 However, radiologists not infrequently report a classical Scheuermann’s kyphosis as osteomyelitis of the spine, particularly if the end plate irregularities are pronounced and the hypertrophied anterior soft tissues give the appearance of a paravertebral shadow. In this situation the clinical findings of an otherwise healthy child help to differentiate the two. Where the differential diagnosis is occasionally in doubt occurs when there has been a vertical loading with flexion injury to the adolescent thoracic spine. In this case the vertebral bodies are wedged and very closely resemble that of Scheuermann’s disease. Perhaps the two need not be strictly differentiated apart from medicolegally, as Scheuermann’s kyphosis is presumably the result of decreased growth anteriorly leading to further compression and further reduction in growth and may indeed have very minor repetitive trauma as an adjunct.
Former treatment methods, which included bed rest or prolonged periods of recumbency on various frames, are both anachronistic and quite out of proportion to the nature of the underlying process, to which the term pathological is probably not semantically correct. As serious neurological complications are exceedingly rare, treatment concerns itself with body topography and the patient’s, and certainly to a much lesser extent the surgeon’s, view thereof. Type II (thoracolumbar or lumbar) kyphosis is not usually a problem of deformity but of pain and this can generally be alleviated by the traditional conservative measures for back pain, which include extension exercises plus hands-on physiotherapy and a supportive brace or corset for periods of severe pain. 7, 8 Surgical treatment is less often required for type II kyphosis. Type I kyphosis, however, can be a major problem of deformity with potential progression during growth.
5.4.1 Conservative Treatment
Thoracic Scheuermann’s kyphosis is perhaps the only spinal deformity (except early onset idiopathic scoliosis) that can be satisfactorily treated conservatively as it is a two-dimensional deformity behind the axis of spinal column rotation and therefore is not rotationally unstable and so does not buckle. Progressive change in vertebral shape producing the condition is caused by the anterior end of the growth plates being under compression, with a consequent reduction in anterior height. Accordingly, treatment that seeks to extend the kyphotic area reduces the amount of compression, 49 and a true physiological reconstitution of normal vertebral shape in the median plane is possible (▶ Fig. 5.9). Conservative treatment for Scheuermann’s kyphosis works by extension. 50 An extension cast or brace that obliterates the lumbar lordosis causes thoracic hyperextension above in order to produce the desired physiological effect, whereas with idiopathic scoliosis the hyperextension so produced will bring the lordosis back toward the sagittal plane with apparent improvement of the scoliosis on X-ray. However, all that will do is to increase the underlying lordosis, the undesirability of which has already been commented upon. 51 Although the Milwaukee brace has been the mainstay of conservative treatment for Scheuermann’s disease, 41 it is not necessary to have an orthosis with super-structure. Indeed the underarm Boston-type brace is better because of its superior powers of obliteration of the lumbar lordosis. Flexion of the spine above can be prevented by increasing the anterior height of the orthosis. We have found that wearing the brace for 12 hours at a time is quite sufficient and patients agree! According to Bradford 41 the indications for brace treatment are initial curves of less than 70 degrees with a Risser sign lower than grade 3 and minimal vertebral wedging in fewer than three vertebrae (is that really Scheuermann’s disease?). However, the critical initial curve magnitude should surely be the level at which the patient finds the deformity unacceptable. Obviously the more growth to go the better the prognosis for conservative treatment. However, minimal vertebral wedging in fewer than three vertebrae will be difficult to equate with the diagnostic criteria of Sorensen 1 and Scheuermann 10 who insist that there must be at least 5 degrees of wedging over three consecutive apical vertebrae.
Fig. 5.9 (a) Lateral radiograph of a 14-year-old with a moderate degree of type I Scheuermann’s disease. He was first treated with an extension cast. (b) Lateral radiograph in an underarm TLSO showing a correction of 50%. (c) Lateral radiograph at the age of 16 on removal of the brace a month previously. The correction has been sustained.