Introduction

Scheuermann’s kyphosis

Scheuermann’s kyphosis is a thoracic hyperkyphosis dueto wedged vertebrae that develop during adolescence. Itwas first described by Holger Werfel Scheuermann in1921 as different from postural kyphosis on the basis ofspine rigidity [36] (Fig. 1). The etiology is thought to bea development error in collagen aggregation, leading to

a disturbance of the enchondral ossification of the ver-tebral endplates [3]. This causes wedge-shaped defor-mation of the vertebrae and increase of kyphosis. Agenetic, hormonal and mechanical etiology has beendiscussed, and an autosomal dominant pattern ofinheritance is now accepted [2, 23]. The incidence ofScheuermann is unknown, probably ranging 1–8% ofthe general population. However, the ratio of boys/girlsis known to be between 2/1 and 7/1 [37, 39]. Clinically,the patients present with a rigid hyperkyphosis in the

Vincent Arlet

Dietrich SchlenzkaScheuermann’s kyphosis: surgicalmanagement

Received: 10 March 2004Accepted: 8 May 2004Published online: 14 April 2005� Springer-Verlag 2005

Abstract Indications for surgery inScheuermann disease are not wellcodified and remain rare, as thenatural history of the disease is inmost cases benign. In the immatureadolescent, conservative treatment,such as bracing or casting, can betried for moderate curves. For largercurves, or in the adult, conservativetreatment is usually not effective,and surgery can be considered. Suchindications are mostly cosmetic forlarge curves above 75�. Pain over thedeformity or in the low back mayrepresent another surgical indica-tion, especially in the adult group.The question of anterior release orstraight posterior fusion has becomemore of an actuality with the adventof powerful, third-generation stiffsegmental instrumentation. How-ever, the long-term results of amodern, posterior-only instrumen-tation fusion are not known. Con-cern about loss of correction, latepseudarthrosis or the need to re-

move instrumentation for infectedhardware or due to late pain at theoperative site must make us carefulabout choosing this method. Veryrigid and large curves still require ananterior release, either done in aconventional or mini-open fashion,or through video-assisted thoraco-scopic surgery. The extent of theposterior instrumentation has nowbeen better defined. One must fusethe whole Cobb angle withouthypercorrection and stop distally,above the first lordotic disc, to avoidsagittal decompensation. New ap-proaches such as short anterior fu-sion with bone-on-bone techniquesand pedicle substraction osteotomieshave not yet been reported in theliterature as having been used fortreating Scheuermann’s kyphosis.These should be considered experi-mental.

Keywords Scheuermann’skyphosis Æ Spinal fusion

Eur Spine J (2005) 14: 817–827DOI 10.1007/s00586-004-0750-0 REVIEW

V. Arlet (&)Department of Orthopaedic Surgery,University of Virginia, 400 Ray HuntDrive, Charlottesville, VA 22908, USAE-mail: Va3e@hscmail.mcc.virginia.edu

D. Schlenzka (&)ORTON Orthopedic Hospital InvalidFoundation, Tenholantie 10,00280 Helsinki, FinlandE-mail: dietrich.schlenzka@invalidisaatio.fiTel.: +358-9-47486624Fax: +358-9-2418415

mid-thoracic or lower-thoracic spine (Fig. 1a, b) with acompensatory hyperlordosis of the cervical and/orlumbar spine. Mild secondary scoliosis with minimal orno rotation at all may be present. Hamstring tightness iscommon. Neurological findings are very rare. Pain mayoccur in the region of the kyphosis or in the hyperlor-dotic areas above or below the main deformity.

Radiographically, Scheuermann is characterized byan increased kyphosis with compensatory lumbar hy-perlordosis (Fig. 1b). Classically, there is wedging ofthree consecutive vertebrae (more than 5�), with end-plate irregularities, loss of disc space height and Schmorlnodes (Fig. 1b) [39]. Aside from these typical featuresoutlined by Sorenson, some authors have considered anyvertebral wedging as a sign of Scheuermann disease [7, 8,9, 45]. On the MRI (Fig. 1b), which should be requestedprior to surgery, one will notice a premature dehydra-tion in the intervertebral discs, as well as an irregularityof ossification in the vertebral bodies.

Fig. 1 a Typical aspect of Scheuermann’s kyphosis with itsincreased rigid kyphosis; b standing-rays show the increasedkyphosis, and spot X-rays show irregularities of the endplates.The MRI shows wedged-shaped vertebrae and dehydrated discs

818

Natural history

Patients usually consult for pain or deformity (Fig. 2).During the early teenage years, symptoms are common

and would, in most cases, decrease in late adolescence orearly adulthood. Sorenson noted that 50% of his teen-agers had thoracic pain, decreasing to 25% after skeletalmaturity [39]. In contrast, Bradford reported an increase

Fig. 2 Sixteen-year-old boywith a 90º Scheuermann’s ky-phosis. a The complaint is 70%cosmetic and 30% back pain.The Cobb angle of the defor-mity measures 90� from T2 toT12. Note that disc L1–L2 isslightly lordotic. On the lateralshoot-through X-rays doneover a bolster, the deformitydecreases to 50�. Surgery con-sists of a posterior spine fusionfrom T2 to L1, avoiding over-correction, and a posterior iliaccrest bone graft; b at 1 yearfollow up, one notes the cor-rection of the deformity ismaintained at 48�. No furtherattempt of correction had beendone at the time of surgery toavoid junctional kyphosis, andno anterior surgery had beendone; patient is symptom free,and reports increased self es-teem

819

in the incidence of disabling back pain in adults. Painwas, therefore, a classic indication of surgery in this agegroup. Murray and Weinstein reported on the naturalhistory of Scheuermann disease over a 32-year period[25]. It appeared to be a benign condition in the majorityof cases. Scheuermann patients seemed to have moreback pain than did healthy controls, although the painrarely interfered with daily activities or professionalcareer. However, their series mostly involved mild-formScheuermann (average kyphotic deformity of 71�).Other series, such as Lowe or Bradford, report severethoracic pain in more severe kyphotic deformities(>75�) [6, 7, 8, 20, 21, 22]. Cardio-respiratory problemsmay occur in very severe deformities (kyphosis >100�[26]. But, surprisingly, most Scheuermann patients showan increase in their functional lung capacities.

Although surgical treatment of adolescent idiopathicscoliosis has gained acceptance and is better codified,indications for surgery in children and adolescents withScheuermann kyphosis are still not well defined, due tothe benign natural history of this condition. Beforeembarking on such a treatment, knowledge of the spine’snormal sagittal alignment must be known. Sagittal spinealignment during growth differs from alignment duringadulthood. At birth, the whole thoracolumbosacralspine is kyphotic. Then, as the child starts to stand, weobserve the apparition of lumbar lordosis. It is onlywhen the child becomes a teenager that the definitivesagittal curves are acquired. The average thoracic ky-phosis increases, during the growth of a child, from 20�in childhood to 25� in teenage years and 40� in adult-hood [13]. There is also a great variety of normal sagittalprofiles. The range of normal thoracic kyphosis is be-tween 10� and 50�, increasing in the elderly [4, 13, 15].The concept of pelvic incidence has recently beenintroduced by Duval Beaupere [18]. This is the anglemade by the line joining the middle of the sacrum to thehip joint, with the other line perpendicular to the sacralplateau. It is the only sagittal parameter that is fixed ineach individual. Pelvic incidence has been shown tocorrelate with the sagittal balance of the patient, and, asa rule of thumb, pelvic incidence should be approxi-mately 10� less than the lumbar lordosis. However, nostudy has focused on any possible relationship betweenincreased pelvic incidence and Scheuermann kyphosis.

Treatment of Scheuermann’s kyphosis

Non-operative treatment

Non-operative treatment is classically indicated duringthe growth period if thoracic kyphosis exceeds 40–45�and if radiological signs of the disease are present. Itincludes exercise, bracing and casting. Until now, exer-cise has not been shown to improve kyphosis. It does

have the advantage of increasing the patient’s awarenessof his own condition. Bracing and casting are of valueonly in patients with mobile kyphotic deformity andwith a sufficient amount of growth remaining. One yearof remaining growth is usually what is accepted asadequate to start an orthopedic treatment. Curves above75� will have a higher failure rate of brace treatment [35].Bracing and/or casting is known to become ineffectiveonce the patient’s Risser sign is 4 or 5. In cases of rigidScheuermann, cast treatment has been advocated bysome authors to be predictive of a successful outcome[30]. However, the largest series of conservative treat-ment remains the Bradford series with the Milwaukeebrace treatment [6]. With such treatment, compliantpatients had stabilization or a small improvement oftheir deformity. Patients with initial curves above 75�required surgery in 30% of the cases [35]. Because of thenon-compliant nature of the Milwaukee brace treat-ment, other braces such as the modified Boston or themodified Milwaukee have been tried and have also beenshown to be effective [14]. Brace treatment must usuallybe carried a minimum of 18 months to have an effect onvertebral wedging [24].

Operative treatment

A formal indication for surgery would be a neurologiccomplication appearing in the context of Scheuermann’s disease. Such complications, fortunately exceptional,would require neurologic decompression through ananterior thoracotomy or a posterolateral decompression[5, 9, 10].

Apart from these exceptional neurologic complica-tions there is no evidence-based criteria for an indicationof surgery. According to the literature, operative treat-ment should be considered in patients presenting with akyphotic deformity over 75�, significant pain that hasnot responded to conservative measures and/or respi-ratory problems due to severe kyphosis usually above100� [22, 32].

Because the natural history of Scheuermann’s ky-phosis is benign overall, surgery can only be done on acase-by-case basis, considering its potential complica-tions. The preoperative workup will focus on the pa-tient’s pain and/or cosmetic concerns, trying to identifythe motivation of the patient. Clinical examination willlook for stiff hamstrings as well as a popliteal angle ofless than 30� and any subtle neurological findings. Tighthamstrings have recently been emphasized as animportant possible cause of sagittal decompensation[17]. Radiographs will include long, standing scoliosisfilms and a hyper-reduction film that is best performedas a lateral shoot-through over a sand bag positioned atthe apex of the kyphosis (Fig. 2). Pulmonary functiontests are not necessary, as they are usually normal oreven increased in the case of Scheuermann [25]. An MRI

820

before surgery is recommended to rule out any excep-tional thoracic disc herniation, epidural cyst or possiblespinal stenosis. The literature has shown such excep-tional cases in various reports of neurologic complica-tions in Scheuermann [9, 10, 34, 41, 44, 47]. The MRIwill also assess the lumbar spine discs, as disc degener-ation of the lumbar spine may explain, in some cases, thepain rather than the deformity itself. One will also lookfor a possible spondylolysis frequently observed in thesepatients [27]. As the operation is essentially cosmetic,clinical photographs are mandatory (Fig. 2a).

Several questions must be answered prior to anysurgical decision: Does the curve need an anterior re-lease? What are the levels to be included in the spinefusion? Which correction technique should be used,and what is the amount of correction we should aimat?

Classic indications of anterior release are curves thatdo not correct over a sand bag to less than 50� [20, 21,22, 43]. However, with the advent of modern and stiffposterior instrumentation, along with spine-shorteningtechniques, these figures may not hold true, as the pos-sible correction achieved with third-generation segmen-tal instrumentation systems seems to make everythingpossible (Fig. 2).

Ponte described his posterior shortening procedurewith removal of the superior and inferior facets at theapex of the deformity to help reduce the curve [31](Fig. 3). Posterior fusion alone, however, exposes thepatient to loss of progression over time, as was firstpointed out by Bradford after simple Harrington pos-terior instrumentation [7]. With pedicle screw instru-mentation, the long-term result of surgery is not known.However, one will never be wrong with an anterior fu-sion, because the thoracoscopic procedures render theanterior release less morbid and may prevent loss ofcorrection in the future [1, 19, 26]. Literature has shownthat thoracoscopic anterior release is effective in Sche-uermann. Its definitive advantages over classic openthoracotomies are cosmesis and reduced morbidity.Thoracoscopic anterior release can also be performed inthe prone position, allowing for a simultaneous anteriorand posterior approach to the spine [1 ]. We recommendperforming an anterior release of 5–6 levels around theapex of the curve and grafting the intervertebral discspaces with pieces of autologous rib. This means that fora classic Scheuermann where the apex is at T8, onewould release from T5–T6 down to T10–T11. This hasthe advantage of working above the diaphragm. How-ever, in heavy patients and in Scheuermann, where thefusion extends into the lumbar spine at L2, anteriorstructural grafting may be warranted to prevent late lossof correction. Approach of the thoracolumbar junctiondown to L2 was a classic indication to take down thediaphragm. However, with sharp-tip angle retractorsinserted into the vertebral bodies and attached to ring

retractors, it is possible to get to the disc L1–L2 with justa buttonhole incision in the diaphragm.

The incidence of late operative-site pain ranges5–10% in patients with segmental posterior instrumen-tation, and removal of posterior instrumentation maylead to recurrence of the deformity, even with a solidposterior fusion [11]. Therefore, one may consider ananterior fusion of large and stiff Scheuermann’s diseasesafer for a long-lasting correction. In the adult patientswith Scheuermann’s disease and anterior bony bridging,the anterior release will, naturally, be necessary.

Considering these factors, using modern segmentalinstrumentation in 2005, should we or should we notperform an anterior release in the surgical candidatewith Scheuermann’s kyphosis? To summarize the liter-ature: initially in 1975, Bradford recommended the useof anterior fusion to prevent loss of correction [7]. Usinghard rods in 1986, Chopin stated that posterior fusionwas enough if the patient’s spine was immature [40]. ForSturm in 1993 posterior fusion was enough if one usedlarge Harrington rods. However, his series dealt withmoderate curves (average 71�) [42]. Papagelopoulos, in2001, recommended only posterior fusion for moderatecurves [29]. More recently, Hosman did not find anydifference in outcome between posterior-only surgeryand anterior and posterior correction, for a mean curveof 78� [16]. At present, no one knows if posterior surgeryonly will hold with an extended follow-up of severaldecades.

How many levels should be fused posteriorly? Theliterature is not very extensive on this very hot andcritical topic. Fusing too short naturally engenders therisk of junctional kyphosis either at the top or bottom ofthe instrumentation [20, 21, 22]. The current, generallyaccepted recommendations are to include the wholekyphotic Cobb angle and stop distally above the firstlordotic disc [20, 21, 22, 28]. In addition, it is ouropinion that the fusion should be balanced on either sideof the apex of the curve, and that the gravity linedropped down from C7 should pass by the middle of thelast fused vertebra. The minimum number of vertebraethat are required to be fused for a thoracic Scheuermannwith an apex at T8 is usually a spine fusion from T2down to L1 or even L2. In the case of a thoracolumbarScheuermann, one may have to extend the fusion distallyas low as L3.

Which techniques of posterior correction should beused? (Fig. 4) Ponte’s spine-shortening technique shouldbe well-known in cases of stiff curves, where correction isnot feasible [31]. Basically, one will remove the posteriorspinous processes at the apex, remove the ligament fla-vum, and perform complete superior and inferior facetresection (Fig. 4). These posterior resections have to bedone at the apex of the curve after the various implantshave been inserted. Although there are many reports onthe pedicle-substraction-osteotomy techniques employed

821

Fig. 3 Fifteen-year-old over-weight boy with a painful formof low thoracic Scheuermann’sdisease. a The deformity ismeasured at 77� by the Cobbangle method. However, on thesand-bag lateral shoot-through,the deformity decreases to only58� Cobb angle. Anterior re-lease and posterior instrumen-tation from T3 to L2 is carriedout in the same setting; b clini-cal aspect 2 years after surgery.The patient reports increasedself esteem, and has no morepain

822

for kyphotic deformities such as flat-back syndrome,congenital kyphosis or ankylosing spondylitis, noavailable report was found in the literature onScheuermann. Such substraction osteotomies wouldhave to be done at the level of the thoracic spine andwould carry an inherent neurologic risk. Likewise,approaches such as short anterior bone-on-bonetechniques have not been published in peer reviewjournals on Scheuermann kyphosis.

The spinal instrumentation should consist of a strongfoundation made of pedicle screws, supplemented withlaminar hooks if necessary in the lumbar spine and inthe upper thoracic spine. The top of the construct shouldbe solid, with multiple anchorage points made of eitherpedicle screws or a different combination of laminarhooks, pedicle hooks and screws(Figs. 2, 3, 4). At thetop, if using hooks, one should employ the claws prin-ciples. There is basically no need to instrument the apexof the deformity. Hooks alone should not be used at theapex of the kyphosis or at the distal end of the construct,as they can be potentially dangerous. They invade thespinal canal and will result in a medial positioning of therods. During the correction maneuvers, such medialpositioning of the rods may press on the thin and frag-ilized lamina at the apex, with possible fracture of thelaminae and neurologic complications.

Four different correction techniques can be used:cantilevering, compression and double-rod techniques(four rods) (Fig. 4); and in situ bending may be used tofine tune the final correction or to undo what appears tobe too much correction. The literature does not seem tofavor one technique over another. Compression wasclassic and the only method available when using Har-rington rod instrumentation [7, 29]. At present, mostsurgeons would prefer cantilevering the rods from thetop of the spine down to the lower thoracic or lumbarpedicle screws. At any rate, comparing compression withcantilevering, one may wonder if the end result wouldnot be the same, because cantilevering is necessarilyassociated with segmental compression. Whatever tech-nique is used, one should use stiff and large diameterrods, since small-diameter or regular titanium rods aretoo flexible, with possible recurrence of the deformity.Posterior bone grafting from the iliac crest should beperformed after thorough spinal decortication of theposterior elements. The goal is to achieve a thick fusionmass that will resist the test of time.

How much correction should we aim for, and isstraighter better? The Scoliosis Research Society (SRS)has defined normal thoracic kyphosis as between 20�and 40�. Bernhart and Bridwell have stated that ky-phosis up to 50� is perfectly normal. What is not knownis the amount of thoracic kyphosis that an individualshould have, depending on his pelvic incidence. At anyrate, most current authors seem to favor normalizingthoracic kyphosis that is above 50�, down into the

40–50� range, in order to avoid postoperative imbalance[16, 17, 22]. Therefore, straighter is not better in Sche-uermann’s kyphosis.

With the advent of strong segmental posteriorinstrumentation, indications for postoperative bracinghave relaxed. During the Harrington era, postoperativebracing was advised for a period of several months.Today, it is often not prescribed, as in the case ofpostoperative scoliosis patients. Forces are, however,totally different, and with an active teenager, willing toresume physical activities as soon as possible regardlessof the doctor’s advice, bracing for the first 3 months is areasonable choice, as recommended by Hosmans [16].

Surgery on Scheuermann’s disease is not benign andcomplications can arise.

Neurologic complications can arise during the cor-rection maneuvers, because of a rare intracanalarproblem, or from a surgical-technique problem. Theexact rate of neurologic complications is not known inScheuermann surgery but is probably higher than foridiopathic scoliosis. Other complications, such as death,infections and gastrointestinal obstruction [29], arecommon to any corrective procedure for spinal defor-mities.

Junctional kyphosis would occur in 20–30% of thecases, according to Lowe [22] (Fig. 5). Some series re-port an incidence even higher than 50% [33]. In thepathogenesis of junctional kyphosis, it was believed,initially, that the passage of sublaminar wires, as in theLuque technique, would result in disruption of thesuperior and infraspinatus ligament with partial removalof the flavum [12]. Persistence of this complication withmodern instrumentation that respects such structuresprompted other authors to propose other explanations.Lowe gives us three explanations: (1) fusion that is tooshort, i.e., stopping short of the first lordotic disc at thebottom, with a resultant distal junctional kyphosis; (2)fusion that is too short proximally and that does notinclude the whole kyphosis on the top, with a resultantproximal-junctional kyphosis and a goose-neck appear-ance; and (3) hypercorrection seems to be a factor, andone should not correct the kyphosis to more than 50%of its initial value. Lastly, but rarely mentioned, it is ourbelief that insufficient control of the distal verte-bra—with only hooks, or pedicle screws that are tooshort, or lack of anterior column support—may allowthe last fused segment into recurrence of some kyphosisand overall junctional kyphosis of the construct (seeFig. 5). We, therefore, recommend inserting large-diameter pedicle screws in the last distal-instrumentedvertebrae and making them as long as possible to pre-vent segmental kyphosis. Natural, the use of polyaxialscrews at the end of the construct also exposes the pa-tient to the risk of junction kyphosis as they do notcontrol the distal vertebra in the same way as monoaxialscrews. Distal and proximal pedicle screws can also be

823

824

protected by laminar hooks to help prevent their pull-out. However, we do not recommend the insertion oftwo laminar hooks at the same level, for fear of an over-large invasion of the spinal canal. In heavy patients, oneshould not hesitate to add an anterior column supportdistally. Hosmans recently published a paper on theimportance of tight hamstrings in the surgical correction[17]. Patients with tight hamstrings were found to have asignificant risk of postoperative imbalance. Further-more, the limitations in the lumbar and pelvic range ofmotion were predicted by tight hamstrings. However,Hosmans did not think that hamstring stretching was ofany value in preventing imbalance of the spine. Post-operatively bracing a patient with Scheuermann’s dis-ease may help prevention of distal junctional kyphosis.

Results of surgery on Scheuermann’s kyphosis can beanalyzed according to the two major indications forwhich the surgery was carried out: pain and cosmeticdeformity. As far as pain is concerned, all series reportan improvement in the amount of back pain, ranging66—90% [6, 7, 16, 20, 21, 22, 42]. Hosmans showed amarked improvement in the Oswestry Disability Score,from an average of 23 preoperatively down to 6.6 atfollow-up. However, neck pain did not seem to improveafter surgery. Interestingly, there did not seem to be anyrelationship between amount of correction and amountof residual back pain. As far as satisfaction is concerned,most series report a very high satisfaction rate, up to96% [6, 7, 16, 20, 21, 22, 42]. Since no cosmetic scale isavailable for assessment of Scheuermann’s kyphosis,and since the SRS instrument on Scheuermann’s has notbeen published, it is necessary to judge cosmetic cor-

Fig. 4 Techniques of posterior corrections. a Posterior shorteningprocedure according to Ponte. The superior and inferior facets andthe ligamentum flavum are totally at the apex of the deformity;b example of implant strategy for correction of Scheuermann’skyphosis. The top is instrumented according to claws principles.One will use supra-laminar hooks and pedicle hooks arranged inclaw configuration (these may be replaced with pedicle screws if oneis used to high thoracic-pedicle screws). No implants are inserted atthe apex of the deformity. The bottom foundation is made ofseveral pedicle screws, possibly supplemented with laminar hooks;c principles of correction: one will use the compression technique(not shown), the cantilever techniques or the four-rods technique;d example of a four-rod technique used to correct the deformity

b

Fig. 5 Junctional kyphosis. a Scheuermann’s kyphosis in a 15-year-old boy. The deformity is measured at 90�. Note that disc L1—L2is parallel and not lordotic. The deformity is very supple, correctingto 35�, as seen on the lateral shoot-through film; b postoperative X-rays of a posterior-fusion-only from T3 to L1. The correction seemsvery satisfactory. Patient does not wear a brace; c 2 monthspostoperatively, there has been a distal junctional kyphosis,fortunately asymptomatic. Possible explanations: the fusion is tooshort, as the disc L1–L2 was only parallel and not lordotic; thepatient did not wear a brace; the pedicle screws in L1 are too shortand backed out somehow

c

825

rection on plain X-rays, representing an extrapolation ofthe cosmetic results. The correction rate given in thevarious surgical series shows a correction from anaverage 70–75� initial Cobb angle to 40–45� at last fol-low-up, i.e., a correction rate of 40–50%. Bradford ini-tially reported a loss of correction after Harringtonposterior instrumentation. This modified the indicationsof surgery and, classically, curves of more than 75� witha hyperextension radiograph of less than 50� shouldhave an anterior release [7, 8]. This has not been dem-onstrated as valid with third-generation spinal instru-mentation.

Hosmans reported a correction loss of only 1.4� afterposterior-only instrumentation [16]. Cosmetic correctionof Scheuermann’s kyphosis should be assessed accordingto patient satisfaction and improvement of perceivedself-image, as well as by independent judgment of clini-cal photographs, before and after surgery, by nonmedical judges. Literature definitively lacks such infor-mation. We hope that future results of Scheuermanncorrective surgery will be based not only on the purepercentage of Cobb angle correction but also on out-come questionnaires such as the scoliosis SRS 24; thepatient’s sagittal balance; and the residual lumbar andcervical lordosis on either side of the fusion, because

once a segment of the spine becomes fused, one shouldalso take interest in the non-fused segment of the spine.

Conclusions

Indications for surgery in patients with Scheuermann’skyphosis should be assessed on a case-by-case basis.Spinal cord compression is exceptional in Scheuer-mann’s kyphosis but occurs in both teenagers andadults. They represent a formal indication for surgery.Other indications are based on the amount of cosmeticdeformity and its acceptance by the patient. Most au-thors agree that curves above 70–75� are an indicationfor surgery. Pain, especially in an adult, may representanother reason for surgery. New technologies such asvideo-assisted thoracoscopic anterior release and fusion,mini-open approaches, and stiff third-generation seg-mental instrumentation, have rendered the need for aformal open anterior release very rare. Proper selectionof levels and avoidance of over correction willlimit complications such as junctional kyphosis. Mostpatients will have satisfactory results after surgery.Particular attention to details will avoid most compli-cations.

References

1. Arlet V (2000) Anterior thoracoscopicspine release in deformity surgery: ameta-analysis and review. Eur Spine J 9[Suppl 1]: S17–23

2. Ascani E, LaRosa G (1994) Scheuer-mann’ kyphosis. In: Weinstein SL (ed)The pediatric spine. Principles andpractice. Raven Press, New York, pp557–585

3. Aufdenmaur E (1981) Juvenile kyphosis(Scheuermann’s disease) Radiography,histology and pathogenesis. Clin Ort-hop 154:166–174

4. Bernhardt M, Bridwell KH (1989) Seg-mental analysis of the sagittal planealignment of the normal thoracic andlumbar spines and thoracolumbarjunction. Spine 14:717–721

5. Bhojraj SY, Dandawate AV (1994)Progressive cord compression second-ary to thoracic disc lesions in Scheuer-mann’s kyphosis managed byposterolateral decompression, inter-body fusion and pedicular fixation. Anew approach to management of a rareclinical entity. Eur Spine J 3:66–69

6. Bradford DS, Moe JH, Montalvo FJ,Winter RB (1974) Scheuermann’s ky-phosis and roundback deformity. Re-sults of Milwaukee brace treatment. JBone Joint Surg Am 56:740–758

7. Bradford DS, Moe JH, Montalvo FJ,Winter RB (1975) Scheuermann’s ky-phosis. Results of surgical treatment byposterior spine arthrodesis in twenty-two patients. J Bone Joint Surg Am57:439–448

8. Bradford DS, Ahmed KB, Moe JH,Winter RB, Lonstein JE (1980) The sur-gical management of patients withScheuermann’s disease: a review oftwenty-four cases managed by combinedanterior and posterior spine fusion.J Bone Joint Surg Am 62:705–712

9. Bruns I, Heise U (1994) Spastic para-paresis in Scheuermann disease: a casereport. Z Orthop Ihre Grenzgeb132:390–393

10. Chiu KY, Luk KD (1995) Cord com-pression caused by multiple disc herni-ations and intraspinal cyst inScheuermann’s disease. Spine 20:1075–1079

11. Cook S, Asher M, Lai SM, Shobe J(2000) Reoperation after primary pos-terior instrumentation and fusion foridiopathic scoliosis. Toward defininglate operative site pain of unknowncause. Spine 25(4):2516–2517

12. Coscia MF, Bradford DS, Ogilvie JW(1988) Scheuermann’s kyphosis resultsin 19 cases treated by spinal arthrodesisand Luque instrumentation. OrthopTrans 12:255

13. Fon GT, Pitt MJ, Thies AC Jr (1980)Thoracic kyphosis: range in normalsubjects. AJR Am J Roentgenol134:979–983

14. Gutowski WT, Renshaw TS (1988)Orthotic results in adolescent kyphosis.Spine 13:485-489

15. Hammerberg EM, Wood KB (2003)Sagittal profile of the elderly. J SpinalDisord Tech 2003 16:44–50

16. Hosman AJ, Langeloo DD, de KleuverM, Anderson PG, Veth RP, Slot GH(2002) Analysis of the sagittal planeafter surgical management for Scheuer-mann’s disease: a view on over correc-tion and the use of an anterior release.Spine 2:167–175

17. Hosman AJ, de Kleuver M, AndersonPG, van Limbeek J, Langeloo DD,Veth RP, Slot GH (2003) Scheuermannkyphosis: the importance of tight ham-strings in the surgical correction. Spine19:2252–2259

826

18. Legaye J, Duval-Beaupere G, HecquetJ, Marty C (1998) Pelvic incidence: afundamental pelvic parameter for three-dimensional regulation of spinal sagittalcurves. Eur Spine J 7:99–103

19. Liljenqvist U, Steinbeck J, Niemeyer T,Halm H, Winkelmann W (1999) Tho-racoscopic interventions in deformitiesof the thoracic spine. Z Orthop IhreGrenzgeb. 137:496–502

20. Lowe TG (1987) Double L-rod instru-mentation in the treatment of severekyphosis secondary to Scheuermann’sdisease. Spine 12: 336–341

21. Lowe TG (1999) Scheuermann’s dis-ease. Orthop Clin North Am 30:475–487

22. Lowe TG, Kasten MD (1994) Ananalysis of sagittal curves and balanceafter Cotrel-Dubousset instrumentationfor kyphosis secondary to Scheuer-mann’s disease. A review of 32 patients.Spine 19:1680–1685

23. McKenzie L, Sillence D (1992) FamilialScheuermann disease: a genetic andlinkage study. J Med Genet 29:41–45

24. Montgomery SP, Erwin WE (1981)Scheuermann’s kyphosis—long-termresults of Milwaukee braces treatment.Spine 6:5–8

25. Murray PM, Weinstein SL, Spratt KF(1993) The natural history and long-term follow-up of Scheuermann ky-phosis. J Bone Joint Surg Am 75 (2):236–248

26. Newton PO, Shea KG, Granlund KF(2000) Defining the pediatric spinalthoracoscopy learning curve: sixty-fiveconsecutive cases. Spine 25:1028–1035

27. Ogilvie JW, Sherman J (1987) Spond-ylolysis in Scheuermann’s disease. Spine12:251–253

28. Otsuka NY, Hall JE, Mah JY (1990)Posterior fusion for Scheuermann’skyphosis. Clin Orthop 251:134–139

29. Papagelopoulos PJ, Klassen RA,Peterson HA, Dekutoski MB (2001)Surgical treatment of Scheuermann’sdisease with segmental compressioninstrumentation. Clin Orthop 386:139–149

30. Platero D, Luna JD, Pedraza V (1997)Juvenile kyphosis: effects of differentvariables on conservative treatmentoutcome. Acta Orthop Belg 63:194–201

31. Ponte A, Siccardi GL, Ligure P (1995)Posterior shortening procedure by seg-mental closing wedge osteotomies (ab-stract). J Pediatr Orthop 15:404

32. Poolman RW, Been HD, Ubags LH(2002) Clinical outcome and radio-graphic results after operative treatmentof Scheuermann’s disease. Eur Spine J11:561–569

33. Reinhardt P, Basset GS (1990) Shortsegmental kyphosis following fusion forScheuermann’s disease. J Spinal Disord3(2):162–168

34. Ryan MD, Taylor TK (1982) Acutespinal cord compression in Scheuer-mann’s disease. J Bone Joint Surg Br64:409–412

35. Sachs B, Bradford D, Winter R,Lonstein J, Moe J, Willson S (1987)Scheuermann kyphosis. Follow-up ofMilwaukee-brace treatment. J BoneJoint Surg Am 69:50–57

36. Scheuermann HW (1921) KyphosisDorsalis Juvenilis. Orthop Chir 41:305

37. Scoles PV, Latimer BM, Di GiovanniBF, Vargo E, Bauza S, Jellema LM(1991) Vertebral alterations in Scheuer-mann’s kyphosis. Spine 16:509–515

38. Soo CL, Noble PC, Esses SI (2002)Scheuermann kyphosis: long-term fol-low-up. Spine J 2:49–56

39. Sorenson KH (1964) Scheuermann’skyphosis: clinical appearance, radiog-raphy, etiology and prognosis. Mun-sksgaard, Copenhagen

40. Speck GR, Chopin DC (1986) Thesurgical treatment of Scheuermann’skyphosis. J Bone Joint Surg Br 68:189–193

41. Stambough JL, VanLoveren HR,Cheeks ML (1992) Spinal cord com-pression in Scheuermann’s kyphosis:case report. Neurosurgery 30(1):127–130

42. Sturm PF, Dobson JC, Armstrong GW(1993) The surgical management ofScheuermann’s disease. Spine 18:685–691

43. Tribus CB (1998) Scheuermann’s ky-phosis in adolescents and adults: diag-nosis and management. J Am AcadOrthop Surg 6:36–43

44. Tribus CB (2001) Transient paraparesis:a complication of the surgical manage-ment of Scheuermann’s kyphosis sec-ondary to thoracic stenosis. Spine26:1086–1089

45. Wenger DR, Frick SL (1999) Scheuer-mann kyphosis. Spine 24:2630–2639

46. Wood KB, Garvey TA, Gundry C,Heithoff KB (1995) Magnetic resonanceimaging of the thoracic spine. Evalua-tion of asymptomatic individuals.J Bone Joint Surg Am 11:1631–1638

47. Yablon JS, Kasdon DL, Levine H(1988) Thoracic cord compression inScheuermann’s disease. Spine 8:896–898

827