Methods

A comparative retrospective study in patients who had surgical correction of early onset scoliosis was performedin a single institution. The study was approved by Sytenko Institute of Spine and Joint Pathology Review Board and by Ethical Committee. Twenty-six skeletal immature consecutive patients who underwent anterior convex growth arrest with posterior growing spinal instrumentation are included in this study. Inclusion criteria were: patients with early onset scoliosis, Risser 0, open triradiate cartilage. In all patients spinal derotation has been performed. The first group (13 pts) underwent hybrid instrumentation (thoracic spine hooks and lumbar spine screws) and the second group (13pts) treated with screws only.

All patients in groups had complete radiographic data, including preoperative standing anteposterior, lateral and bending radiographs. All radiographs during follow-up were available. Preoperative, postoperative and follow-up films have been analyzed. Computer tomography scans were used for more precise measurements. Thoracic kyphosis (T5-T12 angle), postoperative proximal segmental kyphosis and proximal junctional kyphosis were studied preoperatively, postoperatively and after 4 years. Proximal junctional kyphosis has been defined as sagittal Cobb angle between the upper end instrumented vertebra and one vertebra cephalad measuring over 10^° [8]. Proximal segmental kyphosis is defined between segment of the spine at the upper end of growing construct and spinal segment proximal to the construct [6]. Spinal growth was measured by length determination of the rod’s upper segment after surgical treatment and during follow-up. The operative data that were assessed and compared for each group included number of the instrumented vertebras and upper level of the spinal instrumentation. All assessed parameters in both groups were analyzed and compared to each other. Statistical comparison of data was performed using two means and standard deviations. Level of statistical significance was set at p<0.05.

Indications for surgery

Surgery was recommended for the skeletal immature patients with progressive scoliosis with a Cobb angle over 40^°, Risser 0 and open triradiate cartilage. In all patients anterior convex growth arrest and posterior growing spinal instrumentation were performed.

Selection of the fusion levels

Fusion levels were defined using anterior, lateral and supine bending films. In sagittal plane all pathologic curves were included into instrumentation zone. In the transverse plane, the upper instrumented vertebra was defined as a rotational neutral vertebra on standing anteposterior film. Lower instrumented vertebra was defined using central sacral line on supine bending film.

Surgical technique

Anterior convex growth arrest was performed via open thoracotomy approach. Incision was made over the rib corresponding to the end vertebra of the curve. After subperiostal rib resection parietal pleura was incised and reflected of the spine one level above and below the involved area. Segmental vessels were legated and divided. After apical disc identification partial discectomy and convex epiphyseodesis were performed. Intervertebral bone grafting with autorib pieces was performed. Chest tube was placed and thoracic cage was closed. Depending on patient’s condition posterior spinal instrumentation was performed the same day or after 3-5 days breaks.

Posterior approach was performed through midline incision and main curve and secondary curves were subperiostaly exposed. Spinal anchoring was performed using conventional surgical technique. Pedicular hooks with additional internal screw for the secure fixation (1^st group) or pedicular screws (2^nd group) were put into the thoracic spine. Lumbar spine instrumented with pedicular screws. Rods were inserted into bone anchors and seated; deformity correction was performed by 90^° concave rod derotation. Two medial screws and crosslinks were locked if predicted growth was more than 4 cm and other anchors were loose allowing the rods to slide during the patient’s growth. Bottom screw tightening was used if anticipated growth was less than 4 cm. The brace was recommended full time after surgery and thereafter only when the patient was out of bed.

After growth end (closed triradiate cartilage, Risser 4, menarche) posterior spinal fusion was performed. The surgery included instrumented spine exploration, posterior fusion with autograft obtained from rib resected after convex thoracoplasty and construct transformation into rigid fixation by nut tightening.

Results

The first group (hybrid fixation) included thirteen patients with the mean age of 9,1 years at the moment of index surgery. Eleven (81%) patients had main thoracic, one had double curve (9,5%) and one had thoracolumbar deformity (9,5%). Results of radiographic measurements are shown in Table 1.

Preoperative mean thoracic kyphosis angle was 10.8^°±5.33. Mean number of the instrumented vertebras in the first group was 12. In seven patients (53%) T4 was the upper instrumented vertebra, in four patients (30.7%) T5, T3 and T6 were the upper instrumented vertebras in other two patients (29,3%). Mean growth of the instrumented spine during follow-up was 7.1 mm per year. Postoperative spinal kyphosis was 26.2^°±10.83 on the average, mean proximal segmental kyphosis was 8.9^°±4.7, proximal junctional kyphosis angle was 5.6^°±3.32.

After 4 years thoracic kyphosis was 29.8^°±11.96, mean proximal segmental kyphosis and proximal junctional kyphosis angles were 11.4^°±8.9 and 7.6^°±3.01. Two patients with upper hooks pull-out had cephalad instrumentation extension during final fusion (Figure 1) and in 4 patients hook displacement without loosening was observed (Figure 2). Final fusion was performed in all the patients out of this group.

The second group of patients (screw fixation) included thirteen female patients with the mean age of 8.2 years. Curve patterns distribution was the same as in the first group: eleven main thoracic, one double deformity and one thoracolumbar deformity. Before the index meanthoracic kyphosis was 40.6^°±18.4 (Figure 1). Number of the levels fixed was 11.9. Six patients (46%) were instrumented with the upper instrumented vertebra T4, four patients (30%) had upper instrumentation level at T5. Upper instrumented levels at T2, T3 and T7 were instrumented in three more patients. Mean growth of the instrumented spine was 7.2 mm per year. Postoperative kyphosis in the second group was 30.8^°±6,9 on the average. Proximal segmental kyphosis after the surgery was 12.8^°±6.57 and proximal junctional kyphosis was 9.5^°±5.32. Four years after deformity correction mean thoracic kyphosis was 37.7^°±10.1, proximal segmental kyphosis was 18.7^°±4.1 and proximal junctional kyphosis was 10.7^°±5.7. In 4 patients out of the screw group proximal junctional kyphosis greater than 10^° was observed. In one patient with thoracolumbar curve and proximal junctional kyphosis extended fusion two levels above were performed (Figure 3). In other three patients junctional hyperkyphosis didn’t impact final outcome. Six patients of the 2^nd group were finally fused.

Discussion

Proximal junctional kyphosis is a recently recognized situation after posterior spinal instrumentation in idiopathic scoliosis. Kim Y.J. analyzed the long-term proximal junctional change in adolescent idiopathic scoliosis 5 years after surgery. The study included 193 consecutive adolescent idiopathic scoliosis patients after posterior instrumented fusion. The incidence of proximal junctional kyphosis was 26% (50 patients) after 7.3 years postoperatively. Thoracoplasty, hybrid instrumentation and a preoperative larger sagittal thoracic Cobb angle were found as the risk factors for developing of proximal junctional kyphosis [9].

In another study Wang J et al. found proximal junctional kyphosis in 35 out of 123 patients with an overall incidence of 28%. Authors of the paper concluded that inducing factors are: intraoperative decrease in thoracic kyphosis greater than 10^°, thoracoplasty, the use of a pedicle screw at the top vertebra, autogenous bone graft and fusion below L2 [10].

Helgeson M.D. compared the incidence and risk factors for proximal junctional kyphosis in adolescent idiopathic scoliosis following posterior spinal fusion using different fixation strategies (hook, pedicle screw, or hybrid anchoring). Preoperative and postoperative radiographs of 283 patients were reviewed retrospectively. The incidence of proximal junctional kyphosis was 0% in hook group, 2.3% in hybrid group, 8.1% in screw group and 5.6% in screws with hooks only at the top level group. It was suggested that hook fixation at the upper-instrumented vertebrae could decrease the likelihood of junctional problems development [11].

Distraction methods such as Harrington instrumentation, dual growing rod technique [12] and VEPTR are recognized as the most effective procedures for the early onset spinal deformities [13]. Lengthening from 6 to 12 months promotes continued growth of the spine and delay of a formal spinal fusion closer to skeletal maturity. Restoration of the sagittal spinal balance is an important issue in young children who require growing instrumentation [14]. Proximal junctional kyphosis is the potential complication of these surgeries reported in the literature. Higher risks of proximal junctional kyphosis were found for older hyperkyphotic children and in patients with positive postoperative sagittal balance [15].

D. Skaggs presented a retrospective review of 32 cases with growing rods for early onset scoliosis. Average age at the initial surgery was 4 years and follow-up was 24-88 months. Eighteen patients (56%) out of this group had proximal junctional kyphosis and 8 patients out of them had upper anchor failure with required revision in 7 patients. Among 4 patients with proximal junctional kyphosis who underwent spinal fusion, 3 patients had fusion and instrumentation two levels cephalad to implant. Junctional kyphosis was more common in children with dual rods (62%) than with single rod (38%) and in spine-to-spine constructs (59%) compared to upper hooks on ribs constructs [5].

Theoretically, utilization of VEPTR should minimize the risk of proximal junctional kyphosis since the upper anchor attached to the rib. Li Y. in his study assessed proximal spinal kyphosis in 68 patients who had VEPTR surgery. Five patients (7%) were found to have proximal segmental kyphosis. Before the intervention to progressive proximal segmental kyphosis average angle was 66^°, mean increase in proximal segmental kyphosis was 25^°. In all of the patients proximal segmental kyphosis developed within the first years after initial surgery. Authors concluded that risk factor for development of such complication in VEPTR patients is preexisting thoracic kyphosis, and the deformity can progress to severe and require revision surgery [6].

Reinker K. et al. retrospectively reviewed 14 patients who had VEPTR surgery of early-onset kyphoscoliosis. In spite of scoliotic deformity stabilization kyphosis increased by 22^° at the last follow-up. Proximal cradle pull-out was periodically observed. Distal anchors placed too proximally had inadequate lever arms to control kyphosis. It was claimed in this study that attachment of the upper attachment point to the second rib bilaterally and distal extension of hybrid constructs to the pelvis would avoid kyphosis progression [16].

Polysegmental growing spinal instrumentation with apical attachment is another option for the early onset scoliosis treatment. Debnath U. and associates report long term results of growing Luque-trolley construct in growing children. The study cohort consisted of 37 patients: 7 patients had Luque-trolley instrumentation alone and 30 patients had convex fusion and posterior Luque-trolley instrumentation. Complications regarding development of junctional kyphosis were evaluated. Junctional kyphosis was observed in 8 cases: proximal-3 patients, distal-2 patients, apical-3 patients which was corrected at the time of definitive spinal fusion. Authors of this study concluded that patients with high concave rib-spinal angle and upper end vertebra tilt at initial presentation should be closely monitored for deterioration and junctional kyphosis. Spinal growth that exceeds the capacity of Luque-trolley to elongate leads to apical junctional kyphosis which needs regular surveillance [17].

In our study comparative analysis of hybrid versus screw growing polysegmental instrumentation was performed to evaluate the rate of proximal junctional kyphosis in skeletal immature children. There were similar curve etiology in both groups with a difference in the age of 9.1 years in the 1^stgroup and 8,2 years in the 2^nd group. We found an important difference in thoracic kyphosis parameters: 10.8^° (1^st group) versus 40.6^° (2^nd group). Polysegmental anchor placement and rod derotation maneuver result in normal sagittal contour restoration in all the patients, and postoperative values of thoracic kyphosis were normal. We observed increasing of thoracic kyphosis, proximal junctional kyphosis and proximal segmental kyphosis in both groups, but in hybrid patients mean values of these parameters were less than in screw group. We found proximal junctional kyphosis over 10^° only in 1 patient in the 1^st group and in 4 patients in the 2^nd group. Two patients with hook fixation at the thoracic spine had pull-out and 4 patients had hooks displacement after 4 years follow-up. Thereby thoracic hook fixation provides better control of the sagittal spinal shape as patient is growing but screw fixation avoids displacement of the anchor at the upper instrumented level. In our opinion, the best ways to avoid development of proximal junctional kyphosis in children after growing spinal instrumentation is utilization of screw fixation up to T2 and rod derotation for the deformity correction. As far as during spinal growth sagittal shape of the spine remains irreversible after the surgery, sagittal alignment should be restored according to the normative data for the adulthood (Figure 4).

Conclusion

In spite of preexisting hypo- or hyperkyphosis, utilization of the growing polysegmental construct and performing spinal derotation restore normal sagittal profile in growing EOS patients.

Growth sparing surgery utilizing hook fixation in the thoracic spine decreases likelihood of proximal segmental kyphosis and proximal junctional kyphosis in skeletal immature EOS patients, but increases the rate of upper anchor displacement.

Summary

The purpose of this study is to define if polysegmental growing instrumentation and derotation for the spinal correction result in proximal segmental kyphosis and proximal junctional kyphosis. Twenty-six consecutive patients divided into two groups, who had undergone anterior convex growth arrest with posterior growing spinal instrumentation, are included in this study. We found proximal junctional kyphosis over 10^° only in 1 patient in the 1^st group and in 4 patients in the 2^nd group. In spite of preexisting hypo- or hyperkyphosis, utilization of the growing polysegmental construct that we propose and performing spinal derotation restore normal sagittal profile in growing patients. Growth sparing surgery utilizing hook fixation of the thoracic spine decreases likelihood of proximal segmental kyphosis and proximal junctional kyphosis in skeletal immature EOS patients, but increases the rate of upper anchor displacement. Twenty-six consecutive patients who had undergone anterior convex growth arrest with posterior growing spinal instrumentation are included in this study. Growth sparing surgery utilizing hook fixation of the thoracic spine decreases likelihood of proximal segmental kyphosis and proximal junctional kyphosis in skeletal immature EOS patients, but increases the rate of upper anchor displacement.

Figure 1: (A,E) A 9-year-old female with an 91^° with juvenile idiopathic scoliosis and and thoracic kyphosis 52^°, Risser 1. (B, F) Patient underwent anterior convex growth arrest and posterior spinal instrumentation with growing hybrid system. Upper instrumented vertebra is T4. (C,G) After 4 years she had hook pull-out radiographs show hook pull-out, proximal segmental kyphosis was 27^°, proximal junctional kyphosis was 18^°. (D,H). Final fusion with instrumentation extension performed, concave rod was removed because of subcutaneous prominence.

Figure 2: (A) Female patient with juvenile idiopathic scoliosis 90^° diagnosed in the age of 4 years. (B) hybrid growing spinal instrumentation was performed, upper instrumented vertebra was T4. (D) spinal derogation increased thoracic kyphosis to 35^° postoperatively. (E,F) 4 years post she complained of upper thoracic pain. On radiographs rotational hook displacement without loosening was diagnosed.

Figure 3: (A,E) Radiographs of the patient with thoracic juvenile idiopathic scoliosis 100 who underwent convex growth arrest and growing transpedicular instrumentation T4-L4 (B,F). (C,G) After 4 years proximal junctional kyphosis was 20,6^°, proximal segmental kyphosis was 28^°. Patient developed transitional kyphotic deformity and anchor subcutaneous prominence. (D,H) Prominent acres were removed, final fusion and instrumentation extension to T2 vertebra performed.

Figure 4: (A,B)Patient with infantile thoracic idiopathic scoliosis 100° and thoracic kyphosis 46°. (C,D) Growing transpedicular instrumentation T2-L4 and convex rod derotation were performed. Postoperative Cobb angle 29^° and thoracic kyphosis 39^°.

Table 1: Results of radiographic measurements data for the two groups. 

• Akbarnia BA, Marks DS (2000) Instrumentation with limited arthrodesis for the treatment of progressive early-onset scoliosis. Spine: State of the Art Reviews 14: 181-189.
• Harrington PR (1962) Treatment of scoliosis. Correction and internal fixation by spine instrumentation. J Bone Joint Surg Am 44: 591-610.
• Moe JH, Kharrat K, Winter RB, Cummine JL (1984) Harrington instrumentation without fusion plus external orthotic support for the treatment of difficult curvature problems in young children. Clin Orthop Relat Res pp: 35-45.
• Skaggs DL, Akbarnia BA, Flynn JM, Myung KS, Sponseller PD, et al. (2014) A classification of growth friendly spine implants. J Pediatr Orthop 34: 260-274.
• Skaggs D, Myung K, Lee C (2011) Proximal Junctional Kyphosis in Distraction-BasedGrowing Rods. The 5th International Congress on Early Onset Scoliosis and GrowingSpine (ICEOS): Orlando, Florida. Journal of Children’s Orthopaedics: Springer-Verlag pp 387-401.
• Li Y, Karlin L (2011) Proximal Segmental Kyphosis After Vertical Expandable ProstheticTitanium Rib Insertion. The 5th International Congress on Early Onset Scoliosis andGrowing Spine (ICEOS); Orlando, Florida. Journal of Children’s Orthopaedics: Springer-Verlag pp: 387-401.
• Mezentsev A, Petrenko D, Radchenko VA (2014) Idiopathic scoliosis: principles of managementIn: Bhave A. Modern techniques of spine surgery pp: 175-191.
• Kim HJ, Lenke LG, Shaffrey CI, Van Alstyne EM, Skelly AC (2012) Proximal junctional kyphosis as a distinct form of adjacent segment pathology after spinal deformity surgery: a systematic review. Spine 37: 144-164.
• Kim YJ, Bridwell KH, Lenke LG, Kim J, Cho SK (2005) Proximal junctional kyphosis in adolescent idiopathic scoliosis following segmental posterior spinal instrumentation and fusion: minimum 5-year follow-up. Spine 30: 2045-2050.
• Wang J, Zhao Y, Shen B, Wang C, Li M (2010) Risk factor analysis of proximal junctional kyphosis after posterior fusion in patients with idiopathic scoliosis. Injury 41: 415-420.
• Helgeson MD, Shah SA, Newton PO, Clements DH, Betz RR, et al. (2010) Evaluation of proximal junctional kyphosis in adolescent idiopathic scoliosis following pedicle screw, hook, or hybrid instrumentation. Spine 35:177-181.
• Akbarnia BA, Marks DS, Boachie-Adjie O, Thompson AG, Asher MA (2005) Dual growing rod technique for the treatment of progressive early onset scoliosis: A multi-center study. Spine 30: 46-57.
• Campbell RM Jr, Smith MD (2007) Thoracic insufficiency syndrome and exotic scoliosis. J Bone Joint Surg Am 1: 108-22.
• Cil A, Yazici M, Uzumcugil A, Kandemir U, Alanay Y, et al. (2005) The evolution of sagittal segmental alignment of the spine during childhood. Spine 30: 93-100.
• El-Hawary R, Sturm P, Cahill P, Samandi A (2011) Sagittal Spinopelvic Parameter Help Predict the Risk of Proximal Junctional Kyphosis for Children Treated with Posterior Distraction Based Implants Symposium: 5^th International Congress on Early Onset Scoliosis and Growing SpineOrlando, Florida.
• Reinker K, Simmons JW, Patil V, Stinson Z (2010) Can VEPTR® Control Progression of Early-onset Kyphoscoliosis?: A Cohort Study of VEPTR® Patients With Severe Kyphoscoliosis Symposium: Early Onset Scoliosis 469: 1342-1248.
• Debnath U, Harshavardhana N, Mehdian H, McConnell J (2008) Timing of Definitive Fusion in Early Onset Scoliosis (EOS) treated by Growing Rod (GR) Construct – A Retrospective study with a minimum follow-up of 13 years Symposium: 2^nd International Congress on Early Onset Scoliosis and Growing Spine, Montréal, Québec.