Rib hump is one of the main concerns in patients with adolescent idiopathic scoliosis (AIS).1,2 Surgical methods to address rib hump include thoracoplasty or vertebral column derotation (DVR) using all pedicle screw constructs.1,3 Thoracoplasty has been shown to be detrimental for pulmonary function.4,5 Use of thoracoplasty and/or DVR may increase surgical time and blood loss.6
Lee et al1 described the concept of DVR using all pedicle screw constructs. Using postoperative computed tomographic scans, they were able to demonstrate 43% axial plane correction of the vertebral bodies using DVR as compared with only 2.5% without it. However, they did not report the effect of DVR on rib hump, and it seemed that some of their patients had additional formal thoracoplasty. Suk et al2 reported 58% improvement in rib deformity using thoracoplasty and 72% improvement with the use of both DVR and thoracoplasty. However, they did not report the effect of DVR only on rib deformity. Recently, Hwang et al7,8 reported a mean improvement of 54% in rib prominence using DVR. Samdani et al6 observed similar rib deformity correction using thoracoplasty and DVR in patients with mild prominence (≤9°), whereas in patients with larger prominence, thoracoplasty was needed to obtain correction. Spontaneous apical vertebral derotation has been reported using pedicle screws even without formal DVR.9 On the contrary, forceful spinal derotation has been reported to cause plowing of thoracic pedicle screws and, therefore, aortic abutment during spinal deformity correction.10 However, no study thus far has directly evaluated the effectiveness of DVR on thoracic rib prominence and radiographical parameters as compared with pedicle screw instrumentation without DVR.
From a technical point of view, DVR can be performed using segmental or en bloc techniques.6 All pedicle screw constructs have a tendency to flatten thoracic kyphosis, but recent evidence suggests that DVR does not further increase this risk as compared with pedicle screws only without DVR.11 Some of these devices allow compression on the convexity only, which may further increase the risk of thoracic kyphosis flattening, whereas others allow lifting the low-lying thoracic apex upward at the time of spinal derotation. Most studies evaluating DVR report, however, both segmental and en bloc DVR together, have been retrospective, and are purely register-based or multisurgeon studies, thus further decreasing the available information on the details of surgical techniques used.6,7
MATERIALS AND METHODS
Seventy-two consecutive adolescents (aged 18 yr or younger at the time of surgery) operated for a structural thoracic idiopathic scoliosis (Lenke 1–4, 6) using all pedicle screw constructs at our hospitals were prospectively followed up for a 2-year period (Table 1). Six patients had juvenile idiopathic scoliosis (JIS) and the remainder had AIS with at least 45° major curve preoperatively. The first 24 consecutive patients operated between 2006 and 2008 without DVR served as a control group (N-DVR group) and the following 48 consecutive patients were operated using DVR (DVR group) between 2008 and 2010.
The study design was a comparative review of 2 cohorts with prospective data collection on patients operated using all pedicle screw constructs without (n = 24) or with DVR (n = 48) for structural thoracic idiopathic scoliosis. Clinical medical records, radiographs of the spine, and health-related quality of life (Scoliosis Research Society [SRS]-24 outcome questionnaire) were prospectively recorded and none of these patients were lost during minimum 2-year follow-up. Follow-up examinations were performed by 1 of 2 orthopedic spine surgeons (I.H. or M.M.) before surgery, on the day when the patient was discharged from the hospital, 6 months, and at 2 years after surgery. All patients were operated by 1 orthopedic spine surgeon (I.H.). Standing posteroanterior and side radiographs of the whole spine (scoliosis) were obtained preoperatively and at 6 and 24 months. At least 2 independent observers measured all radiographs (K.M., T.J., and K.K.). Spinal radiographs in case of difficult deformity were measured on the basis of a consensus decision and in borderline cases, the more conservative option was chosen. Spinal bending radiographs were obtained preoperatively to identify structural curves.12 The SRS-24 questionnaire13 was mailed to the patients and answers verified during the follow-up visits. The study was carried out from August 2006 to December 2012.
All patients were operated prone and posterior elements of the spine were exposed carefully with electrocautery. Fourteen (58%) patients without and 34 (71%) patients with DVR underwent Ponte procedure14 (P = 0.29). Mean (range) number of Ponte osteotomies was similar in the study groups (2.1 [0–6] for N-DVR group and 2.7 [0–7] for DVR group, P = 0.26). The levels of Ponte osteotomies in the N-DVR versus DVR group were as follows: Th 5/6 1 versus 4, Th6/7 4 versus 14, Th7/8 13 versus 23, Th8/9 13 versus 30, Th9/10 13 versus 28, Th10/11 6 versus 17, Th 11/12 2 versus 5, Th12/L1 1 versus 5, and L1/2 1 versus 5.
Pedicle screws were inserted with the freehand technique on the basis of posterior bony elements according to Kim et al.15 A single multiaxial reduction screw was used at the apical concave side, whereas the rest of the apical thoracic pedicle screws were monoaxial. All patients were instrumented with 5.5 mm or 6.35 Ti Alloy rods (Medtronics Spinal and Biologics, Memphis, TN).
Concave rod was slightly overcontoured into kyphosis to account for possible rod flattening while performing rod rotation. Rod insertion was started in both groups top to bottom. Rod insertion into pedicle screw heads was performed combining direct translation and rod rotation.
Correction of spinal deformity was obtained by concave rod derotation in the first 24 consecutive patients (N-DVR group). Final correction in both groups was obtained using coronal in situ bending, with the handles of the coronal in situ benders holded as low as possible to produce both scoliosis and hypokyphosis correction. Convex rod was inserted in situ without compression of the main thoracic curve to prevent flattening of the thoracic kyphosis. In case of structural upper thoracic and/or thoracolumbar/lumbar curves (concave rod), compression of screw heads was performed as necessary.
En bloc DVR was performed using the DVR device (VCM; Medtronics Spinal and Biologics) in the next 48 consecutive patients (DVR group). The VCM instrument was inserted into 3 apical pedicle screw pairs in the main thoracic curve. In a typical case, it was attached bilaterally into T8, T9, and T10 vertebral bodies with the exception of T9 concave multiaxial reduction screw. The device was used to derotate the thoracic spine by lifting up the low-lying apical major thoracic concave area and in this manner providing both DVR and hypokyphosis correction. En bloc DVR was carried out at the same time when the concave rod was rotated with single, concave rod in place. Spinal fusion was carried out using own bone material from facetectomies and osteomies with tricalcium phosphate and hydroxyapatite graft extenders (BCP and Nanostim; Medtronics Spinal and Biologics). Thoracoplasty has not been performed to any of the patients. Spinal cord monitoring (MEP, SSEP, and lumbar nerve root electromyography with or without pedicle screw stimulation) was performed in all of the operations.
Standard standing posteroanterior and lateral radiographs of the entire spine were obtained pre- and postoperatively and at follow-up visits. The proximal thoracic, main thoracic, and thoracolumbar/lumbar curves, coronal balance, and pelvic obliquity were measured from the posteroanterior radiographs, and thoracic kyphosis (T5–T12), lumbar lordosis (T12–S1), segmental kyphosis or lordosis, and sagittal balance were measured from the lateral radiographs.16,17 Correction of spinal rotation postoperatively was analyzed using the Upasani scoring method for the main thoracic curve.18
The patients underwent a systematic physical examination preoperatively, before discharge, and at follow-up visits. This included coronal and sagittal balance evaluation, measurement of thoracic and lumbar rib hump measurement using a scoliometer,19 and neurological examination of the lower legs. Thoracic rib hump measurements were available preoperatively and at final follow-up in 21 (88%) and 39 (81%) of the N-DVR and DVR patients, respectively.
The SRS-24 questionnaire13 was translated into Finnish and was mailed to all patients before their follow-up visit. SRS-24 questionnaires were available at final follow-up in 20 (83%) and 42 (88%) of the N-DVR and DVR patients, respectively. Filling of SRS-24 questionnaire was performed at home either by the patient or by caregivers, depending on the patient ability and returned at the follow-up visits. The answers were checked during the physical examination.
Values are given as means, standard deviations (SD), and ranges. A 2-tailed independent t test was used to calculate the level of significance for continuous variables (unpaired for between and paired for within group comparison). For categorical variables, χ2 test was used. P values below 0.05 were considered statistically significant.
We obtained permission to perform this study from the ethics committee of the hospital where the study was conducted. All subjects gave informed consent to participate in the study.
The preoperative curve flexibility of the major thoracic curve was similar in both groups in the bending radiographs (23% ± 18% for the N-DVR group and 25% ± 19% for the DVR group, respectively, P = 0.65). Preoperatively, the mean (SD) main thoracic curve was 56° ± 9° and 57° ± 11° and was corrected to 16° ± 6° in both groups at 2-year follow-up (not significant) (Table 2). No statistical difference was observed in the upper thoracic or lumbar curve correction at 6 months or 2-year follow-up between the study groups (Table 2).
TABLE 2-a. Radiograp...Image Tools
Correction of spinal rotation as assessed by the Upsani score18 was significantly better in the DVR group than in the N-DVR group at 6 months (P = 0.038) and 2-year follow-up (P = 0.039). Correction of spinal rotation decreased significantly in the N-DVR group during the 2-year follow-up (P = 0.016), but no change occurred in the DVR group (Table 2).
Thoracic kyphosis tended to flatten in the N-DVR group, whereas no such change occurred in the DVR group (P = 0.11 between groups at 2-year follow-up) (Table 2). No difference in the mean thoracic kyphosis was noted in patients operated with 5.5-mm or 6.35 Ti Alloy rods. Coronal unbalance (>20 mm) was found in 3 patients (13%) in the N-DVR group and 4 patients (8%) in the DVR groups, respectively (P = 0.57). Positive sagittal balance of 20 mm or more was observed in 7 (29%) and 11 (23%) of the study groups (P = 0.56), but none of the patients had a sagittal vertical line in front of the midpoint of the femoral heads.
Clinical Data, Rib Hump Correction, and SRS-24 Outcome Questionnaire
TABLE 2-b. Radiograp...Image Tools
Patients in the N-DVR group were significantly younger than patients in the DVR group at the time of surgery (mean 13.7 yr vs. 15.1 yr, P = 0.0033), and there tended to be more girls in the N-DVR group than in the DVR group (P = 0.092). Thoracic rib hump averaged 12.3° ± 3.6° versus 14.2° ± 5.0° (P = 0.075) preoperatively and 7.2° ± 3.8° versus 8.3° ± 3.7° at 2-year follow-up in both the N-DVR and DVR groups, respectively (P = 0.30) (Table 3). Thus, the mean (SD) final correction of thoracic rib hump was 40% ± 31% in the N-DVR groups and 44% ± 26% in the DVR group (P = 0.62). No statistically significant differences were observed in the perioperative blood loss or operative time between the study groups (Table 1). No spinal cord or permanent nerve root deficits were observed in the study groups.
The SRS-24 total score averaged 100.4 ± 5.7 points in the N-DVR group and 99.3 ± 9.0 points in the DVR group, respectively, at 2-year follow-up (P = 0.56). No significant differences existed in the 7 main domains of SRS-24 between the study groups at 2-year follow-up (Figure 1).
Matched Cohort Comparison
To minimize the compounding effects of age at surgery, sex, and primary diagnosis as reliably as possible, 2 matched cohorts were formed. For every N-DVR patient (n = 24), a DVR group patient, who matched the best for age at surgery (±1 yr), sex, thoracic rib hump size (±2°), primary diagnosis (JIS or AIS), and curve type (Lenke classification), was selected from the rest of the original study population (n = 48). Two N-DVR group patients could not be matched sufficiently and they had to be excluded (Table 4).
Preoperatively, mean main thoracic curve was 56° ± 9.2° and 55° ± 9.5° (P = 0.79) and was corrected to 16° ± 5.7° in both groups at final follow-up (P = 0.92). Thoracic rib hump averaged 12.5° ± 3.7° in N-DVR versus 12.7° ± 4.8° in DVR (P = 0.92) preoperatively and 7.2° ± 4.0° versus 8.5° ± 3.6° at 2-year follow-up (P = 0.35). Correction of thoracic rib hump was 41% ± 32% in N-DVR and 38% ± 31% in DVR groups at 2-year follow-up, respectively (P = 0.973). Radiographical correction of spinal rotation (Upasani score) was significantly better in the DVR group (1.14) than in the N-DVR group (0.73) at 6-month follow-up (P = 0.049) and tended to be better at 2-year follow-up (P = 0.13) (Figures 2 and 3). Thoracic kyphosis (T5–T12) averaged 25° ± 18° in the N-DVR group and 24° ± 11° in the DVR group before surgery and 22° ± 9.3° and 24° ± 9.0° at 2-year follow-up (P = 0.53).
Validity of the Data
The strengths of this study are a single surgeon, prospective data collection with all patients undergoing similar surgery with or without DVR with similar instrumentation and instruments for structural thoracic idiopathic scoliosis. None of the patients underwent thoracoplasty. Despite prospective data collection, not all patients had their rib humps measured at follow-up visits, but radiographical follow-up rate was 100% and more than 80% of the patients had both rib hump measurements and SRS-24 questionnaires available. As the rib hump correction slightly diminished during the follow-up period, it would have been interesting to see the immediate rib deformity correction. However, we did not perform Adams forward bending tests immediately after surgery and, thus, the immediate rib hump correction remains obscure.
In the consecutive patient comparison, DVR patients were older and there were more males. It is possible that older age and male sex make spinal deformity correction more difficult.20 To overcome this bias in the prospective consecutive follow-up study, an additional matched cohort comparison was performed. According to this matched cohort comparison, DVR provided better correction of spinal rotation and preserved thoracic kyphosis better but did not allow more rib hump correction. Thus, our findings remained essentially the same even with the matched cohort comparison, suggesting that the different demographic characteristics do not explain the differences observed.
There is probably a learning curve effect in this series, and although the surgical times were the same in the 2 groups, the en bloc DVR and more Ponte osteotomies were performed in the DVR group. The corrective maneuvers performed in the patient groups were nearly identical and were performed by a single orthopedic spine surgeon with the exception of the use of the en bloc DVR in the latter group. Thus, the authors think that these observed differences are valid.
Comparison With Previous Data
Suk et al2 reported a 38% improvement in rib deformity angle with scoliosis correction in the absence of thoracoplasty or DVR. This 38% is very close to our 40% in the N-DVR group at 2-year follow-up. Suk et al2 also reported 58% improvement with thoracoplasty and 72% improvement with the use of both thoracoplasty and DVR. However, they did not report the correction of rib hump with DVR only.2 Recently, Hwang et al7,8 reported approximately 50% reduction in the rib hump deformity as assessed by scoliometer, which is very close to our 51% at 6 months and 44% at 2-year follow-up in the DVR group.
Recently, Hwang et al21 evaluated effectiveness of DVR on thoracolumbar (Lenke 5C) curves in 34 patients. They were not able to demonstrate any significant differences in the thoracolumbar rib hump or radiographical parameters between the 19 patients operated with DVR and the 15 patients operated without it. The limitations of the study include various DVR techniques (segmental or en bloc or both), multisurgeon design (the details of surgical correction may vary), and the retrospective nature of the study. In addition, thoracolumbar rib hump is probably much more flexible and easier to correct than main thoracic rib hump as it locates in the more mobile segment of spine.
Although direct vertebral column en bloc derotation produces immediate rib hump correction at the time of surgery, it seems that some of this correction will be lost during 2-year follow-up. Spinal instrumentation is rigid and no loss of correction occurred in the coronal plane. Correction of spinal rotation was significantly better in the DVR group than in the N-DVR group at 2-year follow-up. It is possible that the uninstrumented rib deformity recurs like a spring. Another possible explanation for rib prominence recurrence could be continued thoracic cage growth, which we call as “rib crankshaft.” The final volume of thoracic cage growth will be reached by the age of 15 years,22 and more than half of our patients were below this age at the time of surgery.
It should also be noted that the N-DVR group underwent correction of the spinal deformity using monosegmental thoracic pedicle screws apically, which have been shown to be significantly more effective than multiaxial pedicle screws for rotational correction of AIS.23 Use of transverse connectors has been shown to increase rotational stability of the segmental pedicle screw instrumentation.24 The absence of transverse connectors may have affected the rotational stability of our constructs and thus resulted in partial loss of spinal rotational correction. However, no loss of spinal rotation correction was observed in the DVR group during follow-up. One more option to enhance rotational correction of the rib hump could be its overcorrection during primary deformity surgery. It should be noted, however, that this maneuver can increase the risk of plowing of pedicle screws as well as the asymmetry of the anterior part of thoracic cage.
In contrast with the findings by Hwang et al,11 flattening of thoracic hypokyphosis did not occur in the DVR group. Interestingly, this flattening tendency was observed only in the N-DVR group. One explanation could be the ability to lift up the low-lying main thoracic apical area, with the derotation device used in this study. DVR instruments, which allow not only compression but also lifting up the concave side, may thus increase the possibility to maintain thoracic kyphosis.
* Total pedicle screw instrumentation with monoaxial screws apically provides excellent radiographical coronal correction of structural thoracic idiopathic scoliosis.
* No significant differences were observed in thoracic rib hump correction between no derotation and en bloc derotation groups.
* Radiographical correction of spinal rotation is significantly better with en bloc derotation at 2-year follow-up.
* En bloc DVR may help preserving thoracic kyphosis.
The authors thank research nurses Katariina Mattila and Katariina Kauste for help with radiographical analyses.
The funding body did not play a role in the investigation. The funds were used only for salaries of the researchers and research nurses.
1. Lee SM, Suk SI, Chung ER. Direct vertebral rotation: a new technique of three-dimensional deformity correction with segmental pedicle screw fixation in adolescent idiopathic scoliosis. Spine 2004;29:343–9.
2. Suk SI, Kim JH, Kim SS, et al. Thoracoplasty in thoracic adolescent idiopathic scoliosis. Spine 2008;33:1061–7.
3. Ledonio CG, Polly DW, Vitale MG, et al. Pediatric pedicle screws: comparative effectiveness and safety. J Bone Joint Surg Am 2011;93:1227–34.
4. Newton PO, Faro FD, Gollogly S, et al. Results of preoperative pulmonary function testing of adolescent idiopathic scoliosis. J Bone Joint Surg Am 2005;87:1937–46.
5. Kim YJ, Lenke LG, Bridwell KH, et al. Pulmonary function in adolescent idiopathic scoliosis relative to surgical procedure. J Bone Joint Surg Am 2005;87:1534–41.
6. Samdani AF, Hwang SW, Miyanji F, et al. Direct vertebral body derotation, thoracoplasty, or both. Which is better with respect to inclinometer and Scoliosis Research Society-22 scores. Spine 2012;37:E849–53.
7. Hwang SW, Samdani AF, Lonner B, et al. Impact of direct vertebral body derotation on rib prominence. Are preoperative factors predictive of changes in rib prominence? Spine 2012;37:E86–9.
8. Hwang SW, Samdani AF, Cahill PJ. The impact of segmental and en bloc derotation maneuvers on scoliosis correction and rib prominence in adolescent idiopathic scoliosis. J Neurosurg Spine 2012;16:345–50.
9. Luk KD, Cheung WY, Wong Y, et al. The predictive value of the fulcrum bending radiograph in spontaneous apical vertebral derotation in adolescent idiopathic scoliosis. Spine 2012;37:E922–6.
10. Wagner MR, Flores JB, Sanpera I, et al. Aortic abutment after direct vertebral rotation. Plowing of pedicle screws. Spine 2011;36:243–7.
11. Hwang SW, Samdani AF, Gressot LV, et al. Effect of direct vertebral body derotation on the sagittal profile in adolescent idiopathic scoliosis. Eur Spine J 2012;21:31–9.
12. Lenke LG, Betz RB, Harms J, et al. Adolescent idiopathic scoliosis. A new classification to determine extent of spinal arthrodesis. J Bone Joint Surg Am 2001;83:1169–81.
13. Haher TR, Gorup JM, Shin TM. Results of the Scoliosis Research Society Instrument for evaluation of surgical outcome in adolescent idiopathic scoliosis. A multicenter study of 244 patients. Spine 1999;24:1435–40.
14. Ponte A, Siccardi GL. The biomechanical advantage of an innovative posterior technique for correction of Scheuermann's kyphosis. Annals of the First Combined Meeting of the Leading European Spine Societies. Eurospine 1996;5:91.
15. Kim YJ, Lenke LG, Bridwell KH, et al. Free hand pedicle screw placement in the thoracic spine. Is it safe? Spine 2004;29:333–42.
16. Cobb J. Outline for the study of scoliosis. Instr Course Lect AAOS 1948;5:261–75.
17. O'Brien MF, Kuklo TR, Blanke KM, et al. Spinal Deformity Study Group. Radiographic Measurement Manual. Memphis, TN: Medtronic Sofamor Danek Inc; 2004.
18. Upasani VV, Chambers RC, Dalal AH, et al. Grading apical vertebral rotation without a computed tomography scan: a clinically relevant system based on the radiographic appearance of bilateral pedicle screws. Spine 2009;34:1855–62.
19. Bunnel WP. An objective criterion for scoliosis screening. J Bone Joint Surg Am 1984;66:1381–7.
20. Helenius I, Remes V, Yrjönen T, et al. Does gender affect outcome of surgery in adolescent idiopathic scoliosis. Spine 2005;30:462–7.
21. Hwang SW, Dubaz OM, Ames R, et al. The impact of direct vertebral derotation on the lumbar prominence in Lenke type 5C curves. J Neurosurg Spine 2012;17:308–12.
22. Dimeglio A. Growth in pediatric orthopedics. In: Morrissy RT, Weinstein SL, ed. Lovell and Winter's Pediatric Orthopaedics. Philadelphia, PA: Lippincott Williams & Wilkins; 2006:35–65.
23. Kuklo TR, Potter BK, Polly DW Jr, et al. Monaxial versus multiaxial thoracic pedicle screws in the correction of adolescent idiopathic scoliosis. Spine 2005;30:2113–20.
24. Kuklo TR, Dmitriev AE, Cardoso MJ, et al. Biomechanical contribution of transverse connectors to segmental stability following long segment instrumentation with thoracic pedicle screws. Spine 2008;33:E482–7.
adolescent idiopathic scoliosis; pedicle screws; direct vertebral column derotation