Fusion May Not Be a Necessary Procedure for Surgically Treated Burst Fractures of the Thoracolumbar and Lumbar Spines: A Follow-up of at Least Ten Years

Chou, Po-Hsin MD; Ma, Hsiao-Li MD; Wang, Shih-Tien MD; Liu, Chien-Lin MD; Chang, Ming-Chau MD; Yu, Wing-Kwong MD

Journal of Bone & Joint Surgery - American Volume: 15 October 2014 - Volume 96 - Issue 20 - p 1724–1731
doi: 10.2106/JBJS.M.01486
Scientific Articles

Background: The surgical results of treating thoracolumbar and lumbar burst fractures were reported to be comparable between patients with and without fusion in an intermediate-term follow-up. To our knowledge, no prior report has compared the results of fusion and non-fusion with long-term follow-up.

Methods: This study was designed to provide long-term evaluation of patients with a burst fracture of the thoracolumbar and lumbar spine treated with short-segment fixation who were randomly assigned to a fusion or non-fusion group. Patients older than sixty years of age at the time of injury and those who were lost to follow-up were excluded. Functional outcomes were evaluated using the Greenough Low-Back Outcome Score and the visual analog scale for back pain. Radiographic outcomes were focused on the vertebral body height of the injured vertebra, the kyphotic angle, and the regional segmental motion.

Results: Twenty-two patients were enrolled in the non-fusion group, and twenty-four patients were enrolled in the fusion group. The average follow-up period was 134 months (range, 121 to 161 months). The average preoperative kyphotic angle was 16.4° for the non-fusion group and 19.5° for the fusion group. The average postoperative kyphotic angle was 1.5° for the non-fusion group and 4.0° for the fusion group. At the time of the latest follow-up, the average kyphotic angle was 13.8° for the non-fusion group and 14.7° for the fusion group. The average kyphotic angle between the two groups was similar at all follow-up times. A progressive decrease of the kyphotic angle was significant (p < 0.05) with time, regardless of fusion. The radiographic outcomes were similar between these two groups at all follow-up times, as were functional outcomes. More patients in the non-fusion group underwent additional surgery to remove implants. Regional segmental motion was preserved in the non-fusion group, with a mean motion (and standard deviation) of 4.2° ± 1.9°.

Conclusions: The long-term results of short segmental fixation with and without fusion for burst fractures of the thoracolumbar and lumbar spine were comparable. Regional segmental motion could be preserved without fusion, and bone graft donor site complications could be eliminated.

Level of Evidence: Therapeutic Level II. See Instructions for Authors for a complete description of levels of evidence.

1School of Medicine, National Yang-Ming University, Taipei, No. 155, Sec. 2, Linong Street, Taipei, 112, Taiwan, Republic of China. E-mail address for S.-T. Wang: stwang@vghtpe.gov.tw

2Department of Orthopedics and Traumatology, Taipei Veterans General Hospital, No. 201, Sec. 2, Shipai Road, Taipei, 112, Taiwan, Republic of China

Article Outline

According to the Denis classification1, burst fracture is defined as a two or three-column injury to the spine. There continues to be considerable controversy regarding the treatment of burst fractures of the thoracolumbar and lumbar spine2-8. Nonoperative treatment for burst fractures without neurologic deficit has been reported to yield good results, although a greater residual kyphotic angle was noted7-11. However, greater residual deformity does not correlate with symptoms at the time of follow-up9. Good clinical results have been reported with posterior reduction with short segment transpedicular screw fixation and posterior fusion3,12-15 or posterolateral fusion16,17 and have been reported retrospectively with short-segment fixation without fusion2,18. Several prospective studies16,19, including our own work20, showed comparable functional and radiographic results for fusion and non-fusion in surgically treated burst fractures with mean follow-up times ranging from 23.9 to seventy-two months. A long-term comparative study is needed to assess functional and radiographic outcomes, especially preservation of the motion segment in the long term, to determine whether or not fusion should be a routine procedure for surgically treated burst fractures of the thoracolumbar and lumbar spines. We report a long-term comparative study of fusion and non-fusion based on our previous work20, with an average follow-up time of 134 months.

Back to Top | Article Outline

Materials and Methods

This long-term study was based on our previous prospective randomized study20. From June 1996 to June 2003, fifty-eight patients with thoracolumbar and lumbar burst fractures were treated with posterior short-segment transpedicular screw fixation with one lordosing screw at the injured level and were enrolled in the study. Two groups were randomly determined by a roll of the dice at the time that surgical treatment was scheduled. The surgical indications and inclusion and exclusion criteria were described in detail in our previous report20. The additional exclusion criteria used in this long-term study were those lost to follow-up and those with injury at an age of more than sixty years to eliminate the possibility of age-related disc degeneration. The details of the surgical procedures and postoperative care of the patients were described in our previous report20. The research protocol was approved by the institutional review board at our hospital.

Serial supine anteroposterior and lateral radiographs were made immediately postoperatively and postoperatively at six weeks, three months, six months, twelve months, and then annually. Supine dynamic flexion and extension lateral radiographs were made postoperatively at two years and at the latest follow-up.

The injured vertebral body height was divided into anterior, middle, and posterior thirds, and each was measured in units of centimeters. Using the Cobb method9, the kyphotic angle was measured between the end plates with one body above and one body below the injured level. Other radiographic parameters such as adjacent spondylolisthesis or retrolisthesis, implant loosening or breakage, and formation of bridging callus were also assessed during follow-up. Radiographic spondylolisthesis or retrolisthesis of the adjacent motion segment was defined as slippage of the adjacent vertebra of >5 mm at one level caudal or cephalic to the fixation level in the static lateral radiographs at the latest follow-up21. Postoperative regional segmental motion, as well as the motion of the injured intervetebral disc and uninjured intervertebral disc around the injured vertebra, was calculated with the angular change using supine dynamic flexion-extension lateral radiographs. The method of measuring vertebral body height and regional segmental motion was based on that of Mumford et al.9.

All measurements were performed by two of the study authors (P.-H.C. and H.-L.M.), who were not involved in the surgery. All radiographic measurements were obtained three times by one of the two authors, with the arithmetic mean recorded as our data. To alleviate bias, the same protractor was used during measurements22.

Solid fusion was defined in the fusion group as an angular change of <5° of the fused segments in dynamic flexion and extension lateral radiographs21 and in the non-fusion group as no change in the injured vertebral body height.

Functional outcomes were evaluated using the Greenough Low-Back Outcome Score23 and the visual analog scale (VAS) for back pain at the time of the latest follow-up. The patient’s neurologic status was also evaluated with the Frankel grading system24 at the time of the latest follow-up. This study was registered at ClinicalTrials.gov (NCT01979198).

Statistical analysis was performed with the SPSS for Windows statistical package, version 15.0 (SPSS, Chicago, Illinois). The Student t test for continuous data and the chi-square test for categorical data were used for comparisons of variable parameters between the fusion and non-fusion groups at various follow-up times. The repeated-measure analysis of variance (ANOVA) test was used to discover whether the progressive change in the kyphotic angle and injured vertebral body height may alter with time. A significant difference was determined at p < 0.05. We used the t test to compare whether each parameter was significantly different (p < 0.05) between the two groups. (The significance of a Type-I error was set at p = 0.05.) Using the t test, we evaluated the power and investigated whether these comparative results were convincing or not. We used G*Power software (Heinrich-Heine Universität Düsseldorf, Düsseldorf, Germany) to calculate the power in each comparison between two groups.

Back to Top | Article Outline

Source of Funding

There was no external source of funding.

Back to Top | Article Outline

Results

Forty-six patients were enrolled in this long-term study: twenty-two in the non-fusion group and twenty-four in the fusion group. Of the twelve patients who were excluded, ten were lost to follow-up with a follow-up rate of 82.1% and two were older than sixty years of age at the time of injury. The overall mean age of the patients at the time of the operation was 39.4 years (range, sixteen to fifty-three years): thirty-nine years for the fusion group and forty years for the non-fusion group. The overall mean body mass index was 23.8 kg/m2 (range, 16.7 to 33.9 kg/m2): 24.5 kg/m2 in the non-fusion group and 23.5 kg/m2 in the fusion group.

According to the Denis classification1, in the non-fusion group, six patients were Type A and sixteen patients were Type B; in the fusion group, five patients were Type A and nineteen patients were Type B. In the non-fusion group, two patients were injured at T12, seven patients were injured at L1, and thirteen patients were injured at L2. In the fusion group, seven patients were injured at T12, twelve patients were injured at L1, and five patients were injured at L2. Clinical and radiographic examinations were available for the forty-six patients for an overall average of 134 months of follow-up (range, 121 to 161 months): 137 months for the non-fusion group and 132 months for the fusion group. No significant difference (p ≥ 0.05) was present in the demographic data between these two groups (Table I).

The average preoperative kyphotic angle was 16.4° in the non-fusion group and 19.5° in the fusion group. The average postoperative kyphotic angle was 1.5° in the non-fusion group and 4.0° in the fusion group. At the time of the latest follow-up, the average kyphotic angle was 13.8° in the non-fusion group and 14.7° in the fusion group. The average kyphotic angles of the two groups at all follow-up times were similar. Compared with the preoperative kyphotic angle, the average loss of correction was 10.7° in the fusion group and 12.3° in the non-fusion group, which was not significant (p = 0.42) (Table II).

The overall average preoperative vertebral body height at the anterior third was 1.58 cm: 1.58 cm in the fusion group and 1.59 cm in the non-fusion group. The average postoperative vertebral body height at the anterior third was 2.5 cm: 2.51 cm in the fusion group and 2.49 cm in the non-fusion group. At the time of the latest follow-up, the vertebral body height at the anterior third was 2.24 cm: 2.23 cm in the fusion group and 2.27 cm in the non-fusion group (Table II) (see Appendix). The measured middle and posterior thirds of the vertebral body height at various times (preoperative, postoperative, and latest follow-up) are shown in Table II and the Appendix.

A progressive loss of correction in the kyphotic angle was observed with time regardless of fusion and was significant (p < 0.05) (Fig. 1). A progressive decrease in the vertebral body height of the injured level was observed as time went on, and was not significant (p ≥ 0.05) (see Appendix). At all follow-up times, there was no significant difference (p ≥ 0.05) between the two groups in the loss of correction in the kyphotic angle (Fig. 1) and in the vertebral body height (see Appendix).

The mean regional segmental motion (and standard deviation) in the non-fusion group at the time of the latest follow-up was 4.2° ± 1.9° (range, 3° to 8°), similar to that at the time of the postoperative two-year follow-up (4.6° ± 2.4°) (Table II) (Fig. 2). The average motion was 0° for the injured intervertebral disc and 4.0° (range, 3° to 8°) for the uninjured intervertebral disc, which indicated that the regional segmental motion was affected mainly by the motion of the uninjured intervertebral disc. Eighteen patients (ten in the non-fusion group and eight in the fusion group) had bridging callus at the injured intervertebral disc during follow-up.

Twenty-seven patients (eighteen in the non-fusion group and nine in the fusion group) had their spinal implants removed during follow-up. Eight patients had screw breakage during follow-up: three in the non-fusion group and five in the fusion group (Table II). Bone graft donor site pain was noted in six patients in the fusion group. Neither spondylolisthesis nor retrolisthesis was observed at the adjacent level in either group.

The mean functional outcomes were 2.1 points (range, 0 to 4 points) for the VAS and 65.3 points (range, 52 to 75 points) for the Low-Back Outcome Score in the non-fusion group and 2 points (range, 0 to 4 points) for the VAS and 65.8 points (range, 55 to 75 points) for the Low-Back Outcome Score in the fusion group. The functional outcomes were similar between these two groups at the time of the latest follow-up using the Low-Back Outcome Score and the VAS for back pain (Table II). Neither nonunion nor postoperative infection was observed in either group at the time of the latest follow-up. On the basis of the Frankel grading system24, the neurologic status remained unchanged at the time of the latest follow-up as compared with that at the time of the two-year follow-up20.

Power analysis showed that G*Power reached 0.99 for the parameter of regional segmental motion. However, G*Power for the other parameters was insufficient to reach a conclusion.

Back to Top | Article Outline

Discussion

The goals of treatment for spinal injuries are the restoration of a pain-free state and the preservation of the dynamic and protective function of the spine. In general, a surgical approach is chosen in cases in which conservative treatment is unlikely to produce satisfactory results25. The roles of spinal instrumentation are fracture reduction, maintenance of reduction, rigid fixation, and enhancement of osseous union or fusion. Several transpedicular screw instrumentation systems have been reported to have satisfactory results in treating burst fractures of the thoracolumbar and lumbar spine with fusion12,13,15. However, satisfactory results have also been reported, using the VAS for back pain, the Low-Back Outcome Score, or the Short-Form 36 (SF-36), with the non-fusion method for posterior fixation in a number of studies2,17-19, as well as in our previous work20.

Significant progression of regional kyphosis was observed as time went on in both the non-fusion and fusion groups, but without significance between the two groups. However, significance was observed in progressive kyphosis with time (14.1°) compared with the immediate postoperative kyphotic angle (2.8°), regardless of non-fusion or fusion. The same phenomenon was also reported in loss of correction with either fusion17,26,27 or non-fusion2,17,18. Defino and Canto26 reported progressive kyphosis at the final follow-up, even when using transpedicular bone-grafting along with posterolateral fusion to prevent further progressive kyphosis. On the basis of these reports, progressive kyphosis may be inevitable in spite of fusion, and the residual deformity did not correlate with the symptoms at the time of follow-up9.

Removing implants was suggested at one year postoperatively for those patients in the non-fusion group, because of the possibility of implant failure. For those patients in the fusion group, implants were removed only when the patient requested it. In Taiwanese culture, many people believe it is not good for one’s health to have metallic implants inside the body. Therefore, nine patients in the fusion group strongly requested removal of the implants, either for cultural reasons or back soreness. In the non-fusion group, eighteen patients underwent another surgery of implant removal, and the other four patients did not want another surgery. In all, eight patients had screw breakage: three patients in the non-fusion group and five patients in the fusion group.

A small, insignificant decrease in the injured vertebral body height was found in our study. Similar results were also noted in other reports17,26,27, which used different methods to evaluate absolute height (measured in centimeters) or relative ratio to adjacent vertebrae (measured in percentages). Therefore, the decreased vertebral body height may not play an important role in progressive kyphosis.

Haschtmann et al.28 concluded that vertebral end-plate trauma induces intervertebral disc cell apoptosis and necrosis in both nucleus and anulus tissues. Burst fracture could also contribute to either superior or inferior end-plate fracture at the injured vertebrae. Therefore, either the superior or inferior disc heights at the injured vertebrae would inevitably decrease because of further disc degeneration caused by apoptosis associated with the prior injury17,26,27.

As a result of our study, we believe that the progressive loss of disc height at the injured disc plays a more important role in worsening kyphosis than the mild degree of decreased vertebral body height. Wang et al.17 also concluded that the collapse of the disc space was the major contributor to the loss of correction.

In this study with a follow-up of at least ten years, we found that regional segmental motion could be preserved with a mean motion of 4.2°, compared with 4.6° in the prior two-year follow-up. The mean motion of the injured intervertebral disc was 0° and the mean motion of the uninjured intervertebral disc was 4.0° (range, 3° to 8°). This indicated that most of the regional segmental motion may be attributed to the uninjured intervertebral disc. In the fusion group, the regional segmental motion remained similar to that measured at the two-year follow-up, because of the solid union at the posterior portion of these fused motion segments.

This study had several limitations. It was not a multicenter trial and the patient number enrolled in both groups was small, with less-than-desired power. Therefore, the comparison may not truly reflect substantive differences between these two groups. We made efforts to contact all patients to perform the latest follow-up, although there were still difficulties in conducting a thorough follow-up because of individual factors. Moreover, the G*Power in several parameters was not enough to reach a conclusion regarding those parameters, and we did not check the intraobserver and interobserver reliability, which has been reported to have approximately 5° to 7° of interobserver variability22.

In conclusion, this study found that radiographic and functional outcomes were similar between the non-fusion group and the fusion group after short-segment transpedicular screw fixation in thoracolumbar and lumbar burst fractures, and regional segmental motion in the non-fusion group was maintained compared with that at the two-year follow-up. The advantages of the non-fusion procedure reported in our previous study20 were fewer bone graft donor site complications, the saving of more motion segments, and reduced blood loss and operative time. We conclude that posterior fusion with autograft need not be a routine procedure for surgically treated thoracolumbar and lumbar burst fractures.

Back to Top | Article Outline

Appendix Cited Here...

A figure showing the serial change of injured vertebral body heights at different follow-up times is available with the online version of this article as a data supplement at jbjs.org.

NOTE: We thank Shin-I Huang from the Biostatistics Task Force, Taipei Veterans General Hospital, for statistical assistance.

A commentary by Daisuke Togawa, MD, PhD, is linked to the online version of this article at jbjs.org.

Investigation performed at the School of Medicine, National Yang-Ming University, and the Department of Orthopedics and Traumatology, Taipei Veterans General Hospital, Taipei, Taiwan, Republic of China

Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. None of the authors, or their institution(s), have had any financial relationship, in the thirty-six months prior to submission of this work, with any entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. Also, no author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.

Back to Top | Article Outline

References

1. Denis F. Spinal instability as defined by the three-column spine concept in acute spinal trauma. Clin Orthop Relat Res. 1984 ;(189):65–76.
2. Sanderson PL, Fraser RD, Hall DJ, Cain CM, Osti OL, Potter GR. Short segment fixation of thoracolumbar burst fractures without fusion. Eur Spine J. 1999;8(6):495–500.
3. Parker JW, Lane JR, Karaikovic EE, Gaines RW. Successful short-segment instrumentation and fusion for thoracolumbar spine fractures: a consecutive 41/2-year series. Spine (Phila Pa 1976). 2000 ;25(9):1157–70.
4. Marco RA, Kushwaha VP. Thoracolumbar burst fractures treated with posterior decompression and pedicle screw instrumentation supplemented with balloon-assisted vertebroplasty and calcium phosphate reconstruction. J Bone Joint Surg Am. 2009 ;91(1):20–8.
5. Reinhold M, Knop C, Beisse R, Audigé L, Kandziora F, Pizanis A, Pranzl R, Gercek E, Schultheiss M, Weckbach A, Bühren V, Blauth M. Operative treatment of 733 patients with acute thoracolumbar spinal injuries: comprehensive results from the second, prospective, Internet-based multicenter study of the Spine Study Group of the German Association of Trauma Surgery. Eur Spine J. 2010 ;19(10):1657–76. Epub 2010 May 25.
6. Verlaan JJ, Diekerhof CH, Buskens E, van der Tweel I, Verbout AJ, Dhert WJ, Oner FC. Surgical treatment of traumatic fractures of the thoracic and lumbar spine: a systematic review of the literature on techniques, complications, and outcome. Spine (Phila Pa 1976). 2004 ;29(7):803–14.
7. Tian NF, Wu YS, Zhang XL, Wu XL, Chi YL, Mao FM. Fusion versus nonfusion for surgically treated thoracolumbar burst fractures: a meta-analysis. PLoS One. 2013;8(5):e63995. Epub 2013 May 21.
8. Gnanenthiran SR, Adie S, Harris IA. Nonoperative versus operative treatment for thoracolumbar burst fractures without neurologic deficit: a meta-analysis. Clin Orthop Relat Res. 2012 ;470(2):567–77. Epub 2011 Nov 5.
9. Mumford J, Weinstein JN, Spratt KF, Goel VK. Thoracolumbar burst fractures. The clinical efficacy and outcome of nonoperative management. Spine (Phila Pa 1976). 1993 ;18(8):955–70.
10. Weinstein JN, Collalto P, Lehmann TR. Thoracolumbar “burst” fractures treated conservatively: a long-term follow-up. Spine (Phila Pa 1976). 1988 ;13(1):33–8.
11. Shen WJ, Shen YS. Nonsurgical treatment of three-column thoracolumbar junction burst fractures without neurologic deficit. Spine (Phila Pa 1976). 1999 ;24(4):412–5.
12. Liu CL, Wang ST, Lin HJ, Kao HC, Yu WK, Lo WH. AO fixateur interne in treating burst fractures of the thoracolumbar spine. Zhonghua Yi Xue Za Zhi (Taipei). 1999 ;62(9):619–25.
13. Lin HJ, Wang ST, Liu CL, Kao HC, Yu WK, Lo WH. Reduction-fixation system in treating burst fractures of thoracolumbar spine. J Spine Surg ROC. 1996;2:102–11.
14. Liu CL, Dai LG, Wang ST, Yu WK, Lo WH. Surgical treatment of thoracic spine burst fractures. J Spine Surg ROC. 1996;2:92–101.
15. Liu CL, Wang ST, Lin HJ, Kao HC, Yu WK, Lo WH. Roy-Camille platting system in treating burst fractures of thoracolumbar spine. J Orthop Surg ROC. 1996;13:309–16.
16. Dai LY, Jiang LS, Jiang SD. Posterior short-segment fixation with or without fusion for thoracolumbar burst fractures. a five to seven-year prospective randomized study. J Bone Joint Surg Am. 2009 ;91(5):1033–41.
17. Wang XY, Dai LY, Xu HZ, Chi YL. Kyphosis recurrence after posterior short-segment fixation in thoracolumbar burst fractures. J Neurosurg Spine. 2008 ;8(3):246–54.
18. Kim YM, Kim DS, Choi ES, Shon HC, Park KJ, Cho BK, Jeong JJ, Cha YC, Park JK. Nonfusion method in thoracolumbar and lumbar spinal fractures. Spine (Phila Pa 1976). 2011 ;36(2):170–6.
19. Jindal N, Sankhala SS, Bachhal V. The role of fusion in the management of burst fractures of the thoracolumbar spine treated by short segment pedicle screw fixation: a prospective randomised trial. J Bone Joint Surg Br. 2012 ;94(8):1101–6.
20. Wang ST, Ma HL, Liu CL, Yu WK, Chang MC, Chen TH. Is fusion necessary for surgically treated burst fractures of the thoracolumbar and lumbar spine?: a prospective, randomized study. Spine (Phila Pa 1976). 2006 ;31(23):2646–52; discussion 2653.
21. McAfee PC, Boden SD, Brantigan JW, Fraser RD, Kuslich SD, Oxland TR, Panjabi MM, Ray CD, Zdeblick TA. Symposium: a critical discrepancy-a criteria of successful arthrodesis following interbody spinal fusions. Spine (Phila Pa 1976). 2001 ;26(3):320–34.
22. Morrissy RT, Goldsmith GS, Hall EC, Kehl D, Cowie GH. Measurement of the Cobb angle on radiographs of patients who have scoliosis. Evaluation of intrinsic error. J Bone Joint Surg Am. 1990 ;72(3):320–7.
23. Greenough CG, Fraser RD. Assessment of outcome in patients with low-back pain. Spine (Phila Pa 1976). 1992 ;17(1):36–41.
24. Frankel HL, Hancock DO, Hyslop G, Melzak J, Michaelis LS, Ungar GH, Vernon JD, Walsh JJ. The value of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetrapelgia. I. Paraplegia. 1969 ;7(3):179–92.
25. Daniaux H, Seykora P, Genelin A, Lang T, Kathrein A. Application of posterior plating and modifications in thoracolumbar spine injuries. Indication, techniques, and results. Spine (Phila Pa 1976). 1991 ;16(3)(Suppl):S125-33.
26. Defino HL, Canto FR. Low thoracic and lumbar burst fractures: radiographic and functional outcomes. Eur Spine J. 2007 ;16(11):1934–43. Epub 2007 Jun 14.
27. Lakshmanan P, Jones A, Mehta J, Ahuja S, Davies PR, Howes JP. Recurrence of kyphosis and its functional implications after surgical stabilization of dorsolumbar unstable burst fractures. Spine J. 2009 ;9(12):1003–9. Epub 2009 Oct 9.
28. Haschtmann D, Stoyanov JV, Gédet P, Ferguson SJ. Vertebral endplate trauma induces disc cell apoptosis and promotes organ degeneration in vitro. Eur Spine J. 2008 ;17(2):289–99. Epub 2007 Oct 10.
Copyright 2014 by The Journal of Bone and Joint Surgery, Incorporated