university of groningen thoracolumbar spinal fractures leferink, … · 2016. 3. 7. · fractures:...

$: University of Groningen Thoracolumbar spinal fractures Leferink, … · 2016. 3. 7. · fractures: thirty percent of all B -type fractures are initially unrecognised Published European$
University of Groningen

Thoracolumbar spinal fracturesLeferink, Vincentius Johannes Maria

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2002

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Leferink, V. J. M. (2002). Thoracolumbar spinal fractures: aspects of epidemiology, classification,radiological results and functional outcome = Thoracolumbale wervelfracturen. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 19-12-2020

https://www.rug.nl/research/portal/en/publications/thoracolumbar-spinal-fractures(86f4a526-f5e8-4b89-a6db-427ef05be188).html



1

THORACOLUMBAR SPINAL FRACTURES

ASPECTS OF EPIDEMIOLOGY, CLASSIFICATION, RADIOLOGICAL RESULTS,

AND FUNCTIONAL OUTCOME

VJM LEFERINK

2

Leferink, V.J.M. Thoracolumbar spinal fractures: aspects of epidemiology, classification, radiological results and functional outcome Thesis Groningen-with references-with summary in Dutch ISBN 90-367-1662-4 Copyright 2002 V.J.M. Leferink All rights reserved. No part of this book may be reproduced or transmitted, in any form or by any means, without written permission of the author. Cover photograph by V.J.M. Leferink

3

RIJKSUNIVERSITEIT GRONINGEN

THORACOLUMBAR SPINAL FRACTURES ASPECTS OF EPIDEMIOLOGY, CLASSIFICATION,

RADIOLOGICAL RESULTS, AND FUNCTIONAL OUTCOME

THORACOLUMBALE WERVELFRACTUREN

ASPECTEN VAN DE EPIDEMIOLOGIE, CLASSIFICATIE, RADIOLOGISCHE EN FUNCTIONELE RESULTATEN

Proefschrift

ter verkrijging van het doctoraat in de Medische Wetenschappen

aan de Rijksuniversiteit Groningen op gezag van de

Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op

woensdag 27 november 2002 om 14.15 uur

door Vincentius Johannes Maria Leferink

geboren op 1 maart 1959 te Herwen en Aerdt

4

Promotor: Prof.dr. H.J. ten Duis Referent: Dr. E.F.M. Veldhuis Beoordelingscommissie: Prof.dr. R. van Schilfgaarde Prof.dr. H.J.Th.M. Haarman Prof.dr. A.J. Verbout Paranimfen: Drs. H.S. Hofker Dr. K.W. Wendt

Voor Anouk en Nicole

THORACOLUMBAR FRACTURES CONTENTS

5

Contents

1. General introduction

9

2. Classificational problems in ligamentary distraction type vertebral fractures: thirty percent of all B-type fractures are initially unrecognised Published European Spine Journal (2002) 11:3:246-250

31

3. Thoracolumbar spinal fractures: radiological results of transpedicular fixation combined with transpedicular cancellous bone graft and posterior fusion in 183 patients

Published European Spine Journal (2001) 10:6:517-523

41

4. Burstfractures of the thoracolumbar spine: changes of the spinal canal during operative treatment and follow-up Accepted European Spine Journal (2002)

55

5. Thoracolumbar spinal fracture: segmental range of motion after dorsal spondylodesis in eighty-two patients, a prospective study Published European Spine Journal (2002) 11:1:2-7

69

6. Functional outcome after 3-8 years in 19 patients with thoracolumbar fractures, treated with dorsal instrumentation and fixation, transpedicular cancellous bone grafting and dorsal spondylodesis Accepted European Spine Journal (2002)

81

7. General discussion, recent, present and future developments

95

8. Abstracts

107

9. Nederlandse samenvattingen

113

10. Dankwoord

119

11. Curriculum vitae

123

12. List of abbreviations

125

THORACOLUMBAR FRACTURES 1 GENERAL INTRODUCTION

9

Chapter 1

General introduction History The first written proof of treatment of spinal fractures was found in the Smith papyrus rolls originating about 1550 years before Christ. Highly specialised doctor-priests took care of the patients with spinal fractures. They performed wound treatment, put on bandages and subscribed rest in the horizontal position [30]. In these days probably only open fractures and fractures with considerable kyphotic deformity were recognized. Hippocrates distinguished spinal fractures with and without neurological deficit. Patients with paralysis would die. Spinal fractures without paralysis were treated by distraction, manual reduction, and rest in supine position [30]. Special tables were designed and used for these treatments by Hippocrates and Oribasius (Fig.1 and Fig.2) [38].

Fig.1 Reduction table used in non operative treatment

of spinal fractures and dislocations by Hippocrates [38]


10

Even some kinds of operative treatments were described, but it is uncertain whether those operations actually were performed. Laminectomy and removal of the narrowing fracture part in case of paralysis was suggested as soon as the 7th century by Paul of Aegina. However, the statement of John Bell: The cutting into a fractured vertebra is a dream (1799), probably provides a better idea of the possibilities in the middle ages. Conservative treatment was the golden standard for a long time. Malgaigne (1847) and Böhler (1932) advocated indirect manipulative anatomical reduction by longitudinal traction and hyperlordosis, immobilisation in a plaster jacket, followed by intensive muscle training (physiotherapeutics avant la lettre) [30].

Fig.2 Table used for nonoperative treatment by Oribasius [38] Nicoll advised a completely functional treatment after he evaluated the treatment of miners with spinal fractures. The anatomical result of the treatment was rather poor, but Nicoll stated that most of the miners had perfect results, because they could perform the heavy mining job, showing … their ability to withstand the conditions of stress that arise in working at the coal-face in cramped positions … without discomfort [31]. One should realise that the first radiographs of the spinal column were only available since 1925, because much higher voltages were needed for trunk radiographs than for extremity radiographs. All earlier remarks about fractures, dislocations and reduction have to be considered reflections of gross distortions of the normal anatomy or findings in post mortal dissection (Fig.3).


11

The history of spinal fracture treatment is described in a concise way by Memmert in Die Wirbelsäule in der Anschauung [30]. Most of the described history is derived from this book, unless other references are mentioned. After 1970 improvements in (radio-)diagnostics, safer anesthesiological techniques, improved intensive care and development of more reliable implants were the prerequisites for further development of operative techniques.

Fig.3 Gross distortions of the normal anatomy Operative treatment of spinal fractures Like in other fields in traumatology stable fixation of fractures and early movement was advocated in spinal fractures since 1960. The idea was that most of the compression type fractures (burst fractures) of the thoracolumbar transition could either be treated by a ventral fusion operation or by a dorsal transpedicular operation.

Fig.4 Harrington rod with laminar hooks


12

With retroperitoneal, transthoracic or transperitoneal access a direct ventral reduction and reconstruction with plate osteosynthesis was performed. In the dorsal procedure fracture reduction was achieved by antikyphosing and dorsal compression. This indirect reduction was completed by ways of stabilization known from scoliosis treatment. Both procedures were frequently combined with decompression corporectomy or laminectomy. A positive effect of these additional procedures could never be proved [2]. After the application of dorsal long segment fixation (5 segments) with Weiss springs, Roy-Camille plates, Harrington rods (Fig.4), or with other implants used in scoliosis treatment, shorter segment fixation systems were developed to treat spinal fractures: the external fixator (Magerl) and internal fixators (Dick, Olerud) [12;22;27;32;40;57]. Also new anterior systems were developed: Kostuik, Dunn, Kaneda [10;14;23] and many other implants. History of treatment of spinal fractures in the University Hospital Groningen Harrington rods Operative treatment including surgical stabilization with Harrington’s instruments was practised in the University Hospital Groningen between 1979 and 1987 [6;12]. The impact of this treatment on the anatomy was evaluated radiologically (plain radiographs and conventional tomography) and the findings are described in the thesis of one of the co-authors of several chapters of this thesis [51]. In those days preoperative halobitibial traction was used for reduction in patients with spinal fractures (Fig.5).

5a 5b

Fig.5 Halobitibial traction preceded operative treatment (5a halo; 5b bitibial pins)


13

In more than 80 percent of the operatively treated patients halobitibial traction preceded operative treatment. The mean duration of the traction was 15 days. The complications associated with halobitibial traction and the operative procedure were numerous. Halobitibial traction was frequently associated with pin tract infection at the tibia (10 percent). In two out of 49 patients the traction had to be discontinued for psychological reasons. One of these patients was psychiatrically treated before. Reported complications related to the operative treatment with Harrington rods were infection, implant failure or hook dislocation, bleeding and hyperpathy of the lower dermatomes (cumulative percentage of complications 20 percent) [51]. Internal fixation according to Dick [8] The "fixateur interne" was presented in 1987 as a new device for posterior spinal fracture surgery. This device was derived from the Magerl external short segment spinal fixator [27]. It consists of four long Schanz screws, which are inserted from a posterior approach through the pedicles into the adjacent vertebral bodies, and of two connecting threaded longitudinal rods, carrying mobile clamps that can be fixed in the right position by nuts (Fig.6). The long lever arms of the Schanz screws enable manual reduction (Fig.7). The lever arms are shortened with a steel cutter at the end of the operation (Fig.8).

Fig.6 Dick’s internal fixator in model


14

As the device is stable against flexion and rotation by itself, the fixation can be restricted to the immediately adjacent vertebral bodies of a lesion, leaving the rest of the spine mobile. In fracture treatment the instrumentation is combined with a direct repair of the anterior loss of bone stock by a transpedicular bone grafting procedure from the same dorsal approach [5]. Instead of the unilateral transpedicular cancellous bone graft, as advocated by Daniaux, we have performed a bilateral cancellous bone graft in order to deposit as much bone in the reduced vertebral body as possible. A dorsolateral spondylodesis was done with the rest of the cancellous bone graft at the level of destructed endplates; for example the upper segment in incomplete burst fractures, and both segments in complete burst fractures. A standard treatment and follow up protocol was used. Only minor changes in treatment were allowed. Therefore the treatment of patients in 1988 was very similar to the treatment in 1997.

Fig.7 Acting mechanism and applied antikyphosing forces (1) in the internal fixator. Fixation of nut with the same instrument (2)

In February 1996, in our hospital the Dick fixator was replaced by the stainless steel Universal Spinal System (USS) derived from the Dick system. Since 1998, titanium USS-implants are used.


15

Fig.8 Universal Spine System USS (titanium) with shortened Schanz screws

Since 1988 over 350 patients with a spinal fracture of the thoracolumbar junction have been operated according to this system in our hospital. All three implant systems act according to the same principles (Fig.7). Rehabilitation and follow up Much attention has to be paid to correct rehabilitation and psychological support. In our hospital the intensive collaboration with the colleagues in the nearby rehabilitation centre Beatrixoord has proved to be very valuable. In the postoperative period, the patients were treated by physical therapy in bed during two weeks. Afterwards they were mobilised in a simple thoracolumbar reclination orthesis until nine months after the operation (Fig.9). At nine months the hardware was removed. Afterwards the temporarily fixed segment has the opportunity to regain its mobility. If the spondylodesis was bridging the same levels as the internal fixator no hardware removal was done. All patients and their radiographs were examined at regular intervals (at 3, 6, 9, 12 and 24 months). Epidemiology Descriptives and epidemiological changes in the period 1970-1999 The group of operatively treated patients in this study forms only a small part of all patients with spinal fractures admitted to our hospital. Most patients with thoracolumbar spinal fractures are treated conservatively.


16

Fig.9 Thoracolumbar orthesis Comprehensive epidemiological data of trauma patients with spinal fractures are scarce [4]. Many epidemiological reports about spinal fractures focus on osteoporosis as an etiologic factor [3;15]. Studies concerning only operatively treated patients with spinal fractures show selective and biased data that might be useful for capacity planning in hospitals or evaluating results of operative treatment, but not for epidemiological purposes [21;24]. Epidemiological evaluation of subgroups like pediatric cervical spine injuries, spinal fractures in aviators, sports-related spinal injuries give important information, but only about these subgroups, for example in the field of risk evaluation or specific aftertreatment [1;16;37;41]. In our hospital 10.000 new trauma patients are treated each year. The yearly number of patients has been relatively constant in the last three decades. About 200 polytrauma patients (ISS>16) are submitted to the emergency trauma unit of our hospital, and 50-70 more are transferred after initial treatment in other hospitals. The adherence area of our hospital for polytrauma patients, acetabular fractures and spinal fractures amounts to about 2 million people. RLOG Hospital registries of all primary visits of trauma patients like RLOG (Registratie van Letsels en Ongevallen Groningen/Registration of traumatic lesions and accidents Groningen) showed to be relevant in supplying the data to analyse non-


17

fatal injuries [33]. Since 1970 the data of more than 300,000 trauma patients have been included in the RLOG database. The RLOG data survey revealed a slight increase in numbers of patients with spinal fractures during the last three decades (Fig.10). The most frequently diagnosed single spinal fracture is the thoracolumbar fracture without neurological deficit (Table 1 and Table 2). The high prevalence rates in 20-29 years old men represent the young male peak in general in epidemiological studies in traumatology [17]. In patients with spinal fractures of the thoracolumbar transition the observed male/female ratio is 2.5 at age 20-29 and shows inversion at the age of 60. This inversion probably is the result of more pronounced osteoporosis in women as well as their higher life expectancy. All our patients are registered in the trauma registry RLOG according to the International Classification of Disease, version 8, later version 9 (ICD-8 and ICD-9).

0

2 0

4 0

6 0

8 0

100

120

140

7 0 7 1 7 2 7 3 7 4 7 5 7 6 7 7 7 8 7 9 8 0 8 1 8 2 8 3 8 4 8 5 8 6 8 7 8 8 8 9 9 0 9 1 9 2 9 3 9 4 9 5 9 6 9 7 9 8 9 9

Year

n

Total Linear

Fig.10 Number of patients per year with spinal fractures treated in the University

Hospital Groningen between 1970 and 1999 (n=3003) and the linear trend in 30 years


18

Table 1 Number of patients, with monolevel (single) spinal fractures, divided according to level, age-group, gender, and neurological deficit in the period 1970 to 1999 (n=2713)

*without neurological deficit, ** with neurological deficit

Age 0 - 9 10 - 19 20 - 29 30 - 39 40 – 49 50 - 59 60 - 69 70 - 79

80 - 89 >90 Total

Level M F M F M F M F M F M F M F M F M F M F M F

Cervical* 6 1 38 21 100 21 57 20 45 20 25 12 26 16 10 12 12 5 4 0 323 128

Th.lumb* 11 9 114 74 262 107 192 76 139 86 130 98 86 94 58 103 17 49 1 9 1010 705

Sacral* 2 1 10 7 39 22 15 20 29 14 11 10 3 7 5 10 2 6 0 1 116 98

Cervical** 2 1 23 3 24 3 18 7 4 7 9 5 7 5 6 1 2 2 1 0 96 34

Th.lumb** 0 1 12 1 29 7 23 6 11 5 16 4 5 5 5 4 4 1 0 1 105 35

Sacral** 0 0 6 6 8 15 4 8 5 4 1 4 0 2 0 0 0 0 0 0 24 39

Total 21 13 203 112 462 175 309 137 233 136 192 133 127 129 84 130 37 63 6 11 1674 1039

Table 2 Causes of accident in patients with (single) spinal fractures, with and without neurological deficit

Spinal fracture

without neurological deficit

Spinal fracture with neurological

deficit

E-code Level Cause

Cervical Thoraco Lumbar Sacral Cervical

Thoraco lumbar Sacral Total

Private 106 765 93 49 53 32 1098 Sports 24 109 7 4 12 3 159 Violence 4 12 4 0 1 3 24 Traffic 112 293 53 29 25 4 516 Work 2 59 6 4 5 1 77 Self inflicted 4 61 6 3 9 1 84 Unknown 199 416 45 41 35 19 755 Total 451 1715 214 130 140 63 2713

The causes of accident, as registered in the RLOG as E-codes of the ICD, show that most accidents leading to spinal fractures occur in the private environment (Table 2). This includes fall from a height and hit by falling object. Traffic accidents are an important cause as well (Table 2). The percentage of patients with neurological deficit is 22 in the cervical and sacral group, but only 8 in the thoracolumbar group (Table 2). ICD-codes revealed Abbreviated Injury Scale (AIS) scores and these were cumulated to ISS in case of multiple injuries [18]. ISS calculation was performed by means of a Pascal computer program in all patients [18]. The Injury Severity


19

Score (ISS) in monotrauma patients with spinal fractures can only be 4, 9, 16, or 25. Lesions to other AIS-body regions and multiple spinal fractures can result in higher (cumulative) ISS-scores, but also in scores between the above mentioned scores.

Table 3 ISS and number of patients with monolevel (mono) or multilevel (multiple)

spinal fractures, with or without other injuries (n=3003)

Spinal fractures Spinal fractures and other injuries ISS

Mono Multiple Mono Multiple n

4 178 4 76 1 259 5 0 0 53 1 54 6 0 0 18 2 20 8 0 0 57 4 61 9 926 39 101 2 1068 10 0 0 131 5 136 11 0 0 30 1 31 12 0 0 22 0 22 13 0 15 281 27 323 14 0 0 70 12 82 16 75 4 27 1 107 17 0 0 118 11 129 18 0 35 79 14 128 19 0 0 23 7 30 20 0 0 34 3 37 21 0 0 6 1 7 22 0 1 106 39 146 24 0 0 14 1 15 25 1 5 37 5 48 26 0 0 4 0 4 27 0 0 55 18 73 29 0 0 40 10 50 30-39 0 0 29 11 40 40-49 0 0 4 1 5 50-75 0 0 0 1 1 Unknown 53 0 65 9 127 Total 1233 103 1480 187 3003


20

The mean ISS did not change in the studied period (ISS 12), but the percentage of patients that is treated in the outpatient clinic only increased from 60% to more than 70%. Of all spinal fracture patients in 30 years 27.3 percent had an ISS >16 (Table 3). Diagnostics, classification and indications for operative treatment Diagnostics The mechanism and the impact of the trauma, in combination with the complaints of the patient generally give rise to the suspicion of a spinal fracture. Mechanisms are fall or jump from a height, deceleration trauma and direct blow, for example in traffic accidents or sports, but also in the working or private environment. The probability of fractures in minor trauma increases in pathological conditions of the spinal column. Acute back pain, mild or severe, and, in a minority of patients, sensory or motor loss, wounds of the back, or palpable deformities or gaps, add to the suspicion. Plain radiographs of the thoracic spine and the lumbar spine in two directions are the initial forms of imaging for the vast majority of patients. These radiographs are more helpful in proving the existence of a spinal fracture than in excluding it. Special attention should be given to the thoracolumbar junction. Often extra radiographs have to be made of this junction to avoid misinterpretation of lesions at the edge of the beam. Before the early nineties conventional tomography was performed to assess stability and to classify the fracture [28]. Since the eighties computer tomography, including digital sagittal reconstructions, gradually replaced the conventional tomography. Classification A classification should provide, in a descriptive and concise way, information about the severity of the lesion and should give an identification of any lesion by means of a (rather) simple algorithm. In daily use it should give guidance to the form of treatment in relation to prognosis. After classifications of Nicoll [31], Holdsworth [13], Whitesides [58], Louis [25], and Roy-Camille [39] the three column classification of Denis was widely accepted [7]. Further developments by McAfee [29], Ferguson and Allen [11], influenced by the increasing possibilities of CT, lead to the comprehensive classification of Magerl et al [26]. This classification is based on the pathomorphological characteristics of the injury. Three (supposed) mechanisms of injury, of which the effect is shown in the radiographs and CT scans, give name to the three main categories: A=compression, B=distraction and C=rotation type fractures. In each group


21

morphological criteria lead to further classification in (three) subgroups, et cetera (Table 4).

Table 4 Comprehensive Classification, groups and subgroups

A1.1 Endplate impaction A1.2 Wedge impaction A1 Impaction fracture A1.3 Vertebral body collapse A2.1 Sagittal split fracture A2.2 Coronal split fracture A2 Split fracture A2.3 Pincer fracture A3.1 Incomplete burst fracture A3.2 Burst-split fracture

A Compression injury

A3 Burst fracture A3.3 Complete burst fracture B1.1 With disc disruption B1 Posterior ligamentary

lesion B1.2 With type A fracture B2.1 Transverse bicolumn B2.2 With disc disruption

B2 Posterior osseous lesion

B2.3 With type A fracture B3.1 With subluxation B3.2 With spondylolysis

B Distraction injury

B3 Anterior disc rupture B3.3 With posterior dislocation C1.1 Rotational wedge fracture C1.2 Rotational split fracture C1 Type A with rotation C1.3 Rotational burst fracture C2.1 B1 lesion with rotation C2.2 B2 lesion with rotation C2 Type B with rotation C2.3 B3 lesion with rotation C3.1 Slice fracture

C Rotation injury

C3 Rotational shear injury C3.2 Oblique fracture

In general A type fractures are less severe than B type fractures, and B type fractures less severe than C type fractures (Fig.11). In the subgroups the lesions are ranked as well. For example: an incomplete burst fracture is classified as A3.1, but in case of additional dorsal ligament rupture (caused by distraction) as B1.2. Note that isolated spinal process or transverse process fractures are not included in the classification [26]. We started to use the Magerl (AO) classification in 1994 because it allows categorization of injuries to all relevant parts of the spine. In this study we


22

classified all fractures according to this classification, including the fractures of the patients treated between 1988 and 1994.

Fig.11 Comprehensive Classification: A type fractures (compression type), B type fractures (distraction type) and C type fractures (rotation type)


23

To our opinion, the classification scheme will have to be revised. Based on additional MRI information, it will be clear which kinds of soft-tissue injuries should be considered indicative of various types and subgroups [34-36]. Indications for operative treatment In the last decade the following fracture types formed an indication for operative treatment: type A fractures (Fig.11): A1.2 fractures, only when the wedge angle of the body is more than 15 degrees, A2.3 fractures and A3 fractures (Table 4). B and C type fractures are usually operated, although in some cases there might be reasons to prescribe conservative treatment. Age, degree of dislocation of the fracture, medical history and other injuries are other important factors in decision making. Outcome Desperation when the neck or the back has broken is still a common reaction of the patient and his family. The public generally accepts that a spinal injury causes death or at least paralysis or paraplegia. Though Trickey stated in 1976 that spinal fractures should be treated in the same way as other fractures and dislocations and that nearly always there will be a complete recovery [42]. Many studies focus on the radiological outcome, but like in other fractures functional outcome does not necessarily correlate with the radiological result. The neurological sequelae strongly influence the functional outcome; however the treatment modality has only a marginal influence on the neurological outcome. An important outcome study is the meta analysis of Dickman [9]. In his study the performance of four different groups of operative treatment was compared, i.e. pedicle screw fixation devices, hook-rod systems, Luque instrumentation, and anterior instrumentation devices. Journal articles between 1975 and 1993 were evaluated for sufficient clinical and radiographical outcome data. One hundred and sixty-eight articles on thoracolumbar fractures were studied of which 110 were rejected after review. Fifty-eight articles could be included in the meta-analysis. Fusion and residual pain were outcome measures. They were similar in different groups and seemed not to be the problem in either treatment. The complication rate was similar as well, up to 27 percent! The higher loss of fixation in hook-rod systems (9.5%) compared to fixation problems in pedicle screws (3.3%) was statistically significant. In a more recent study it seems that the dorsal approach is accompanied with less complications than the anterior approach (14.1% versus 29.7%) [19]. In our series of 183 operatively treated patients between 1988 and 1996 we did not observe any deep infection nor


24

neurological complications, but we noticed three superficial infections after implant removal. One major complication occurred: intraoperatively a retractor hook was placed behind a rib and this caused an intrathoracal bleeding complication that at first was misdiagnosed. The patient died of haemorrhagic shock despite thoracotomy. Influence of insurance and social security on the outcome It is generally accepted that minor impairments may have major consequences in working ability and sports leisure. It is logical that this accounts for major impairments as well. In this respect the results of Nicoll concerning miners with conservatively treated spinal fractures and their return to work status can only be seen as an illustration of the consequences of injury and treatment in those days [31]. Even in huge multicentre studies like the study performed by the Working Group on Spine of the German Trauma Association it is very difficult to draw general conclusions about specific therapies, for example anterior and posterior techniques in relation to outcome and work status [20]. In the Netherlands it takes about a year before it will be clear whether a patient can return to work and about two years before the final assessment about definite inability to work and social security benefits can be made in severely or multiple injured patients [49]. The research history in the Department of Surgery/Traumatology of the University Hospital Groningen, concerning outcome in trauma and in thoracolumbar fractures facilitates the performance of research in this field. A thesis on the evaluation of the conventional radiological examination of thoracolumbar fractures was published in 1988 [28]. The formerly mentioned study on the results of the operative treatment of thoracolumbar fractures with Harrington rods was published in 1993 [51]. Many studies on the evaluation of short-term and long-term outcome in (major) trauma were published in collaboration with the Department of Rehabilitation [43-50]. Studies on the effects of deceleration trauma of the neck (whiplash) were published in collaboration with the Departments of Anaesthesiology and Psychology [52-56]. Outline of this thesis Chapter 1 provides information about the history of spinal trauma and the conservative and operative treatments, about the epidemiology of the study populations in this thesis, diagnostic procedures, classifications, operative procedures, rehabilitation, and the follow up schemes, as applied to the patients described in this study.


25

With respect to the new developments in the field of diagnostics and classifications of spinal fractures the following questions will be answered in Chapter 2: • What percentage of B type fractures are initially unrecognised and regarded

as A type fractures? • Which are characteristic radiological qualities of initially unrecognised B type

fractures? In the field of results of the operative treatment of spinal fractures we studied radiological data in order to answer the following questions in Chapter 3: • Does the vertebral body collapse after removal of the implants, despite

transpedicular bone grafting? • Is correction of the regional angle maintained after surgery? • And if not, does the loss of the intervertebral angle contribute to this change? • What is the influence of implant failure on the radiological measurements? The second set of questions concerning radiological results will be answered in Chapter 4: • Are there any changes of the posterior segmental height (PSH) during

operation and in the course of further treatment? • Do the bony parts in the spinal canal in plain radiographs disappear in the

course of treatment, and how many patients have developed a normal width of the spinal canal two years after the initial treatment?

• Are there changes of the midsagittal diameter of the spinal canal as measured in CT-slices?

The third set of questions about radiological outcome concerns the functional aspects of remaining mobility after operative treatment in Chapter 5: • Does dorsal spondylodesis of one segment also cause loss of range of motion

(ROM) at other segments or does it result in increased ROM at the surrounding segments?

• Does dorsal spondylodesis result in ankylosis of the affected intervertebral disc space?

Finally functional outcome tests and questionnaires try to reveal the answers to the following questions in Chapter 6: • Which are the impairments in operatively treated patients with

a thoracolumbar burst fracture? • What is their ability to perform in activities of daily life? • What is their return to work status and quality of life?


26

In the Chapters 7, 8, and 9 we summarize our findings in a general discussion focusing on the consequences for future management of spinal fractures. We point out ongoing and future research in the field of spinal fractures. References 1. Belmont P-JJ, Taylor KF, Mason KT, Shawen SB, Polly D-WJ, Klemme WR (2001)

Incidence, epidemiology, and occupational outcomes of thoracolumbar fractures among U.S. Army aviators. J Trauma 50:855-861

2. Boerger TO, Limb D, Dickson RA (2000) Does 'canal clearance' affect neurological outcome after thoracolumbar burst fractures? J Bone Joint Surg Br 82:629-635

3. Cooper C, Atkinson EJ, Kotowicz M, O'Fallon WM, Melton LJ (1992) Secular trends in the incidence of postmenopausal vertebral fractures. Calcif Tissue Int 51:100-104

4. Cooper C, Atkinson EJ, O'Fallon WM, Melton LJ (1992) Incidence of clinically diagnosed vertebral fractures: a population-based study in Rochester, Minnesota, 1985-1989. J Bone Miner Res 7:221-227

5. Daniaux H (1986) Transpedikuläre Reposition und Spongiosaplastik bei Wirbelkörperbrüchen der unteren Brust- und Lendenwirbelsäule. Unfallchirurg 89:197-213

6. Dekutoski MB, Conlan ES, Salciccioli GG (1993) Spinal mobility and deformity after Harrington rod stabilization and limited arthrodesis of thoracolumbar fractures. J Bone Joint Surg Am 75:168-176

7. Denis F (1983) The three column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine 8:817-831

8. Dick W (1987) The "fixateur interne" as a versatile implant for spine surgery. Spine

12:882-900 9. Dickman CA, Yahiro MA, Lu HT, Melkerson MN (1994) Surgical treatment

alternatives for fixation of unstable fractures of the thoracic and lumbar spine. A meta-analysis. Spine 19:2266S-2273S

10. Dunn HK (1984) Anterior stabilization of thoracolumbar injuries. Clin Orthop 189:116-124

11. Ferguson RL, Allen-BL J (1984) A mechanistic classification of thoracolumbar spine fractures. Clin Orthop 189:77-88


27

12. Harrington PR (1988) The history and development of Harrington instrumentation. By Paul R. Harrington, 1973. Clin Orthop 227:3-5

13. Holdsworth FW (1963) Fractures, dislocations, and fracture-dislocations of the spine. J Bone Joint Surg Br 45:6-20

14. Kaneda K, Taneichi H, Abumi K, Hashimoto T, Satoh S, Fujiya M (1997) Anterior decompression and stabilization with the Kaneda device for thoracolumbar burst fractures associated with neurological deficits. J Bone Joint Surg Am 79:69-83

15. Kannus P, Niemi S, Palvanen M, Parkkari J (2000) Continuously increasing number and incidence of fall-induced, fracture-associated, spinal cord injuries in elderly persons. Arch Intern Med 160:2145-2149

16. Katoh S, Shingu H, Ikata T, Iwatsubo E (1996) Sports-related spinal cord injury in Japan (from the nationwide spinal cord injury registry between 1990 and 1992). Spinal Cord 34:416-421

17. Kingma J (1994) The young male peak in different categories of trauma victims. Percept Mot Skills 79:920-922

18. Kingma J, ten Vergert E, Werkman HA, ten Duis HJ, Klasen HJ (1994) A Turbo Pascal program to convert ICD-9CM coded injury diagnoses into injury severity scores: ICDTOAIS. Percept Mot Skills 78:915-936

19. Knop C, Bastian L, Lange U, Oeser M, Zdichavsky M, Blauth M (2002) Complications in surgical treatment of thoracolumbar injuries. Eur Spine J Springer link DOI 10.1007/s00586-001-0382-6

20. Knop C, Blauth M, Bühren V, Arand M, Egbers HJ, Hax PM, Nothwang J, Oestern

HJ, Pizanis A, Roth R, Weckbach A, Wentzensen A (2001) Operative Behandlung von Verletzungen des thorakolumbalen Übergangs-Teil 3: Nachuntersuchung. Ergebnisse einer prospektiven multizentrischen Studie der Arbeitsgemeinschaft "Wirbelsäule" der Deutschen Gesellschaft für Unfallchirurgie. Unfallchirurg 104:583-600

21. Knop C, Fabian HF, Bastian L, Blauth M (2001) Late results of thoracolumbar

fractures after posterior instrumentation and transpedicular bone grafting. Spine 26:88-99

22. Korovessis PG, Baikousis A, Stamatakis M (1997) Use of the Texas Scottish Rite Hospital instrumentation in the treatment of thoracolumbar injuries. Spine 22:882-888


28

23. Kostuik JP (1984) Anterior fixation for fractures of the thoracic and lumbar spine with or without neurologic involvement. Clin Orthop 189:103-115

24. Leferink VJM, Zimmerman KW, Veldhuis EFM, ten Vergert EM, ten Duis HJ (2001) Thoracolumbar spinal fractures: radiological results of transpedicular fixation combined with transpedicular cancellous bone graft and posterior fusion in 183 patients. Eur Spine J 10:517-523

25. Louis R (1977) Fractures instables du rachis. III. L'instabilite. A. Les theories de l'instabilite. Rev Chir Orthop.Reparatrice Appar Mot 63:423-425

26. Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S (1994) A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 3:184-201

27. Magerl FP (1984) Stabilization of the lower thoracic and lumbar spine with external skeletal fixation. Clin Orthop 189:125-141

28. Martijn A (1988) Thoracale en lumbale wervelfracturen: een evaluatie van conventionele röntgendiagnostiek. 1-139. Thesis Rijksuniversiteit Groningen.

29. McAfee PC, Yuan HA, Fredrickson BE, Lubicky JP (1983) The value of computed tomography in thoracolumbar fractures. An analysis of one hundred consecutive cases and a new classification. J Bone Joint Surg Am 65:461-473

30. Memmert M (1999) Ein Versuch, die Geschichte der Wirbelsäulenchirurgie zu

umreissen. In: Memmert M, Memmert G (eds) Die Wirbelsäule in der Anschauung. Spurensuche in Kunst, Geschichte und Sprache. Springer-Verlag, Berlin Heidelberg, pp 247-270

31. Nicoll EA (1949) Fractures of the dorso-lumbar spine. J Bone Joint Surg Br

31:376-394

32. Olerud S, Karlstrom G, Sjostrom L (1988) Transpedicular fixation of thoracolumbar vertebral fractures. Clin Orthop 227:44-51

33. Oskam J, Kingma J, Klasen HJ (1994) The Groningen trauma study: injury patterns in a Dutch trauma centre. Eur J Emergency Med 1:167-172

34. Öner FC, van Gils AP, Dhert WJ, Verbout AJ (1999) MRI findings of thoracolumbar spine fractures: a categorisation based on MRI examinations of 100 fractures. Skeletal Radiol 28:433-443

35. Öner FC, van der Rijt RR, Ramos LM, Dhert WJ, Verbout AJ (1998) Changes in the disc space after fractures of the thoracolumbar spine. J Bone Joint Surg Br 80:833-839


29

36. Öner FC, van Gils APG, Faber JAJ, Dhert WJA, Verbout AJ (2002) Some complications of common treatment schemes of thoracolumbar spine fractures can be predicted with magnetic resonance imaging. Spine 27:629-636

37. Patel JC, Tepas JJ, Mollitt DL, Pieper P (2001) Pediatric cervical spine injuries: defining the disease. J Pediatr Surg 36:373-376

38. Richter AL (1828) Brüche und Verrenkungen der Knochen. Tafel XXX. Theodor Christ Friedr Enslin. Berlin

39. Roy-Camille R, Saillant G (1984) Les traumatismes du rachis sans complication neurologique. Int Orthop 8:155-162

40. Roy-Camille R, Saillant G, Berteaux D, Salgado V (1976) Osteosynthesis of thoraco-lumbar spine fractures with metal plates screwed through the vertebral pedicles. Reconstr Surg Traumat 15 :2-16

41. Shingu H, Ohama M, Ikata T, Katoh S, Akatsu T (1995) A nationwide epidemiological survey of spinal cord injuries in Japan from January 1990 to December 1992. Paraplegia 33:183-188

42. Trickey EL (1976) Injuries of the Spine. In: Wilson JN (ed) Watson-Jones Fractures

and Joint Injuries. Churchill Livingstone, Edinburgh London and New York, pp 798-849

43. van der Sluis CK, Eisma WH, Groothoff JW, ten Duis HJ (1998) Long-term physical,

psychological and social consequences of a fracture of the ankle. Injury 29:277-280

44. van der Sluis CK, Eisma WH, Groothoff JW, ten Duis HJ (1998) Long-term physical, psychological and social consequences of severe injuries. Injury 29:281-285

45. van der Sluis CK, Geertzen JH, Werkman HA, ten Duis HJ (1994) Epidemiologische gegevens van ernstig gewonden; een retrospectief onderzoek over de periode 1985-1989. Ned Tijdschr Geneeskd 138:2285-2290

46. van der Sluis CK, Kingma J, Eisma WH, ten Duis HJ (1997) Pediatric polytrauma: short-term and long-term outcomes. J Trauma 43:501-506

47. van der Sluis CK, Klasen HJ, Eisma WH, ten Duis HJ (1996) Major trauma in young and old: what is the difference? J Trauma 40:78-82

48. van der Sluis CK, Kucey DS, Brenneman FD, Hunter GA, Maggisano R, ten Duis HJ (1997) Long-term outcomes after upper limb arterial injuries. Can J Surg 40:265-270


30

49. van der Sluis CK, ten Duis HJ, Geertzen JH (1995) Multiple injuries: an overview of the outcome. J Trauma 38:681-686

50. van der Sluis CK, Timmer HW, Eisma WH, ten Duis HJ (1997) Outcome in elderly injured patients: injury severity versus host factors. Injury 28:588-592

51. Veldhuis EFM (1993) De behandeling van ernstig instabiele letsels van de thoracale en lumbale wervelkolom op het Strykerframe met halobitibiale tractie en Harringtonstabilisatie. Thesis Rijksuniversiteit Groningen. pp 1-135.

52. Versteegen GJ, Kingma J, Meijler WJ, ten Duis HJ (1998) Neck sprain in patients injured in car accidents: a retrospective study covering the period 1970-1994. Eur Spine J 7:195-200

53. Versteegen GJ, Kingma J, Meijler WJ, ten Duis HJ (1998) Neck sprain not arising from car accidents: a retrospective study covering 25 years. Eur Spine J 7:201-205

54. Versteegen GJ, Kingma J, Meijler WJ, ten Duis HJ (2000) Neck sprain after motor

vehicle accidents in drivers and passengers. Eur Spine J 9:547-552

55. Versteegen GJ, Kingma J, ten Duis HJ (2001) Sprain of the neck in clinically treated patients in The Netherlands: an inventory of different categories of car accidents. Percept Mot Skills 92:655-658

56. Versteegen GJ, van Es FD, Kingma J, Meijler WJ, ten Duis HJ (2001) Applying the Quebec Task Force criteria as a frame of reference for studies of whiplash injuries. Injury 32:185-193

57. Weiss M (1975) Dynamic spine alloplasty (spring loading corrective devices) after fracture and spinal cord injury. Clin Orthop 112:150-158

58. Whitesides TEJ (1977) Traumatic kyphosis of the thoracolumbar spine. Clin Orthop 128: 78-92

THORACOLUMBAR FRACTURES 2 CLASSIFICATIONAL PROBLEMS

31

Chapter 2 Classificational problems in ligamentary distraction type vertebral fractures: 30% of all B-type fractures are initially unrecognised V.J.M. Leferink, E.F.M. Veldhuis, K.W.Zimmerman, H.J. ten Duis, E.M. ten Vergert Published in European Spine Journal (2002) 3:246-250 © Springer

Introduction Judgement about instability of spinal fractures and fracture dislocations is a continuing field of research. Although many authors have pointed out that lesions should be divided into stable and unstable lesions, all structures contribute to stability. All lesions of structures result in a certain instability, but complete instability is rare. An important contribution to stability is given by the posterior dorsal complex, which is formed by the posterior interspinous ligament and the supraspinous ligament. It is one of the merits of the comprehensive classification (CC) that instability is now increasingly considered as a continuous scale. A-type fractures are more stable than B-type fractures, which are, in turn, more stable than C-type lesions. A1 is more stable than A2, etcetera [12]. It is difficult to find good criteria for the preoperative diagnosis of a ligamentary B-type lesion, unless there is gross dislocation or a palpable interspinous gap [12]. A preoperative A-type spinal fracture diagnosis was quite often found during operation to be an unrecognised B-type lesion. We analysed our data in order to find, in retrospect, characteristic radiological qualities of the initially unrecognised B-type fractures. All patient data, including those from patients treated before 1994, were revised according to the CC for this study.


32

Materials and methods Between 1988 and 1996 we performed an operative treatment on 160 patients with thoracolumbar A-type or B-type fractures, classified preoperatively. All fractures were treated with instrumental angular reduction, distraction and stabilisation using the Dick internal fixator [6]. Since 1995 we have used the Universal Spine System, Synthes®. The procedure was combined with transpedicular cancellous bone grafting of the vertebral body in most of the cases (according to Daniaux) [3;5] and posterior spondylodesis [3]. After revision of X-rays, CT-scans and operation records we concluded that, during operation, 17 of the 128 A-type fractures appeared to be B-type lesions after exploration of the dorsal ligaments . A total of 32 B-type lesions were identified preoperatively. Analysis of characteristics of patients with A-type fractures (without the unrecognised B-type fractures), initially unrecognised B-type (uB) fractures, and B-type fractures (without the unrecognised B-type fractures) was performed. We analysed the age of the patients, the respective fracture levels, neurological deficit, anterior wedge angles (AWA), anterior corporal height (ACH), posterior corporal height (PCH), and the percentage of frontal corporal collapse (FCC) (Fig.1). The measurements were compared using the t-test, and p<0.005 was considered significant.

Fig.1 Measurements in lateral plain X-rays: AWA anterior wedge angle, ACH anterior corporal height, PCH posterior corporal height


33

Results The mean age of the patients in each group did not show a significant difference (Table 1). The group of unrecognised B-fractures had a more caudal fracture level than the recognised B-type fractures. The fracture levels of the A-group and uB-group patients were not shown to be significantly different using the t-test. The percentage of patients with spinal fractures with neurological deficit is 16% in the A-type fracture group, 12% in the uB-fracture group and 50% in the B-type group. The preoperative classification of patients in the A-group and uB-group is compared in Table 2. Patients in the uB-group have more than proportional relatively simple preoperative A-fractures.

Table 1 Age, fracture level and neurological deficit in A-type, uB-type and B-type fractures. Fracture level 11 means 11th thoracic vertebra,

level 12 12th thoracic, level 13 1st lumbar vertebra, etcetera.

Fracture type A uB B Number (n) 111 17 32 Age (years) 36.3 32.9 32.0 Fracture level 13.2 13.5 12.8 Neurological deficit (%) 16 12 50

Table 2 Preoperative classification in 111 A-type fractures and 17 unrecognised B-type fractures (uB)

Fracture

class A uB Subclass A uB

A1.1 1 0 A1.2 13 4 A1 15 5 A1.3 1 1 A2.1 1 0 A2.2 0 0 A2 4 0 A2.3 3 0 A3.1 55 5 A3.2 21 5 A3 92 12 A3.3 16 2


34

Table 3 Anterior wedge angle, anterior and posterior corporal height, and frontal corporal collapse in A-fractures, uB-fractures and B-fractures.

* Statistical significant difference, p<0.05

Fracture type A uB B

AWA (degrees) 17.8 17.5 20.0 ACH (mm) 23.3 25.6 24.3 PCH (mm) *36.9 *40.2 38.6 FCC (%) 34.2 30.9 29.8

Fig.2 Lateral X-ray of a patient with an initially unrecognised B-type fracture

Fig.3 Normal interspinous distance in sagittal 2D-reconstruction of a CT scan of a patient

with an initially unrecognised B-type fracture (same patient as in Fig.2)


35

AWA and ACH did not show significant differences between the groups (Table 3). The mean PCH of the uB-group was higher than the PCH of the A-group. No differences were measured between the uB-group and the B-group. The mean percentages of frontal corporal collapse (FCC) did not show a significant difference. Discussion One of the problems with comparing different studies in this field is the lack of uniformity in the use of classifications before CC was introduced [12]. Nicoll classified vertebral fractures on an anatomical basis, in four categories: anterior wedge angle, lateral wedge angle, fracture dislocation, and isolated fracture of the neural arch [17]. Holdsworth introduced the mechanism of injury into his classification: stable simple wedge fractures, stable compression or burst fractures, flexion-rotation fractures, instable fracture dislocations (extension type and rotational type) [9]. He stressed the value of palpation of the posterior ligamentary complex, consisting of interspinous and supraspinous ligaments, in order to judge stability. Chance described the osseous distraction fracture in 1948 and suggested it was a relatively stable fracture that should give a near 100% prognosis [2]. Nowadays, we consider these osseous (and ligamentary) posterior lesions, flexion distraction or B-type lesions as rather unstable [12]. Whitesides introduced the two-column concept: a pressure-resistant ventral column of vertebral bodies and discs, and a posterior column of elements with tensile strength [22]. The role and importance of the structures of the segment moyen around the spinal canal and the middle column (ventral to the spinal canal) were described by Roy-Camille and by Denis in the early 1980s, and the three-column concept was widely accepted [4;22]. Ferguson added the mechanism of injury and gave seven categories [7]. The load-sharing classification of McCormack, and the previously mentioned classification of Ferguson, tried to provide a classification on which to base the choice of therapy [7;13]. The CC 1994 provides a degree of instability as a reflection of a progressive scale of morphological damage, rather than dividing between stable and unstable fractures [12]. Some difficulties have still not been solved, for example: differentiation between A-(compression) type and ligamentary B-(distraction) type vertebral fractures sometimes is difficult, and consideration of the condition of the ligaments is essential. In some cases these dorsal ligaments are evaluable only in an indirect way in plain X-rays, conventional tomographies and CT-scans. A (large) gap


36

between the spinous processes indicates rupture of the interspinous and supraspinous ligament. Osseous distraction lesions, such as the Chance-type fracture, can be recognised more easily in plain lateral X-rays [2]. About 30% of B-type vertebral fractures are misinterpreted and considered to be A-type fractures, according to the CC [12], since ligamentary lesions are not recognised in plain X-rays and CT-scans (Fig.2 and Fig.3). This phenomenon leads to the intra-operative correction of the preoperatively type-A classified fractures of the thoracolumbar vertebral column in ligamentary distraction (B1.2) type fractures (Fig.4). The difference in PCH between A-lesions and uB-lesions might be an explanation for the existence of dorsal ligamentary lesions in the uB-group. The (intact or relatively high) posterior wall in the uB-lesions acts as a fulcrum. In further traumatic flexion and ventral compression the dorsal ligament stretches until it tears. In cases of spontaneous reduction of the dorsal structures the lesion can easily be misinterpreted. The PCH in B-lesions, however, does not differ from either A-lesions or uB-lesions. This means that a typical A-type fracture with a rather high PCH should give rise to the suspicion that this fracture might be a less stable B-type fracture. An AWA of more than 16-20° in an A-type fracture has been suggested to be a type B-lesion [1]. In vitro studies of flexion-distraction injuries of the lumbar spine, and the effect on stability, showed the first sign of permanent deformation at a flexion of 15.8°. Total disrupture occurred at 19.6° [14;15]. We could not confirm this kind of discrimination between A-type, uB-type, and B-type lesions (Table 3). Although uB-fractures are more caudal than recognised B-fractures, the difference is too small to be of clinical relevance. The difference regarding neurological deficit suggests that patients without neurological deficit have a larger chance of being misdiagnosed and, possibly, under-treated. Combined with the observed preoperative classifications, this is an even stronger possibility in rather simple fractures (A1.2 and A1.3) without neurological deficit and in more caudal lesions (L2, L3 and L4). Determination of the distance between the spinous processes is helpful, unless reduction of the fracture, fracture dislocation, or ligamentary distraction has occurred spontaneously or as a result of repositioning the patient during transport or diagnostics. Measuring differences between the interspinous distances on anteroposterior radiographs will therefore probably be of limited value. However, a difference in interspinous distance exceeding 7 mm should give rise to the suspicion of a ligamentary distraction lesion [16].


37

Fig.4 Peroperative findings of the ruptured interspinous ligament, as found during dorsal

exploration. (4a photograph, 4b schematic drawing) The left side is cranial, the right side is caudal. Arrow ligamentary tear, * spinal process

Physical examination can reveal a palpable interruption of the chain of spinous processes and dorsal interspinal ligaments [9]. MRI, as well as sonographical investigation, can provide extra information about the integrity of ligaments [8;10;11;21]. The accuracy of MRI (fat-suppressed T2-weighted sagittal sequence) in detecting interspinous ligament injuries is 97%. The positive predictive value is 96.7% and the negative predictive value 100%. In contrast, the same study neither revealed a relation between the findings on palpation and the operative findings, nor between plain radiographs (interspinous gap) and the


38

operative findings [10]. In the study of Williams et al., instability of spinal fractures was diagnosed in 50% of the fractures on plain radiograph review. MRI instability was diagnosed in 73% [23]. In some of these cases there was only indirect evidence of posterior column damage on MRI, for example haemorrhage, but no definable gap in the interspinous ligament. In these patients stretched and functionless ligaments were found at operation [23]. An MRI-investigation instead of a preoperative CT scan might also be considered [10]. MRI can be helpful in the evaluation of the status of the posterior longitudinal ligament in decision-making about the treatment of individual patients, when posterior instrumentation is considered. The continuity of the interspinous ligaments is of special importance when anterior stabilisation is considered [21]. Öner proposed a preliminary classification scheme of the observed status of the respective ligamentary, disc and osseous structures in three or four categories each. In this way the degree of instability of the anterior longitudinal ligament, posterior longitudinal ligament, posterior longitudinal complex, cranial endplate, caudal endplate, disc, and the vertebral body respectively is registered [19]. MRI can also be used to evaluate the postoperative condition of soft tissue and bone [18;20]. When only a CT scan is performed preoperatively, operative treatment with dorsal exploration eventually shows the definite condition of the dorsal ligaments. In conservative treatment, under-treatment may occur in unrecognised B-type fractures (Table 2). In future classifications, or modifications to the CC, a way of visualisation of the posterior ligaments should be applied. Conclusions Thirty percent of B-type fractures are misdiagnosed when plain X-rays and CT scans with 2D reconstructions are used as the only preoperative diagnostic tools. A large PCH with a normal interspinous distance should give rise to the suspicion of a B-type lesion. A large AWA does not point to a ligamentary B-type fracture. References 1. Blauth M, Bastian L, Jeanneret B, Knop C, Moulin P, Müller-Vahl H, Schmidt U,

Schratt HE, Wippermann B (1998) Wirbelsäule. In: Tscherne H, Blauth M (eds) Tscherne Unfallchirurgie. Springer, Berlin Heidelberg New York, p 275

2. Chance GQ (1948) Note on a type of flexion fracture of the spine. Br J Radiol

21:452-453


39

3. Daniaux H (1982) Technik und erste Ergebnisse der transpedikulären Spongiosaplastik bei Kompressionsbrüchen im Lendenwirbelsäulenbereich. Acta Chir Austr 43 [Suppl]:79

4. Dick W (1984) Osteosynthese schwerer Verletzungen der Brust- und

Lendenwirbelsäule mit dem Fixateur interne. Langenbecks Arch Chir 364:343-346 5. Dick W (1987) Innere Fixation von Brust- und Lendenwirbelfrakturen. In: Burri C,

Harder F, Bauer R (eds) Aktuelle Probleme in Chirurgie und Orthopädie 2(28). Verlag Hans Huber, Bern Switzerland, pp 53-108


12:882-900 7. Ferguson RL, Allen-BL J (1984) A mechanistic classification of thoracolumbar spine

fractures. Clin Orthop 189:77-88 8. Gillis C (1999) Spinal ligament pathology. Vet Clin North Am Equine Pract

15:97-101 9. Holdsworth FW (1963) Fractures, dislocations, and fracture-dislocations of the spine.

J Bone Joint Surg Br 45:6-20 10. Lee HM, Kim HS, Kim DJ, Suk KS, Park JO, Kim NH (2000) Reliability of magnetic

resonance imaging in detecting posterior ligament complex injury in thoracolumbar spinal fractures. Spine 25:2079-2084

11. Lehtinen A, Taavitsainen M, Leirisalo-Repo M (1994) Sonographic analysis of

enthesopathy in the lower extremities of patients with spondylarthropathy. Clin Exp Rheumatol 12:143-148

12. Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S (1994) A comprehensive

classification of thoracic and lumbar injuries. Eur Spine J 3:184-201 13. McCormack T, Karaikovic E, Gaines RW (1994) The load sharing classification of

spine fractures. Spine 19:1741-1744 14. Neumann P, Nordwall A, Osvalder AL (1995) Traumatic instability of the lumbar

spine. A dynamic in vitro study of flexion-distraction injury. Spine 20:1111-1121 15. Neumann P, Osvalder AL, Nordwall A, Lovsund P, Hansson T (1992) The

mechanism of initial flexion-distraction injury in the lumbar spine. Spine 17:1083-1090


40

16. Neumann P, Wang Y, Karrholm J, Malchau H, Nordwall A (1999) Determination of inter-spinous process distance in the lumbar spine. Evaluation of reference population to facilitate detection of severe trauma. Eur Spine J 8:272-278

17. Nicoll EA (1949) Fractures of the dorso-lumbar spine. J Bone Joint Surg Br

31:376-394 18. Öner FC, van der Rijt RR, Ramos LM, Dhert WJ, Verbout AJ (1998) Changes in the

disc space after fractures of the thoracolumbar spine. J Bone Joint Surg Br 80:833-839

19. Öner FC, van Gils AP, Dhert WJ, Verbout AJ (1999) MRI findings of thoracolumbar

spine fractures: a categorisation based on MRI examinations of 100 fractures. Skeletal Radiol 28:433-443

20. Rüdig L, Runkel M, Kreitner KF, Seidel T, Degreif J (1997) Kernspintomographische

Untersuchung thorakolumbaler Wirbelfrakturen nach Fixateur-interne-Stabilisierung. Unfallchirurg 100:524-530

21. Saifuddin A, Noordeen H, Taylor BA, Bayley I (1996) The role of imaging in the

diagnosis and management of thoracolumbar burst fractures: current concepts and a review of the literature. Skeletal Radiol 25:603-613

22. Whitesides TEJ (1977) Traumatic kyphosis of the thoracolumbar spine. Clin Orthop

128:78-92 23. Williams RL, Hardman JA, Lyons K (1998) MR imaging of suspected acute spinal

instability. Injury 29:109-113

THORACOLUMBAR FRACTURES 3 RADIOLOGICAL RESULTS

41

Chapter 3

Thoracolumbar spinal fractures: radiological results of transpedicular fixation combined with transpedicular cancellous bone graft and posterior fusion in 183 patients V.J.M. Leferink, K.W. Zimmerman, E.F.M. Veldhuis, E.M. ten Vergert, H.J. ten Duis Published in European Journal (2001) 10:6:517-523 © Springer Introduction The rationale for the operative treatment of thoracolumbar spinal fractures has been the subject of discussion in about 300 articles between 1975 and 1994 [13]. Despite this it remains unclear exactly what changes occur in the disturbed anatomy of the spine during surgery and during the course of treatment. Since Aebi et al. showed that a better anatomical correction could be achieved with the Dick internal fixator technique than with the classical Harrington rods [1-3], we have been treating patients with unstable fractures, and fractures with considerable angulation or neurological deterioration by posterior reduction and stabilization with short segment fixation with Dick's internal fixator. In those cases in which the fracture of the vertebral body was actually reduced, the posterior procedure was combined with a transpedicular autologous cancellous bone graft. Posterior fusion of the intervertebral facet joints only at the level of the destroyed end plates was performed in all cases. As all patients were treated in a strict working protocol, our work resulted in a unique database over a 10-year period. We managed to collect almost complete data during a follow-up period of 2 years for our radiological retrospective analysis.


42

For the study reported here, we tried exclusively to find answers to the following questions: • Does the vertebral body collapse after removal of the implants, despite

transpedicular bone grafting? • Is correction of the regional angle (RA) maintained after surgery? • And if not, does the loss in the intervertebral angle (IVA) contribute to this

change? • What is the influence of implant failure on the radiological measurements? Materials and methods All consecutive patients, aged 18-65 years, with a fracture of the thoracolumbar spine between the 9th thoracic and the 5th lumbar vertebral body, surgically treated at the Traumatology Department of the University Hospital Groningen between March 1988 and August 1996 were included in this study, and of these 51% were referred from other hospitals. Patients with fractures in osteoporotic bone or with other pathological conditions (metastases) were excluded. All the available clinical records, and operative, follow-up, rehabilitation plus all radiographic material (including conventional tomographs and computerized tomographs with 2D reconstructions) were reviewed.

2 36

38

80

31

8 104

10

10

20

30

40

50

60

70

80

90

T9 T10 T11 T12 L1 L2 L3 L4 L5 L6

level

n

Fig.1 Fracture level in 183 patients


43

By definition all patients had sustained a trauma. Falls and jumps from a height accounted for 64.5% of all injuries, and 25.1% were from traffic accidents. Young adults were predominant; the median age was 32 years. Although many of the patients showed temporary sensory loss in the legs, 27% had a definitive objective neurological deficit, varying from (partial) conus-cauda lesions to complete paraplegia. 37% of the patients had other injuries unrelated to the spinal fracture. Most of the patients suffered a 12th thoracic, a 1st or a 2nd lumbar fracture (Fig.1). According to the Comprehensive Classification [19], 128 were type A fractures, 32 type B and 21 type C as identified preoperatively. Two fractures could not be classified according to the Comprehensive Classification because of missing data (Table 1). An example of a type A3.1 fracture is shown in Fig.2.

Table 1 Comprehensive Classification in 183 patients

A1.1 1 A1.2 17 A1 20 A1.3 2 A2.1 1 A2.2 0 A2 4 A2.3 3 A3.1 60 A3.2 26

A 128

A3 104 A3.3 18 B1.1 3 B1.2 13 B1 16 B1.3 0 B2.1 1 B2.2 1 B2 14 B2.3 12 B3.1 1 B3.2 0

B 32

B3 2 B3.3 1 C1.1 3 C1.2 0 C1 15 C1.3 12 C2.1 5 C2.2 0 C2 5 C2.3 0 C3.1 1 C3.2 0

C 21

C3 1 C3.3 0

Unknown 2


44

t=0 t=1 t=9 t=24

Fig.2 AWA, IVA and RA in radiographs during the course of treatment of a type A3.2

fracture in a 39-year-old man (as an example) (t=0 before surgery, t=1 within 1 month of operation, t=9 before

implant removal at 9 months, t=24 24 months after trauma) If possible, the operative treatment was performed on day 4 after trauma (median, range 0-31 days), but 17 patients were operated upon immediately after admission because of neurological impairment, and 28 after more than 10 days. During surgery, 17 out of 105 fractures thought to be type A were found to be type B lesions after exploration of the dorsal ligaments. All fractures were treated by instrumental angular reduction, distraction and stabilization with Dick's internal fixator [10-12]. Since 1995 we have used the Universal Spine System (Synthes®). The procedure was combined with unilateral (1988-1989) or bilateral (1989-1996) transpedicular cancellous bone grafts as described by Daniaux [7;8]. Posterior spondylodesis was added only at the level of the disturbed cranial or caudal end plate. No ventral operations or laminectomy were performed. Approximately 10 days after surgery the patients were transferred to a rehabilitation centre for a mean stay of 4 weeks. Here they were allowed to mobilize in a standard thoracolumbar support orthesis. The brace was worn for 9 months; the first 6 months day and night and the last 3 months only during the day. After 9 months all implants were removed except in five patients in whom the spondylodesis had been performed at the same segments as the internal fixation. At 12 months all patients were instructed to recommence all their former activities. The last follow-up examination was 2 years after the initial operation. One patient died postoperatively of a severe intrathoracal bleeding complication. Three patients died during follow-up due to pulmonary complications in complete paraplegia (4, 8 and 13 months after operative treatment). Two patients performed


45

lethal suicidal attempts during follow-up. Two patients were lost to follow-up. Since 1989 only minor changes in the treatment protocol have been made. We limited our study to four parameters (Fig.2): • Changes in the anterior wedge angle (AWA) of the fractured vertebral body

in all our patients. Additionally, we performed a separate evaluation of those (101) patients with type A3 fractures with a positive AWA on admission.

• Changes in the RA (i.e. the angle formed by the cranial and caudal end plates of the adjacent intact vertebrae).

• Changes in the IVA at the end plate of the fractured vertebral body at the level of the involved disc. The "second" disc's IVA was not directly measured.

• The influence of implant failure upon the changes in AWA, IVA and RA.

Angles were measured in plain radiographs preoperatively, and within 1 month postoperatively, 9 months (before removal of implants, which we did in 97% of the patients) and 24 months postoperatively (all patients). The differences measured in each period were calculated: the perioperative period (period I), the period until implant removal (period II) and the period after implant removal (period III). The Kolmogorov-Smirnov (K-S) test was used to compare the distribution of the angle changes per period with a normal distribution. The K-S test is a one-sample test for goodness-of-fit, like the chi-squared test, but is preferred for small samples because it does not lose information due to combining of categories [22]. Only for angle changes that did not fit a normal distribution a nonparametric test of significance was used (one-sample runs test). Angle changes with a normal distribution were compared using the t-test. Angles of patients with and without implant failure were compared using the unpaired t-test. Results The mean kyphotic AWA on admission was 18.0°. After the operation this was reduced to 5.9°. At 9 months, before removal of the implants, the mean AWA was 6.8° and at 24 months 7.3° (Fig.3, Table 2). The lordotic angle of the IVA on admission was 4.5°, after reduction 4.9°, at 9 months 3.7°, and at 24 months -0.4°. This means that the lordotic angle at the level of the intervertebral space changed to a neutral angle (0.4° kyphosis; almost parallel) after implant removal (Fig.3, Table 2). The mean kyphotic RA on admission was 9.9°. After the operation it was reduced to -0.3°. At 9 months it was 2.3° and had changed to 9.2° at the end of the follow-up (Fig.3, Table 2). At that time the RA was similar to the RA on admission.


46

Fig.3 AWA, IVA and RA in time. Positive values indicate kyphosis;

negative values indicate lordosis

Table 2 Different angles in relation to time in 183 patients. Positive AWA and RA

indicates kyphosis; positive IVA indicates lordosis

AWA IVA RA Month

Mean SD SEM Mean SD SEM Mean SD SEM 0 18.0 7.9 0.88 -4.5 4.8 0.37 9.9 11.6 0.59 1 5.9 5.5 0.87 -4.9 3.5 0.26 -0.3 11.5 0.41 9 6.8 5.7 0.87 -3.7 11.1 0.23 2.3 11.5 0.43 24 7.3 6.2 1.0 0.4 10.0 0.30 9.2 12.8 0.48

The K-S test discriminated between parametric (AWA I, IVA I, IVA III, RA I and RA II) and nonparametric distributions of the calculated angle changes (AWA II, AWA III, IVA II and RA III) (Table 3). AWA I, IVA III, RA I and RA II differed significantly from the zero distribution, according to the appropriate test (t-test or runs test; Table 3). The changes in AWA obtained during operation (period I) were statistically significant. Later changes were not significant. A separate


47

analysis of changes in the AWA in the type A3 cases with a positive AWA (n=101) (mean-2xsd = 18.0-2x7.9 = 2.2°) showed a significant reduction of 11.5° in period I, a small but marginally significant AWA change in period II (0.80°, p=0.003) and a somewhat larger but still very small and non significant AWA change in period III (1.2°, p=0.184). No statistically significant changes in IVA occurred during the course of treatment during periods I and II; the major part of the change in IVA occurred in period III and the difference was significant. The differences of the RA's observed in the studied periods were statistically significant before 9 months after surgery, but the changes in the RA after removal of the implants were nonparametric and not significant (Table 3). Table 3 Distribution type of angle changes and runs/t-test results compared to the zero distribution in 183 patients. AWA data are given for 101 type A3 fractures with a positive AWA. The distributions of the measured changes (PAR parametric, NPAR nonparametric) were tested by Kolmogorow-Smirnov test (column 4) in order to perform the right test for comparing the values with the zero distribution (column 5: Student's t-test for parametric distributions and the runs test for nonparametric distributions). The total numbers of (complete) pairs of measurements are listed in column 6 and the number of runs in the runs test in column 7. The test values in column 8 indicate the difference from the zero distribution, and the p-values in column 9 indicate the statistical significance.

Angle change

Mean SEM

Distribution (4)

Test (5)

Total cases (6)

Runs (7)

Z/T value (8)

p (9)

AWA I 11.5 0.6 PAR T 100 19.451 <0.001 AWA II -0.8 0.3 NPAR Runs 97 70 3.070 0.003 AWA III -1.2 0.9 NPAR Runs 94 69 1.339 0.184 IVA I 0.90 0.6 PAR T 165 0.604 0.547 IVA II -0.064 1.0 NPAR Runs 171 59 0.247 0.805 IVA III -4.1 1.1 PAR T 162 12.579 <0.001 RA I 10.0 0.7 PAR T 169 15.040 <0.001 RA II -3.0 0.4 PAR T 172 8.139 <0.001 RA III -7.0 0.6 NPAR Runs 164 19 1.610 0.107 In 20 patients (10.9%) one or two pedicle screws were shown to be broken on radiological evaluation 9 months after surgery. Analysis of the changes in the AWA, IVA and RA and when compared with measurements in patients without breakage of screws showed significant changes in the AWA and RA between primary operation and implant removal, but no changes in the other periods, and no changes in the IVA (Table 4).


48

Table 4 Differences in angle changes between 20 patients with implant failure and 156 without implant failure (unpaired t-test and significance)

Period Failures (n=20)

Controls (n=156)

Mean difference

p

AWA I 12.15 12.01 0.14 0.937 AWA II -2.15 -0.69 -1.46 0.030 AWA III -1.78 -1.17 -0.61 0.735 IVA I 0.67 -0.40 1.07 0.434 IVA II -1.80 0.85 -2.65 0.144 IVA III -3.35 -4.26 0.91 0.398 RA I 9.00 10.16 -1.16 0.574 RA II -8.75 -2.27 -6.48 <0.001 RA III -5.18 -7.24 2.06 0.309

Discussion Surgery for spinal fractures can be performed with various instrumentation systems including pedicle screws, hook rods (such as Harrington rods), Luque rods, and anterior instrumentation. Considerable controversy exists regarding the clinical outcome with these different instrumentation systems [4]. Thus Dickman et al. [13], who performed a meta-analysis of surgical treatment alternatives comparing the results from the use of four instrumentation systems, were unable to find convincing evidence as to the best method of treatment of unstable fractures of the thoracolumbar spine. They studied 308 reports published between 1975 and 1994. Of these, 250 were excluded for scientific reasons, and from the remaining 58 they concluded that posterior instrumentation with pedicle screws was the best method with regard to fusion rate, functional outcome and incidence of intraoperative and postoperative complications, including pain and neurological complications. Their study of other aspects concerning the efficacy of different implants and methods did not provide meaningful conclusions, and they stress that no prospectively randomised studies have been reported that comprehensively evaluate the results of different spinal implants for spinal trauma. In general, retrospective multicentre trials (for example Esses et al [14]) are not useful for comparing the results of large series of patients treated with different methods because it is very difficult to allow for differences in surgical technique. Esses et al. [14] had to exclude 25% of their patients because of insufficient follow-up data. Although our study was a retrospective radiological study, all our patients were treated in the same centre by two surgeons, recruited


49

from a group of two (later three) senior staff members and a changing number of junior surgeons. The treatment was performed according to a constant working protocol, which included standard intervals for radiological evaluation, giving us the opportunity to evaluate the radiological results. AWA We did not find any significant loss in AWA during the first 9 months following surgical treatment (period II) or after removal of implants (period III), although several authors have suggested that the shape of the fractured vertebral body obtained during the operative procedure changes after removal of the implant, with or without spongioplasty [15]. However, in type A3 fractures with an AWA >2.2° a small but significant loss in AWA was found in the postoperative period [2;8;17;18]. Our favourable results in this respect might have been because of the use of the bilateral transpedicular bone graft procedure. Studies comparing unilateral and bilateral transpedicular cancellous bone grafts have not been performed. The study of Lindsey and Dick [18] showed a small loss in AWA of only 0.5° in 76 patients, and even though all of these patients had a neurological deficit and only 27% of our patients had such a deficit, the results regarding AWA changes are comparable. IVA The IVA at the "affected" segment changed significantly after implant removal. This suggests that the internal fixator and the posterior fusion have a temporary protective effect against collapse of the intervertebral space, but cannot prevent complete disc collapse after implant removal. Independently of weight bearing, this occurred mainly in period III after removal of the implant, as in the study of Lindsey and Dick, in which a 5° loss in the intervertebral space was observed [18]. RA Changes in the RA, comparable to the Cobb angle or kyphosis angle in some other studies [6;15], were observed during all phases of the treatment. Posterior instrumentation resulted in the correction of the RA by 10.0°. Loss of RA during the remainder of the follow-up period accounted for 3.0° and 7.0° in the respective periods resulting in a complete return to the preoperative value. This recurrent kyphosis is comparable to the findings of Knop et al. (10.1° in 16-59 months follow-up, mean 40 months) [15], but much more than those reported by Aebi et al. and Olerud et al. (3.6° in 12 months; 4° in 10 months) [2;20]. In a retrospective study, Crawford and Askin compared two historical groups of patients. They showed that the correction in the RA is greater and maintained better if transpedicular bone grafting of the vertebral body is also


50

performed. This study had a mean follow-up of 9 months. The follow-up was not specified for both groups and therefore it is reasonable to conclude that it was longer for the first historical group (without bone graft) [6]. Certainly the length of follow-up influences the findings (Table 5). Short follow-up shows "good" results and long follow-up shows progressive regression to values comparable to the preoperative RA. This explains why early studies [2-6;8;9;14;17] showed good results from posterior instrumentation. Only recently has this recurrent kyphosis, despite transpedicular fixation, been shown by Speth et al. [23] and Knop et al. [15], but in relatively small numbers of patients and/or with a short follow-up. Non significant changes in RA occurred in period III (p=0.107), although the differences were rather large, they can partially be explained by the nonparametric distribution and therefore the obligatory nonparametric runs test, the relatively large standard deviation, and large standard error of the mean. Possibly, this might have been caused by poor reproducibility of the measurements: the RA may have been influenced by postural factors at the level of the second disc. The IVA at the level of the caudal end plate at t=0 could have been larger as a result of ventral distraction in the preoperative supine position after trauma. The other measurements were done on sitting or standing radiographs. Regression towards the "normal" IVA value of the second disc at 24 months would influence the RA and could be an additional explanation for the relatively large regression of the RA (Fig.3, Table 2).

Table 5 Changes in RA from fracture reduction until end of follow-up

compared to the length of follow-up

Author n Follow-up in months

(mean)

Loss in RA (after operation

to end of follow-up)

Aebi [2] 30 12 (12) 3.6° Crawford [6] 50 ? (9) 5-8° Daniaux [8] 44 6-49 (26) 10.4° Knop [15] 56 16-59 (40) 10.1° Liljenqvist [17] 26 24-59 (34) 7.2° Lindsey [18] 76 24 (24) 8.5° Olerud [20] 20 6-17 (10) 4° Speth [23] 24 18-48 (35) 10-11° This study 183 24 (24) 10.0°


51

Collapse of both discs could be due to degeneration as a response to both trauma and immobilization. Fusing only two of the three instrumented vertebrae has -at least theoretically- the advantage of regaining mobility at the second segment as early as possible after implant removal. This does not mean that the second disc will not be influenced by the procedure. Accelerated degeneration of the facet joints adjacent to a lumbar fusion has been described by Lee [16]. Implant removal at 9 months could interfere with the posterior spondylodesis, accelerating the return to the preoperative RA values. It is likely that avoiding posterior fusion at the second segment and early implant removal do not prevent degeneration of the second disc. Subgroups, for example type A3 fractures with a positive AWA, were only evaluated separately in this study with respect to changes in the AWA. The effect of the posterior, one-level spondylodesis in relation to the segmental range of motion will be presented in a separate report. Canal clearance, posterior wall height restoration and related topics will be the subjects of a third report. Breakage of pedicle screws led to statistically significant changes in the AWA and RA in period II but surprisingly not to differences in the changes in the IVA (Table 4). Therefore we cannot conclude that the (intact) internal fixator only temporarily prevents disc collapse; breakage of pedicle screws is not reflected at disc level. We were not able to determine the exact moments of screw breakage; patients did not report any symptoms and the radiological evidence of implant failure could only be found at the scheduled check-ups, before implant removal. Screw breakage mainly occurred before 1990 when we advised our patients to wear a corset for only 6 months and we removed the implants at 12 months. This suggests material fatigue and a breakage tendency of 5-mm screws. From 1990 we advised patients to wear the brace until implant removal at 9 months to protect the screws against breakage. However, this function of bracing has been questioned by Rohlmann et al., who showed that braces do not prevent stress in internal fixators [21]. Conclusions The reduced vertebral body does not collapse after removal of the implants at 9 months, but internal fixation and posterior spondylodesis cannot prevent collapse of the affected intervertebral space occurring after implant removal. This collapse contributes significantly to the loss of the RA. The RA at the end of 2 years follow up was shown to be almost similar to the preoperative value. Implant failure occurring between primary operation and implant removal had a significant influence on the AWA and RA before implant removal, but did not influence the IVA.


52

References 1. Aebi M (1987) Die operative Behandlung von Wirbelsäulenverletzungen. Fortschr

Med 105:549-551 2. Aebi M, Etter C, Kehl T, Thalgott J (1987) Stabilization of the lower thoracic and

lumbar spine with the internal spinal skeletal fixation system. Indications, techniques, and first results of treatment. Spine 12:544-551

3. Aebi M, Etter C, Kehl T, Thalgott J (1988) The internal skeletal fixation system.

A new treatment of thoracolumbar fractures and other spinal disorders. Clin Orthop 227:30-43

4. Blauth M, Tscherne H, Haas N (1987) Therapeutic concept and results of operative

treatment in acute trauma of the thoracic and lumbar spine: the Hannover experience. J Orthop Trauma 1:240-252

5. Carl AL, Tromanhauser SG, Roger DJ (1992) Pedicle screw instrumentation for

thoracolumbar burst fractures and fracture-dislocations. Spine 17:S317-S324 6. Crawford RJ, Askin GN (1994) Fixation of thoracolumbar fractures with the Dick

fixator: the influence of transpedicular bone grafting. Eur Spine J 3:45-51 7. Daniaux H (1982) Technik und erste Ergebnisse der transpedikulären

Spongiosaplastik bei Kompressionsbrüchen im Lendenwirbelsäulenbereich. Acta Chir Austriaca 43:S79

8. Daniaux H (1986) Transpedikuläre Reposition und Spongiosaplastik bei

Wirbelkörperbrüchen der unteren Brust- und Lendenwirbelsäule. Unfallchirurg 89:197-213

9. Dick W (1984) Osteosynthese schwerer Verletzungen der Brust- und

Lendenwirbelsäule mit dem Fixateur interne. Langenbecks Arch Chir 364:343-346 10. Dick W (1987) Innere Fixation von Brust- und Lendenwirbelfrakturen. In: Burri C,

Harder F, Bauer R (eds) Aktuelle Probleme in Chirurgie und Orthopädie, 2(28). Verlag Hans Huber, Bern, Switzerland, pp 1-137


12:882-900 12. Dick W, Kluger P, Magerl F, Woersdorfer O, Zach G (1985) A new device for

internal fixation of thoracolumbar and lumbar spine fractures: the 'fixateur interne'. Paraplegia 23:225-232


53

13. Dickman CA, Yahiro MA, Lu HT, Melkerson MN (1994) Surgical treatment alternatives for fixation of unstable fractures of the thoracic and lumbar spine. A meta-analysis. Spine 19:S2266-S2273

14. Esses SI, Botsford DJ, Wright T, Bednar D, Bailey S (1991) Operative treatment of

spinal fractures with the AO internal fixator. Spine 16:S146-S150 15. Knop C, Blauth M, Bastian L, Lange U, Kesting J, Tscherne H (1997) Frakturen der

thorakolumbalen Wirbelsäule. Spätergebnisse nach dorsaler Instrumentierung und ihre Konsequenzen. Unfallchirurg 100(8):630-639

16. Lee CK (1988) Accelerated degeneration of the segment adjacent to a lumbar fusion.

Spine 13:375-377 17. Liljenqvist U, Mommsen U (1995) Die operative Behandlung thorakolumbaler

Wirbelsäulenverletzungen mit dem Fixateur interne und transpedikulärer Spongiosaplastik. Unfallchirurgie 21:30-39

18. Lindsey RW, Dick W (1991) The fixateur interne in the reduction and stabilization of

thoracolumbar spine fractures in patients with neurological deficit. Spine 16S140-S145

19. Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S (1994) A comprehensive

classification of thoracic and lumbar injuries. Eur Spine J 3:184-201 20. Olerud S, Karlstrom G, Sjostrom L (1988) Transpedicular fixation of thoracolumbar

vertebral fractures. Clin Orthop 227:44-51 21. Rohlmann A, Bergmann G, Graichen F, Neff G (1999) Braces do not reduce loads on

internal spinal fixation devices. Clin Biomech (Bristol, Avon) 14:97-102 22. Siegel S, Castellan NJ (1988) Nonparametric statistics for the behavioral sciences.

The Kolmogorov-Smirnov one-sample test. McGraw Hill, New York, pp 51-64 23. Speth MJ, Öner FC, Kadic MA, de Klerk LW, Verbout AJ (1995) Recurrent kyphosis

after posterior stabilization of thoracolumbar fractures. 24 cases treated with a Dick internal fixator followed for 1.5-4 years. Acta Orthop Scand 66:406-410


54

THORACOLUMBAR FRACTURES 4 SPINAL CANAL

55

Chapter 4 Burst fractures of the thoracolumbar spine: changes of the spinal canal during operative treatment and follow-up V.J.M. Leferink, J.M.M. Nijboer, K.W. Zimmerman, E.F.M. Veldhuis, E.M. ten Vergert, H.J. ten Duis

Accepted European Spine Journal October 2002 Introduction One of the goals of operative treatment of spinal fractures is restoration of the anatomy of the spinal chain, including the spinal canal. Since 1988 we have been using instrumental decompression of the spinal canal at the thoracolumbar junction in an indirect way, what is said to be by ligamentotaxis through distraction applied via pedicle screws of an internal fixator [1;9-11;16;23-25]. In a dorsal approach all fractures were treated by instrumental angular reduction, distraction, and stabilisation with Dick’s internal fixator [8;9;11]. Since 1995 we have used the Universal Spine System®, Synthes. The procedure was combined with unilateral (1988-1989) or bilateral (1989-1996) transpedicular cancellous bone grafts, as described by Daniaux [5]. Posterior spondylodesis was performed only at the level of the disturbed cranial or caudal end plate [26;27]. Several authors have described spontaneous remodelling of the spinal canal during the course of treatment, with or without instrumentation. This finding has been used as an argument against all operative treatment or against direct (open) manipulation of the bony fragment [3;4;17;21;24;34]. One of the parameters that correlate with spinal canal encroachment is the posterior vertebral height (PVH)


56

[20]. Another important factor concerning the reduction is the antikyphosed position of the disk and its bony fixation as well as the rigid fixation [1;7]. To quantify the process of reduction during operation and spontaneous remodelling during the period of reconvalescence, we studied this process in patients with burst fractures of the thoracolumbar junction who were treated by indirect fracture reduction and internal fixation. There were three questions to answer: • Are there any changes of the posterior segmental height PSH (the sum of

posterior vertebral height and posterior intervertebral height) in plain lateral radiographs during operation and in the course of further treatment

• Does bony narrowing of the spinal canal, as identified in plain lateral radiographs disappear in the course of treatment, and if so, how many of these patients have developed a normal width of the spinal canal two years after the initial treatment.

• Are there changes of the midsagittal diameter of the spinal canal in selected cases with considerable preoperative spinal canal narrowing, as measured in CT-slices.

Presumably the effect of the so-called ligamentotaxis can be measured as changes of the above mentioned parameters in the perioperative period. The PSH reflects the length of the segmental part of the posterior longitudinal ligament that is involved in the ligamentotaxis. The effect of spontaneous remodelling is reflected in changes of the studied parameters after the operation. Materials and methods Between March 1988 and August 1996 183 consecutive patients with thoracolumbar fractures (T9-L5) were treated operatively at the Traumatology Department of the University Hospital Groningen. Of these patients 95 had A3-fractures of the thoracolumbar junction (T9-L2) that were treated with indirect operative reduction and fixation with Dick’s internal fixator, combined with transpedicular cancellous bone grafting and posterior spondylodesis (Table 1). In these patients the preoperative plain lateral radiographs (t=0) were studied, as well as the postoperative radiographs (t=1), the radiographs after 9 months (before implant removal, t=9), and at 24 months (t=24). The sub-classification, according to the comprehensive classification, of the fractures and the percentage of patients with neurological deficit are shown in table 2 [29].


57

Table 1 Level of spinal fractures (n=95)

Level n (%)

T9 1 (1.1) T10 1 (1.1) T11 1 (1.1) T12 24 (25.3) L1 51 (53.7) L2 17 (17.9) Total 95 (100)

Table 2 Comprehensive classification in 95 patients with A3-fractures and the number and percentage of neurological deficit

Classification n Neurological deficit

(%) A3.1 55 9 (16.4) A3.2 25 3 (12.0) A3.3 15 3 (20.0) A3 total 95 15 (15.8)

Posterior segmental height The changes of the posterior segmental height PSH in the course of treatment and follow-up were analysed (Fig.1). Technical differences in radiographic technique were corrected using the proportion of the measured and unchanged height of the adjacent vertebral bodies. Correction of the imaging amplification factor in lateral radiographs (15-20%) was not done, because this would not influence the statistical tests. Most measurements were performed in the center of the photograph. In case of distortion of the radiograph by non centered beam the projected area of the disc was determined from the oval outlines of the rims of the vertebral endplates (Fig.1). Guidelines as described by Frobin were followed, resulting in a relative measurement error of approximately 3%, consisting of a relative error of the posterior vertebral height of 2.2% (0.7mm) and of the disk height of 4.2% (0.5mm) [15].


58

Fig.1 Measurement of posterior segmental height (PSH=PIH+PVH). See text for explanation

Bony narrowing In all 95 patients the plain lateral radiographs were studied and recognisable bony narrowing of the spinal canal (with or without recognition of a fracture part) were registered (Fig.2). After an initial period in which conventional tomographies were used in the preoperative workup, all patients had preoperative CT-scans. In all preoperative CT-scans the midsagittal diameter was measured and patients were classified in 4 groups: no narrowing, less than one third canal narrowing, more than one third, but less than two third narrowing and more than two third canal narrowing [28].

t=0 t=1 t=9 t=24

Fig.2 Recognizable bony narrowing in plain lateral radiographs in patient with A3.1 fracture, preoperative t=0, postoperative t=1 (horizontal arrows pointing at fragment),

at 9 months t=9, and at 24 months t=24 (vertical arrow pointing at completely reduced posterior wall)


59

Midsagittal diameter In 66 patients preoperative CT-scans were available to measure the midsagittal diameter. In 13 patients with a preoperative narrowing of the spinal canal of more than one third we compared the preoperative midsagittal diameter of the spinal canal with the diameter at two years after operation by CT-scanning. In computerized tomographies of the spinal canal preoperatively and after two years the minimum midsagittal diameter of the spinal canal was compared to the mean of the minimum diameter of the spinal canal one level cranial and one level caudal. These percentages represent the remaining space in the spinal canal and changes in diameter represent the summed effect of reduction and remodelling. The CT data of these patients were compared to the data of the plain lateral radiographs. Statistical analysis was performed with the paired samples t-test, a parametric test. Results Posterior segmental height The mean PSH in preoperative radiographs measured 40.5 mm, after operation 43.2 mm, before implant removal 41.0 mm and at 24 months it measured 38.7 mm (Table 3). Analysis of the (calculated) changing of the PSH in paired measurements in the perioperative period, the period until implant removal and the period until 24 months after the initial operation show an increase of the PSH during operation of 2.7 mm (p<0.001). After the operation a decrease of 2.2 and 2.3 mm respectively occurs (p<0.005) (Table 4).

Table 3 Posterior segmental height (summed posterior vertebral height and posterior intervertebral height) and percentages of patients with

recognisable bony canal narrowing in the plain lateral radiographs during the course of treatment (n=95). SD Standard deviation

Period PSH (mm)

SD (mm)

Bony narrowing (%)

Preoperative 40.5 5.7 76.5 Postoperative 43.2 4.1 18.4 At 9 months 41.0 3.9 8.2 At 2 years 38.7 3.8 2.4


60

Table 4 Paired samples t-test comparing the calculated differences of the length of the ligament along PVH and PIVS to the test value zero.

Period 1= perioperative period, period 2= the period between initial operation and implant removal, period 3= period after implant removal until end of follow up at 24

months. CID= confidence interval of the difference, SEM= standard error of the mean, t= test value, df= degrees of freedom

95% CID p Period n Mean (mm)

SD (mm)

SEM (mm) Lower Upper

t df

1 83 2.72 5.72 0.63 1.47 3.97 4.34 82 <0.001 2 82 -2.16 4.60 0.51 -3.17 -1.14 -4.25 81 <0.001 3 83 -2.29 3.29 0.36 -3.01 -1.57 -6.34 82 <0.001

0

20

40

60

80

100

preop postop 9 m 24 m

t

%

Fig.3 Percentage of patients with bony spinal canal narrowing

in plain lateral radiographs in 95 patients with A-type fractures

Table 5 Narrowing of the spinal canal measured in CT-scans of 66 patients with burst fractures and the number and percentage of neurological deficit

Narrowing n (%) Neurological deficit

(%) 0 7 (10.6) 1 (14.3)

>0-1/3 32 (48.5) 4 (12.5)

1/3-2/3 23 (34.8) 5 (21.7)

>2/3 4 (6.1) 0 (0)

Total 66 (100) 10 (15.2)


61

Bony narrowing In plain lateral radiographs recognizable bony encroachment of the vertebral canal was seen preoperatively in 76.5% of the patients, postoperatively in 18.4%, at nine months in 8.2% and at two years in 2.4% (Table 3 and Fig.3). Midsagittal diameter Preoperative CT-scans in 66 patients with A-fractures showed in 89.4% a measurable spinal canal narrowing. Less than one third canal narrowing was found in 48.5%, 34.8% had more than one third spinal canal narrowing, but less than two third. Only 6.1% of the patients had more than two third canal narrowing. Patients with spinal canal narrowing of 1/3 to 2/3 showed the highest percentage of neurological deficit (Table 5). Traumatic narrowing of the spinal canal in the sagittal plane in the preoperative CT-scans left a mean residual midsagittal diameter of 52.3% (n=66). After two years the midsagittal diameter was 78.3% (n=13). Bony narrowing in the lateral radiographs in these patients was recognized in 92% and 15% respectively (Table 6).

Table 6 Residual percentage of midsagittal spinal canal diameter in CT-scan and number (percentage) with recognizable

bony narrowing in lateral radiograph in 13 patients

Canal narrowing in CT

Mean (%) SD

Bony narrowing (%)

Preoperative 47.7 21.6 12 (92) Two year 21.7 20.5 2 (15)

The paired samples t-test showed that the changes between preoperatively and two years postoperatively accounted for 25.0% (2 tailed significance p<0.001) (n=14) (Table 7).

Table 7 Paired t-test, comparing differences between preoperative residual ratio of the midsagittal diameter of the spinal canal and the residual spinal canal at two years

95% CID Mean SD SEM

Lower Upper t df p

Difference 0.25 0.18 0.049 0.14 0.36 5.06 12 <0.001


62

Discussion It is generally accepted that bony encroachment of the spinal canal in the thoracolumbar region by one third or more might jeopardize the spinal cord [18;28]. There is a correlation between the level of the spinal fracture and the probability of neurological deficit. More cranial levels of fracture have a higher probability of neurological deficit [12]. The percentage of spinal canal narrowing as measured in CT-scan has a positive correlation with the probability of neurological deficit as well [12;18]. In this regard one should realize that the measured residual diameter at the preoperative radiographs is larger than the diameter at the moment of trauma [30]. Another aspect of interest is that the midsagittal diameter overestimates the canal narrowing, compared to area surface measuring [13]. Posterior segmental height Through intact spinal ligaments and discs a partial fracture reduction will be induced when the patient is put in a supine position during initial care and transport. Extension does not widen the spinal canal in an unfractured spinal column [20], but the combination of angular fracture reduction and distraction will widen the spinal canal by means of ligamentotaxis [14]. For anatomical reasons ligamentotaxis below L2 level is weak or even absent [24]. If the longitudinal ligaments-especially the posterior longitudinal ligament-are not (completely) disrupted, distraction and antikyphosis (ligamentotaxis) can achieve a reduction of bony fragments narrowing the spinal canal of the injured spine. The effect of the forces conducted via the attachment of the annulus to the endplates by instrumental and postural antikyphosing reduction will add to the restoration of the spinal canal wall. Bony canal narrowing It is the trabecular structure of the spinal body that causes the typical trapezoid fracture part in the posterior wall in burst fractures, which results in narrowing of the spinal canal. The trabeculae are found at the medial corner of the base of the pedicles and extend in a radial array throughout the vertebral body. A stress concentration near the base of the pedicles results in the typical fracture at the posterior wall of the vertebral body in severe compression [19]. In 1991 Johnsson reported about 17 thoracolumbar fractures and concluded that manipulative open reduction of spinal canal wall and bony fragments in the spinal canal is not necessary because spontaneous spinal canal remodelling occurs in operated and non-operated patients. In his study 14 operated and 3 non-operated


63

patients in a follow-up of 1-4 year were evaluated. One of the conclusions was that there were no differences between non-operated and operated cases concerning the restitution of the spinal canal. In his study the measurements were difficult to interpret because laminectomy in some cases or interference from Harrington rods in other cases influenced the quality of the radiographs and the measurements [21]. Scapinelli reported in 1995 about 5 adult patients with thoracolumbar spinal fractures with associated intracanalar displacement of a large bone fragment. Two of these patients had no neurological deficit and three patients had. Four patients underwent operative posterior stabilization by Harrington rods and bone autografts without surgical decompression. These treatments lead to neurological recovery in all but one case. Comparison of computer tomography scans in all patients after 1.5 to 5 years later showed remodelling of the spinal canal. Rhythmic respiratory oscillations in the cerebrospinal fluid pressure were suggested as a factor in the mechanism of bone resorption. It was concluded that removal of intraspinal fragments is not longer necessary [32]. The discussion about laminectomy as the procedure of choice for decompression in patients with neurological deficit will probably never end completely. We earlier reported about 183 patients (17% of them had neurological deficit) treated by dorsal instrumentation, in which we only used operative decompression by laminectomy in one case, because of progressive neurological detoriation after the initial dorsal operative procedure [27]. Boerger [2] showed that no reason for surgical decompression can be found in published research. In the present study we confirm a link between initial canal narrowing and its risk for traumatic neurological deficit. We could not confirm a higher incidence of neurological deficit in the subclassification of more severe A3 fractures, but low numbers of A3.3 type fractures and low numbers of severe (more than 2/3) traumatic canal narrowing make it impossible to draw any conclusion of this observation. One should realize that the midsagittal diameter reduction overestimates the traumatic reduction of the spinal canal compared to CT area measurements [13]. The present study shows that fracture reduction by angular correction and distraction is accompanied with a marked increase of patients with a cleared spinal canal from 23.5% to 81.6%. This very large change is followed by a further increase of the number of patients with complete canal clearance to 91.8% at 9 months postop. At the end of follow up- at 24 months postop -only 2.4% of the patients have residual canal narrowing (Fig.3). Midsagittal diameter The caliber of the vertebral canal in the lumbar region shows a large variation. The anteroposterior diameter in normal individuals decreases from 17.3 mm (range


64

13-22 mm at level L1) to 15.9 mm (range 9-21 mm at level L4) [6;28]. At the thoracolumbar zone cross section of spinal canal shows the transition of an oval format (thoracic spine) to a triangular form with rounded angles (lumbar spine) [28]. The diameter of the spinal cord has a mean caliber of 10 mm, although it might be broader at the thoracolumbar region. Normally this results in considerable spare room around the spinal cord. In 1998 de Klerk et al. reported about a retrospective study of 42 trauma patients with initial spinal canal stenosis of more than 25% that were treated conservatively [22]. Computerized tomography in his study was performed at 12 months to 108 months after the injury. One of their conclusions was that conservative treatment is followed by a marked degree of spontaneous restitution of the deformed spinal canal. They showed that the higher the initial percentage of canal stenosis, the greater the spontaneous reduction. Age at the time of injury was inversely correlated with the reduction in the percentage of spinal canal stenosis. No correlation was found considering the spontaneous reduction of the spinal canal stenosis and the time gone by since the injury. This suggests that the changes have occurred within the first 12 months. In 1992 Gertzbein showed a reduction of the preoperative canal encroachment by distraction forces, delivered by an internal fixator, from 54% to 40% only, but a selection of patients operated within the first four days after their injury showed a reduction from 56% to 38% [16]. The importance of the sagittal alignment with respect to the forces that act on the fracture parts has to be stressed. The antikyphosing reduction is reflected in an increase of the postoperative intervertebral angle [27]. The rigid fixation of the annulus to the upper and lower fragment parts will certainly add to the reduction forces. Segmental stability by rigid fixation will add to the persistence of the reduction force. In our series the lateral radiographs show that in only 18.4% of the cases canal encroachment could be visualized postoperatively. It seems logical to measure the midsagittal diameter in the lateral plain radiographs, but in contrast to the dorsal part of the vertebral body the arch can not be distinguished enough in the lateral radiograph to gain reliable measurements. The incidence of spinal cord injury with spine fractures and dislocations is approximately 14% of the total as ascertained from a survey of these injuries in Northern California [31]. Injuries to the cervical spine most often produced neurological damage, the incidence of neurological deficit being 39% [31]. Patients sustaining fractures of the vertebral bodies and posterior elements with some degree of malalignment of the spine even had a 61% incidence of neurological deficit. In our series we found more than 20% neurological deficit in patients with 1/3 to 2/3 spinal canal narrowing, but we saw 4 patients with a high degree of traumatic canal narrowing (>2/3) without any neurological deficit (Table 3). This


65

is surprising, because canal narrowing influences the risk of neurological involvement. L1-fractures with more than 65% of canal narrowing have a high probability to be accompanied by neurological deficit [18;33]. Conclusions In this study we showed that the initial restoration of the spinal canal by indirect manipulation and disco-ligamentotaxis is incomplete. The posterior segmental height increases in the perioperative period and clearance of the spinal canal is observed in about 75% of the patients with traumatic canal narrowing in the plain lateral radiographs. Later ligamentotaxis does not play a role anymore, because the PSH even diminishes. The clearance of the spinal canal as interpreted from lateral radiographs continues in the course of follow-up. At two years after operation about 97% of all patients with burst fractures have a completely free spinal canal in the lateral radiograph, but not all fractures and fracture parts can be identified in plain lateral radiographs. So plain lateral radiographs seems to overestimate the process of remodelling. This study provides clinical data for the description of (partial) spontaneous remodelling. This phenomenon can be observed in plain lateral radiographs, but not as accurate as in CT-scans. References 1. Aebi M, Etter C, Kehl T, Thalgott J (1987) Stabilization of the lower thoracic and


2. Boerger TO, Limb D, Dickson RA (2000) Does 'canal clearance' affect neurological outcome after thoracolumbar burst fractures? J Bone Joint Surg Br 82:629-635

3. Chakera TM, Bedbrook G, Bradley CM (1988) Spontaneous resolution of spinal

canal deformity after burst-dispersion fracture. Am J Neuroradiol 9:779-785

4. Dai LY (2001) Remodelling of the spinal canal after thoracolumbar burst fractures. Clin Orthop 382:119-123

5. Daniaux H (1982) Technik und erste Ergebnisse der transpedikulären Spongiosaplastik bei Kompressionsbrüchen im Lendenwirbelsäulenbereich. Acta Chir Austriaca, supplement 43:79


66

6. Delmas A, Pineau H (1970) Sur le canal vertébral de la colonne lombaire. CR Assoc Anat 145:135-138

7. Dick W (1984) Osteosynthese schwerer Verletzungen der Brust- und Lendenwirbelsäule mit dem Fixateur interne. Langenbecks Arch Chir 364:343-346

8. Dick W (1987) Innere Fixation von Brust- und Lendenwirbelfrakturen. In: Burri C, Harder F, and Bauer R (eds) Aktuelle Probleme in Chirurgie und Orthopädie, 2(28) Verlag Hans Huber, Bern, Switzerland, pp 1-137

9. Dick W (1987) The "fixateur interne" as a versatile implant for spine surgery. Spine 12:882-900

10. Dick W (1992) Dorsale Stabilisierung bei Brust- und Lendenwirbelverletzungen. Langenbecks Arch Chir Suppl Kongressbd 290-292

11. Dick W, Kluger P, Magerl F, Woersdorfer O, Zach G (1985) A new device for internal fixation of thoracolumbar and lumbar spine fractures: the 'fixateur interne'. Paraplegia 23:225-232

12. Fontijne WP, de Klerk LW, Braakman R, Stijnen T, Tanghe HL, Steenbeek R, van Linge B (1992) CT scan prediction of neurological deficit in thoracolumbar burst fractures. J Bone Joint Surg Br 74:683-685

13. Frank E, Bonsell S (1994) The accuracy of anterior-posterior measurements in the assessment of spinal canal compromise in burst fractures. Neurol Res 16:410-412

14. Fredrickson BE, Mann KA, Yuan HA, Lubicky JP (1988) Reduction of the

intracanal fragment in experimental burst fractures. Spine 13:267-271

15. Frobin W, Brinckmann P, Biggemann M, Tillotson M, Burton K (1997) Precision measurement of disc height, vertebral height and sagittal plane displacement from lateral radiographic views of the lumbar spine. Clin Biomech 12:S1-S63

16. Gertzbein SD, Crowe PJ, Fazl M, Schwartz M, Rowed D (1992) Canal clearance in burst fractures using the AO internal fixator. Spine 17:558-560

17. Ha KI, Han SH, Chung M, Yang BK, Youn GH (1996) A clinical study of the natural remodelling of burst fractures of the lumbar spine. Clin Orthop 323:210-214

18. Hashimoto T, Kaneda K, Abumi K (1988) Relationship between traumatic spinal canal stenosis and neurologic deficits in thoracolumbar burst fractures. Spine 13:1268-1272


67

19. Heggeness MH, Doherty BJ (1997) The trabecular anatomy of thoracolumbar vertebrae: Implications for burst fractures. J Anatomy 191:309-312

20. Isomi T, Panjabi MM, Kato Y, Wang JL (2000) Radiographic parameters for evaluating the neurological spaces in experimental thoracolumbar burst fractures. J Spin Disord 13:404-411

21. Johnsson R, Herrlin K, Hagglund G, Stromqvist B (1991) Spinal canal remodelling after thoracolumbar fractures with intraspinal bone fragments. 17 cases followed 1-4 years. Acta Orthop Scand 62:125-127

22. Klerk de LW, Fontijne WP, Stijnen T, Braakman R, Tanghe HL, van Linge B (1998) Spontaneous remodelling of the spinal canal after conservative management of thoracolumbar burst fractures. Spine 23:1057-1060

23. Kuner EH, Kuner A, Schlickewei W, Mullaji AB (1994) Ligamentotaxis with an internal spinal fixator for thoracolumbar fractures. J Bone Joint Surg Br 76:107-112

24. Kuner EH, Kuner A, Schlickewei W, Wimmer B (1992) Die Bedeutung der

Ligamentotaxis für die Fixateur-interne-Osteosynthese bei Frakturen der Brust- und Lendenwirbelsäule. Chirurg 63:50-55

25. Kuner EH, Schlickewei W, Kuner A, Hauser U (1997) Restoration of the spinal canal by the internal fixator and remodelling. Eur Spine J 6:417-422

26. Leferink VJM, Nijboer JMM, Zimmerman KW, Veldhuis EFM, ten Vergert EM, ten Duis HJ (2002) Thoracolumbar spinal fractures: segmental range of motion after dorsal spondylodesis in 82 patients: a prospective study. Eur Spine J 11:2-7


28. Louis R (1983) Surgery of the spine: surgical anatomy and operative approaches. Springer-Verlag, Berlin, Heidelberg, New York, pp 77-83


30. Panjabi MM, Kifune M, Wen L, Arand M, Oxland TR, Lin RM, Yoon WS, Vasavada A (1995) Dynamic canal encroachment during thoracolumbar burst fractures. J Spin Disord 8:39-48


68

31. Riggins RS, Kraus JF (1977) The risk of neurologic damage with fractures of the vertebrae. J Trauma 17:126-133

32. Scapinelli R, Candiotto S (1995) Spontaneous remodelling of the spinal canal after burst fractures of the low thoracic and lumbar region. J Spin Disord 8:486-493

33. Sjöström L, Karlström G, Pech P, Rauschning W (1996) Indirect spinal canal decompression in burst fractures treated with pedicle screw instrumentation. Spine 21:113-123

34. Wessberg P, Wang Y, Irstam L, Nordwall A (2001) The effect of surgery and

remodelling on spinal canal measurements after thoracolumbar burst fractures. Eur Spine J 10:55-63

THORACOLUMBAR FRACTURES 5 SEGMENTAL RANGE OF MOTION

69

Chapter 5

Thoracolumbar spinal fractures: segmental range of motion after dorsal spondylodesis in 82 patients: a prospective study V.J M. Leferink, J.M.M. Nijboer, K.W. Zimmerman, E.F.M. Veldhuis, E.M. ten Vergert and H.J. ten Duis Published in European Spine Journal (2002) 11:1:2-7 © Springer Introduction Prevention of arthritic pain resulting from movement in subluxated or degenerated thoracolumbar facet joints is one of the main goals of spondylodesis of fractured vertebrae. Little is known about the effect of dorsal spondylodesis after trauma in terms of movement of the adjacent facet joints [5;6]. In vitro stabilisation of calf lumbosacral spine specimens has shown increased mobility of the remaining adjacent segments after dorsal internal fixation [26]. In the canine model, the adjacent segments have shown increased mobility while walking on a treadmill, 12 weeks post fusion. In patients with a fracture, we performed a fusion at the level of the disrupted (upper) endplate, in order to reduce the movements in the facet joints at this level [13]. In order to study the effect of dorsal spondylodesis on intervertebral movement, we measured the sagittal range of motion (ROM) in the vertebral segments above and below the fractured vertebral body to answer the following questions:


70

• Does dorsal spondylodesis of one segment also cause loss of ROM at other segments, or does it result in increased ROM at the surrounding segments?

• Does dorsal spondylodesis result in ankylosis of the affected intervertebral disc space?

Materials and methods Between February 1991 and May 1996, 93 consecutive patients with a fracture of the thoracolumbar spine were treated operatively at the traumatology department of the University Hospital Groningen. Clinical records, operative records, follow-up records and radiographs were analysed. Because of small numbers of patients with T9-T11 and L4-L5 fracture levels, only the data of the T12, L1, L2 and L3 fractures were analysed (n=82). According to Magerl's comprehensive classification, 67 type A fractures, 8 type B fractures and 6 type C fractures were treated; one fracture could not be classified because of missing data (Table 1) [15].

Table 1 Comprehensive classification in 82 patients

Level n Type A Type B Type C Unknown

T12 18 16 0 1 1 L1 42 36 4 2 0 L2 17 11 3 3 0 L3 5 4 1 0 0

Table 2 Accompanying traumatic lesions in 82 patients

Accompanying lesions n %

No other lesions 46 56.0 Neurological deficit 9 11.0 Unrelated other lesions 24 29.3 Unrelated other lesions as well as neurological deficit 3 3.7

Several related and unrelated traumatic diagnoses were made. Cone/caudal or root lesions were identified in 14.6% of the patients (Table 2). Fracture reduction and fixation was performed by means of dorsal instrumentation with the Dick internal


71

fixator or Universal Spine System (Synthes®), combined with two-sided transpedicular cancellous bone graft and dorsal spondylodesis following the methods of Dick, Daniaux and Blauth [2;3;5;6]. Fracture reduction (angular reduction and distraction) was obtained by indirect manipulation using pedicle screws. Cancellous bone was taken from the dorsal iliac bone near the sacro-iliac joint and was put in the reduced vertebral body transpedicularly [3]. The facet joints at the level of the traumatised disc were opened and the cartilage was removed. Cancellous bone was packed around the joints at the dorsolateral side [2]. No ventral operations, discectomies or laminectomies were performed.

Fig.1 Intervertebral angles at different segments. I-IV are the measured segments

Postoperatively, all patients were transferred to a rehabilitation centre, where they were allowed to walk after 10 days in a simple reclination corset, which was worn for 9 months. In the final 3 months, patients only wore the corset in the daytime. After 9 months the implants were removed. Three months later the patients were instructed to resume all former activities. Two years after the spondylodesis, flexion and extension radiographs were obtained, which were analysed in this study.


72

The intervertebral angles at the two segments cranial to the fractured vertebral body and at the two segments caudal to the fractured vertebral body were measured (Fig.1). Measurements were made on plain radiographs while patients were standing with a maximum voluntarily flexed and extended spine. The X-ray beam was directed at the fracture level. Patient data were statistically analysed using SPSS® in comparison to normal movements (Table 3) and with respect to the levels that were intentionally immobilised by dorsal spondylodesis [8].

Table 3 Normal range of motion (ROM, in degrees),

in spinal segments. SD = standard deviation *Presumed values

Segment ROM

(degree) SD

(degree) T10-T11 5.0 *2 T11-T12 5.0 *2 T12-L1 *8.0 *2.2 L1-L2 11.9 2.27 L2-L3 14.5 2.29 L3-L4 15.3 2.04 L4-L5 8.2 2.99

Measured data were compared to normal values and zero-distributions by the Kolmogorov-Smirnov (K-S) one-sample test, a statistical so-called goodness-of-fit test, like the chi-square test. The K-S test determines whether a sample of observed scores can reasonably be thought to have come from a population of scores with a theoretical distribution; in this case observed scores of differences and the zero-distribution. The K-S test is preferred in small numbers of measurements of a continuous (at least ordinal) scale, as in our study, because, unlike the chi-square test, it does not need to combine categories, and thus avoids losing information. The exact K-S test is definitely more powerful than the asymptotic chi-square test in tests with small samples. The tests are equally powerful with (very) large samples [27].

Results Mean patient age in the T12 fracture group was 41.7 years, and this turned out to be slightly higher than in the T11, L1 and L2 fracture groups (35.5, 37.8 and 37.7 years respectively; p<0.05, Wilcoxon test).


73

T12 fractures The results of the measurements of the intervertebral angles at maximum flexion and maximum extension around the T12 vertebral body 2 years after the initial operation are listed in Table 4. At segment T11-T12, no significant movement could be determined. The other segments (T10-T11, T12-L1 and L1-L2) showed a range of motion of 1.50°, 4.61° and 6.56° respectively (p<0.05, p<0.05, p<0.001 respectively, compared to the zero-distribution, K-S test). L1 fractures The ROM values of the intervertebral segments around the first lumbar vertebral body are listed in Table 5. In 42 L1 fractures the mean ROM at level T12-L1 was 0.14° (±1.30°), i.e. no significant movement, at a level where the normal ROM is between 5° and 12°. At the surrounding levels, T11-T12, L1-L2 and L2-L3, the ROM was 2.32°, 4.67° and 9.29° respectively (p<0.001, compared to the zero-distribution, K-S test). L2 fractures No significant ROM was shown at the L1-L2 level, but significant ROM was found at segments T12-L1, L2-L3 and L3-L4 (p<0.001, p=0.005, K-S test, Table 6). L3 fractures Results concerning the third lumbar vertebral body are shown in Table 7. In five patients no significant ROM could be determined in segment L1-L2 (p=0.058) and segment L2-L3 (P=0.182), whereas L3-L4 and L4-L5 were mobile (p<0.005, p<0.05; K-S test). Differences compared to normal ROM values Relative differences to the normal values of segments related to the fracture level are summarised in Table 8 and Fig.2. At all fracture levels, the ROM of the segment adjacent to the upper endplate of the fractured body did not differ from zero (K-S test). Discussion Measurements Radiological measurements in analysis of segmental ROM can best be done after implantation of small tantalum balls, for example during posterolateral fusion in patients suffering from spondylolysis, spondylolisthesis, lumbar disc disorders or facet joint arthritis [9;25]. For the purpose of our study, i.e. merely depicting


74

more or less ROM in single segments in relation to operative fracture therapy, reliable comparison of the intervertebral angles can be done without implantation of radiopaque material. Non-radiological measurement, for example with a Myrin inclinometer, is unreliable in single segments [16]. Measurements of segmental ROM using skin markers, and calculating their relative position, have been found to correlate highly with the radiological position of the lumbar vertebrae [12].

Table 4 Mobility (in degrees) 2 years after operative treatment of T12 fracture in 18 patients (SEM standard error

of the mean, CID confidence interval of the difference)

Paired differences 95% CID Segment

Mean SD SEM Lower Upper t df p

T10-T11 1.50 2.16 0.54 0.35 2.65 2.777 15 0.014 T11-T12 0.33 1.33 0.31 -0.33 0.99 1.065 17 0.302 T12-L1 4.61 7.05 1.66 1.11 8.12 2.776 17 0.013 L1-L2 6.56 4.02 0.95 4.56 8.55 6.922 17 <0.001

Table 5 Mobility (in degrees) 2 years after operative treatment

of L1 fracture in 42 patients



T11-T12 2.32 3.69 0.61 1.09 3.55 3.832 36 <0.001 T12-L1 0.14 1.30 0.20 -0.26 0.55 0.713 41 0.480 L1-L2 4.67 0.92 0.61 3.44 5.89 7.707 41 <0.001 L2-L3 9.29 4.56 4.5 7.70 10.88 11.894 33 <0.001

Table 6 Mobility (in degrees) 2 years after operative treatment

of L2 fracture in 17 patients



T12-L1 3.62 3.10 0.86 1.74 5.49 4.209 12 0.001 L1-L2 0.00 1.77 0.43 -0.91 0.91 0.000 16 1.000 L2-L3 6.65 4.77 1.16 4.20 9.10 5.747 16 <0.001 L3-L4 6.50 7.11 1.90 2.39 10.61 3.420 13 0.005


75

Table 7 Mobility (in degrees) 2 years after operative treatment of L3 fracture in 5 patients



L1-L2 4.8 4.1 1.83 -0.27 9.87 2.626 4 0.058 L2-L3 0.50 0.58 0.29 -0.42 1.42 1.732 3 0.182 L3-L4 10.4 3.6 0.60 5.96 14.84 6.500 4 0.003 L4-L5 12.3 5.1 2.56 4.10 20.40 4.782 3 0.017

New computer-aided instruments, like the SpinalMouse®, combine the data concerning the relative position of the spinal vertebrae with a database of segmental heights and age-specific characteristics. This method will need further research in clinical studies to prove the reliability of external measuring of the segmental angles [24]. Normal movement Little is known about the ROM of the uninjured spine, or of the spine after trauma in general and after operative treatment for spinal fracture in particular. Normal ROM was studied in the 1970s by Louis, White and Panjabi, and resulted in the creation of lists of segmental ROM, as published by Louis [14] (Table 3). These studies revealed the ROM of the total spine; however, the segment T12-L1 has never been studied or published. For practical and mathematical reasons, we assumed this level to have a segmental ROM of 8°, because the facets of T12-L1 have a lumbar orientation, and one would expect the ROM to fit in with the ROM of the other lumbar facet joints [8;19]. Movement of the spinal column after fracture stabilisation Segments with endplate disruption show less mobility than segments without endplate disruption, especially when a formal fusion is present [13]. Although in computer models an internal fixator reduces stress in the bridged discs, and only minimally influences the stresses in adjacent discs, one cannot expect that no changes at all will occur [22]. Recent magnetic resonance imaging (MRI) studies have given some insight into what might happen to the intervertebral disc in the follow-up of operatively and non-operatively treated thoracolumbar fractures. MRI showed that the disc may rupture or may sink through the fractured endplate into the vertebral body. An increase in kyphosis was suggested to be the effect of gradual creeping of the disc into the fractured endplate. Secondary degenerative aspects like desiccation of the nucleus pulposus were not visualised [18].


76

One should realise that segmental flexion and extension are merely the result of a sliding movement in the slightly curved facets. This rotational movement in a sagittal plane has a centre of motion just underneath the cranial endplate [20].

Table 8 Mobility (in degrees) in segments above and below the fracture level (See Fig.1 for segment numbering)

* Significant difference to the zero-distribution

Fracture T12 Fracture L1 Fracture L2 Fracture L3 Segment

Mean SEM Mean SEM Mean SEM Mean SEM

Lindsey total motion

[13] I *1.50 0.54 *2.32 0.61 *3.62 0.86 4.80 1.83 3.22 II 0.33 0.31 0.14 0.20 0.00 0.43 0.50 0.29 1.34 III *4.61 1.66 *4.67 0.61 *6.65 1.16 *10.4 0.60 3.08 IV *6.56 0.95 *9.29 4.50 *6.50 1.90 *12.3 2.56 6.88

Fig.2 Mobility in adjacent segments 2 years postoperatively (See Fig.1 for segment numbering)

I II III IV

T h 1 2

L 1

L 2L 3

0

2

4

6

8

1 0

1 2

1 4graden

segm entfracture level


77

The movement changes when the form of the disc changes. Prevention of motion in the intervertebral facet joints will prevent movement in the intervertebral (disc) space, and will therefore prevent flexion and extension at this level. However, osseous healing of the dorsal fusion of the facet joints does not automatically imply "total stabilisation" of the segment [17]. Moreover, resection of the disc, combined with transpedicular interposition of cancellous bone graft in order to fuse the vertebral bodies results in failure of radiological fusion in 66% of the cases, and in some residual segmental movement [10]. The performance of disc resection is questionable anyhow, since MRI investigations have shown only minor structural damage to the nucleus pulposus and the fibrous disc structures [23]. ROM at adjacent segments Accelerated degeneration of the facets of the segment adjacent to the fusion level, when performed for low back pain, has been described [11]. Hypertrophic degenerative arthritis, as well as severe disc degeneration (without herniation) [3] was found responsible for the recurrence of complaints after fusion. It is not certain whether this category of patients with degenerative disease can be compared to trauma patients [11]. All evaluated segments were less mobile than normal values. This observation was also recorded by Lindsey et al. [13]. It is therefore reasonable to conclude that compensational movement at adjacent segments does not occur, although increased mobility in adjacent segments was shown in the canine model 12 weeks post fusion [4]. The reason for this may be fibrosis by trauma or operative procedure, but it may also be an effect of prolonged partial immobilisation, pain or even psychosomatic mechanisms. We advised 9 months of external immobilisation to protect the patient against implant failure. However, Rohlmann and co-workers recently showed that braces do not reduce the load on internal spinal fixators [21]. So it seems reasonable that shorter immobilisation should be considered in order to optimise the ROM of the unstabilised segments I and IV (Fig.1). No actual ROM values of spinal segments in comparable groups have been published in the literature. We compared our data to the ROM values supplied by White and Panjabi [14]. In a retrospective study about this subject, Lindsey compared the residual segmental mobility in 16 patients with fusion with that in 43 patients without fusion. At all levels, fused segments showed less mobility than unfused segments, and unfused levels were less mobile than normal segments. The mean ROM values of the different segments as found by Lindsey are listed in Table 8 [13]. In our data we observed that, besides the intended loss of motion in the upper disc segment caused by the spondylodesis, more than 50% of the ROM of both spared adjacent levels is lost as well in T10, L1 and


78

L2 fractures. This loss accounts for 7°-13°, which is comparable to a complete loss of ROM of a second segment. Paradox or inverse segmental movement was measured in the upper endplate segment in some patients. Instead of the movement observed normally, a small movement in the opposite direction was measured in these patients. We did not find an explanation for this phenomenon. A similar phenomenon has been described in the cervical spine as a physiologic condition. In the last part of the flexion motion, the C0-C1 segment does not flex further, but extends a little [1;7;20]. Conclusion Dorsal spondylodesis at the level of the disturbed endplate in thoracolumbar spinal fractures leads to immobility in the affected segment, measured on flexion-extension radiographs 2 years after primary operative treatment. More than 50% loss of motion in the two adjacent levels is equivalent to complete loss of ROM in a second segment. References 1. Arlen A (1977) The paradox movement of the atlas in functional diagnostics of the

cervical spine. Man Med 15:16-22 2. Blauth M, Bastian L, Jeanneret B, Knop C, Moulin P, Müller-Vahl H, Schmidt U,

Schratt HE, Wippermann B (1998) Wirbelsäule. In: Tscherne H, Blauth M (eds) Tscherne Unfallchirurgie, vol 3. Springer, Berlin Heidelberg New York, pp 241-372

3. Daniaux H (1982) Technik und erste Ergebnisse der transpedikulären

Spongiosaplastik bei Kompressionsbrüchen im Lendenwirbelsäulenbereich. Acta Chir Austr 43 [Suppl]:79

4. Dekutoski MB, Schendel MJ, Ogilvie JW, Olsewski JM, Wallace LJ, Lewis JL

(1994) Comparison of in vivo and in vitro adjacent segment motion after lumbar fusion. Spine 19:1745-1751


12:882-900 6. Dick W, Kluger P, Magerl F, Woersdorfer O, Zach G (1985) A new device for

internal fixation of thoracolumbar and lumbar spine fractures: the 'fixateur interne'. Paraplegia 23:225-232


79

7. Dvorak J, Froehlich D, Penning L, Baumgartner H, Panjabi MM (1988) Functional

radiographic diagnosis of the cervical spine: flexion/extension. Spine 13:748-755 8. Dvorak J, Panjabi MM, Chang DG, Theiler R, Grob D (1991) Functional

radiographic diagnosis of the lumbar spine. Flexion-extension and lateral bending. Spine 16:562-571

9. Johnsson R, Selvik G, Stromqvist B, Sunden G (1990) Mobility of the lower lumbar

spine after posterolateral fusion determined by röntgen stereo photogrammetric analysis. Spine 15:347-350

10. Knop C, Blauth M, Bastian L, Lange U, Kesting J, Tscherne H (1997) Frakturen der

thorakolumbalen Wirbelsäule. Spätergebnisse nach dorsaler Instrumentierung und ihre Konsequenzen. Unfallchirurgie 100:630-639

11. Lee CK (1988) Accelerated degeneration of the segment adjacent to a lumbar fusion.

Spine 13:375-377 12. Lee YH, Chiou WK, Chen WJ, Lee MY, LinYH (1995) Predictive model of

intersegmental mobility of lumbar spine in the sagittal plane from skin markers. Clin Biomech 10:413-420

13. Lindsey RW, Dick W, Nunchuck S, Zach G (1993) Residual intersegmental spinal

mobility following limited pedicle fixation of thoracolumbar spine fractures with the fixateur interne. Spine 18:474-478

14. Louis R (1983) Surgery of the spine: surgical anatomy and operative approaches.

Springer, Berlin Heidelberg New York, pp 63-71 15. Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S (1994) A comprehensive

classification of thoracic and lumbar injuries. Eur Spine J 3:184-201 16. Mellin G (1986) Measurement of thoracolumbar posture and mobility with a Myrin

inclinometer. Spine 11:759-762 17. Olsson TH, Selvik G, Wilner S (1977) Mobility in the lumbosacral spine after fusion

studied with the aid of roentgen stereophotogrammetry. Clin Orthop 129:181-190 18. Öner FC, van der Rijt RR, Ramos LM, Dhert WJ, Verbout AJ (1998) Changes in the

disc space after fractures of the thoracolumbar spine. J Bone Joint Surg Br 80:833-839

19. Panjabi M, Oxland T, Takata K, Goel VK, Duranceau J, Krag MH (1993) Articular

facets of the human spine. Spine 18:1298-1310


80

20. Penning L (1998) Normale bewegingen van de hals- en lendenwervelkolom:

vergeleken met elkaar en met de wervelkolom van viervoetige zoogdieren. Lemma, Utrecht, pp 30-47

21. Rohlmann A, Bergmann G, Graichen F, Neff G (1999) Braces do not reduce loads

on internal spinal fixation devices. Clin Biomech 14:97-102 22. Rohlmann A, Calisse J, Bergmann G, Weber U (1999) Internal spinal fixator

stiffness has only a minor influence on stresses in the adjacent discs. Spine 24:1192-1195

23. Rudig L, Runkel M, Kreitner KF, Seidel T, Degreif J (1997)

Kernspintomographische Untersuchung thorakolumbaler Wirbelfrakturen nach Fixateur-interne-Stabilisierung. Unfallchirurgie 100:524-530

24. Seichert N, Baumann M, Senn E, Zuckriegl H (1994) Die Rückenmaus - ein analog-

digitales Messgerät zur Erfassung der sagittalen Rückenkontur. Phys Rehabil Kur Med 4:35-43

25. Selvik G (1990) Roentgen stereophotogrammetric analysis. Acta Radiol 31:113-126

26. Shono Y, Kaneda K, Abumi K, McAfee PC, Cunningham BW (1998) Stability of

posterior spinal instrumentation and its effects on adjacent motion segments in the lumbosacral spine. Spine 23:1550-1558

27. Siegel S, Castellan NJ (1988) The Kolmogorov-Smirnov one-sample test. In:

Nonparametric statistics for the behavioral sciences. McGraw Hill, New York, pp 51-64

THORACOLUMBAR FRACTURES 6 FUNCTIONAL OUTCOME

81

Chapter 6

Functional outcome after 3-8 in 19 patients with thoracolumbar burst fractures, treated with dorsal instrumentation and fixation, transpedicular cancellous bone grafting and dorsal spondylodesis V.J.M. Leferink, H.J.E. Keizer, J.K. Oosterhuis, C.K. van der Sluis, H.J. ten Duis Accepted European Spine Journal October 2002 Introduction Clinical studies on functional outcome of the treatment of trauma patients are relatively scarce. As a consequence, little is known about the degree of disability after trauma in general and after spinal fractures in particular [37]. Outcome after thoracolumbar spinal fractures is generally seen as the radiological result of the treatment, by some authors referred to as surrogate outcome [18]. Functional outcome after operative therapy is seldom investigated. The present study describes the functional outcome of operatively treated patients with a thoracolumbar burst fracture, operatively treated with pedicle screw internal fixation, transpedicular cancellous bone grafting, and dorsal spondylodesis [3-5]. Ventral fusion was not pursued. The aim of the study is to develop insight in the impairments in these patients, and also in their ability to participate in daily living, in their possibilities to return to work and in their quality of life as defined by the World Health Organization (WHO) in the International Classification of Function, Disability and Health (ICF) [2;36].


82

Materials and Methods Patients operated for a type A fracture (Comprehensive Classification CC) [23] of the thoracolumbar spine (T10-L4) between 1993 and 1998 in the University Hospital Groningen, the Netherlands, aged between 18 and 60 years, without neurological deficit, were included in the study. Exclusion criteria were spinal disorders in the medical history (including low back pain previously treated by a medical specialist), pathological fractures and insufficient command of the Dutch language. The Medical Ethics Committee of the University Hospital Groningen approved the study protocol (Nr. 99/12/206). Within these criteria a group of 35 patients could be identified. Eleven patients did not respond, four refused to join the study, and one patient agreed to participate in the study, but did not show up at several appointments. Eventually nineteen patients joined the study. The mean age of the respondents was 40.5 years (range 24-57, SD 10.3), ten patients were male and nine were female. Etiologic factors were traffic accidents (n=3), accidental fall from height (n=10) and accidents of sports (horse riding, motor sports and parachute jumping) (n=6). Fracture levels are merely T12 and L1 (Table 1) and the CC of the type A fractures shows 68% A3 fractures (Table 2).

Table 1 Level of spinal fracture in 19 patients

Level n

T10 1 T11 0 T12 8 L1 8 L2 0 L3 2 L4 0 Total 19

Three patients had multiple fractures at other locations. In all patients the Injury Severity Score (ISS) was derived from the codes of the 9th version of the International Classification of Diseases (ICD-9) [12]. Mean ISS was 10.6 (range 9-22). One patient suffered from diabetes mellitus, two from cardiovascular ischaemic disease and two from chronic obstructive pulmonary disease. Respondents did not differ in fracture severity, co-morbidity, age and gender from non-respondents.


83

Operative treatment consisted of fracture reduction and internal fixation using the Universal Spine System (USS, Synthes®) within 8 days, combined with transpedicular cancellous bone grafting to the fractured vertebral body. Dorsal spondylodesis was performed only in A3-type fractures, at the level of the injured endplate. No ventral fusion operations, discectomies and laminectomies were performed. Early postoperative complications were seen in three patients (deep wound infection after implant removal (n=1), temporary bladder dysfunction (n=1), and superficial decubital ulcer (n=1)). Within a week after surgery all patients were transferred to a rehabilitation centre, where they were mobilised with a simple thoracolumbar support orthesis (reclination-brace) within two weeks after operation. In 18 patients the implants were removed at 9 months after the primary operation. In one patient (A3.3 fracture) the implants were left in place, because both segments were stabilised additionally by dorsal spondylodesis.

Table 2 Comprehensive classification in 19 patients

Class n Subclass n

A1.1 0 A1.2 1

A1

2

A1.3 0 A2.1 0 A2.2 0

A2

4

A2.3 4 A3.1 9 A3.2 4

A3

13

A3.3 1 Functional outcome in this study is defined by using the concepts described in the ICF of the WHO. The ICF recognizes restrictions in body function and structure, restrictions in activities, and restrictions in participation/quality of life. In this study restrictions in body function and structure are described by objective findings in radiographical analysis in the course of follow up (anterior wedge angle (AWA) and regional angle (RA)) and in testing the physical capacity of the respondents after 3-8 years (static and dynamic lifting tests, and an ergometric exercise test, see intermezzo p.84). The anterior wedge angle (AWA) and regional angle (RA) were measured at 0, 1, 9, and 24 months in plain transverse radiographs.


84

Intermezzo-physical tests _______________________________________________________________________ Dynamic lifting test The patient is asked to lift a box with a weight from the floor to a 75-cm high table and back to the floor again four times in 20 seconds. The starting weight is for males 5.85 kg, and for females 3.6 kg. After 20 seconds of lifting exercises the patient rests during 20 seconds. After each rest the patient decides if he will stop or go on with a heavier weight (males 4.5 kg more, females 2.5 kg more) [24]. The personal maximum weight is calculated with the formula:

Wmax= 0.6 x Body mass The test is stopped when the cardiac frequency rises above the personal maximum value, when the personal maximum lifting weight is achieved, when the patient cannot perform the exercise within 20 seconds or when the patient wants to stop for other reasons. The personal maximum cardiac frequency (MaxCF) is 85% of the age-related maximum cardiac frequency (MaxCf=(220-age) x 0,85). The highest lifted weight according to the dynamic lifting test is called the maximum lifted load. The maximum lifted load is then compared to the Dutch National Institute for Occupational Safety and Health (NIOSH) norm, according to which norm a person is allowed to lift a maximum load of 14.8 kg during an eight hour working day [24;25;40]. The loading degree is calculated as follows:

Loading degree =max lifted load x 100% 14.8 kg

Static lifting test The static lifting test consists of three tests. In a leg lift test (NIOSH-norm 23 kg), a trunk lift test (NIOSH-norm 14 kg) and an arm lift test (NIOSH-norm 15 kg) the patient is asked to lift an Acceptable Maximum Effort (AME) in three positions [11]. Between the tests one minute of rest is allowed. The test is repeated with either the NIOSH-norm in case the patient lifted a higher weight than the NIOSH-norm at the first attempt, or with 50% of the lifted weight in case the NIOSH-norm was not reached. The loading degree in the leg, trunk and arm lift was calculated as follows:

Loading degree = AME Second lift (NIOSH-norm or 50%AME)

Ergometric test The VO2-max (maximum oxygen uptake in litre per minute) was calculated after a sub maximal bicycle ergometry, in which the cardiac frequency (beats per minute bpm), the working load (Watt) and the number of revolutions (per minute) were measured [40]. The formula is: VO2-max(male)= 174.2 x max working load + 4020

103.2 x cardiac frequency – 6299 VO2-max(fem)= 163.8 x max working load + 3780

104.4 x cardiac frequency – 7514 The starting load at 60 revolutions per minute is 50% of the Lean Body Mass (LBM) for two minutes. The load is raised to 150%, 200% and 250% of the LBM until the cardiac rate is 120 bpm or more. The highest load is performed for 6 minutes. The lifting test and ergometric test findings were compared to normal values. _______________________________________________________________________


85

The differences measured in each period were calculated: the perioperative period (period I: t=0-1 months), the period until implant removal (period II: t=1-9 months) and the period after implant removal (period III: t=9-24 months) [22]. Restrictions in activity and the degree of disablement are assessed by the Dutch versions of two disease specific questionnaires, the Visual Analogue Scale VAS Spine Score and the Roland Morris Disability Questionnaire (RMDQ) [12;27;28]. The VAS Spine Score, developed to be used in spinal fracture patients, asks the patient to rate the functional outcome in 19 items on an analogue 10cm visual scale. The patient's perception of pain and restriction in activities, related to problems of the back, is measured [12]. Higher scores represent better results, recalculated to percentages of the maximum score (0-100%). The RMDQ was developed to measure and register changes over time during the treatment of low back pain. The form consists of 24 statements concerning certain (restrictions of) activities, qualified as positive (restricted) or negative (not restricted) [10;11]. Lower scores on the scale of 0-24 represent better results. Finally, restrictions in participation and aspects of quality of life are described in the Short Form 36 (SF36) and in the Return To Work status. The Dutch version of the Medical Outcome Study MOS 36 item Health Survey or Short Form 36 (SF36) scale contains nine subscales: physical functioning, social functioning, role limitation due to physical problems, role limitation due to emotional problems, mental health, energy and vitality, pain, general perception of health and change in health over the past year [26;39]. Higher scores represent better results. Return To Work status The respondents were asked about their former and actual employment status. Statistical analysis RMDQ, VAS Spine Score, and SF36 were compared to reference data with the Student t-test [15;18;20;41]. Regression analysis was used to determine the correlation between changes in radiological angles and RMDQ, VAS Spine Score, and SF 36, respectively. Results Restrictions in body function and structure Radiographic evaluation shows that the preoperative mean AWA is 16.2 degree, which is reduced to 7.2 degree postoperatively. Until the end of follow up AWA gradually increases to 8.5 degree. RA is reduced from 13.2 to 5.0 degrees in


86

period I. RA increases, mostly after implant removal, to 12.9 degree at 24 months (Table 3). Main loss in AWA and RA occurred in period III (after implant removal) (Table 4). Physical capacity, measured as static and dynamic lifting strength, VO2-max, and loading degree in the bicycle ergometric test show large ranges in all categories (Table 5). Comparison to normal values of healthy probands in comparable age groups shows that arm and trunk lift is within the normal range in 87% and 80% of the patients respectively, and 53% of the patients is able to perform a leg lift load within the normal range (Table 5).

Table 3 Radiographic measurements of AWA and RA in 19 patients in the course of treatment

t (months) 0 1 9 24

AWA 16.2 7.2 7.5 8.5 RA 13.2 5.0 6.1 12.9

Table 4 Changes in AWA and RA in period I, II, and III

Period I II III

AWA difference 8.9 0.1 -1.2 RA difference 8.2 -1.1 -6.8

Table 5 Functional capacity measured by bicycle ergometric test and static and dynamic lifting test in 12-15 patients 3-8 years postop

Test n Mean SD Range %

under norm

t-test p

VO2-max (ml/min.kg) 12 34.0 6.5 20.8 45.5 8.3 0.239 Max leg lift (kg) 15 41.7 21.1 5.0 80.0 46.7 <0.001 Max arm lift (kg) 15 36.6 21.1 5.0 82.5 20.0 0.668 Max trunk lift (kg) 15 26.0 11.9 2.5 45.0 13.3 0.221

Restrictions in activities (disablements) In the RMDQ a mean score of 4.0 positive items (SD 6.0, median 1.0, range 0-20) was found (Table 6). Two out of 19 patients had an extraordinary score (18 and 20). The VAS Spine Score revealed a mean score of 79.4 (median 90.5, SD 25.0,


87

range 17.4-100) (Table 7). The mean value does show a difference to uninjured people (p=0.042). The distribution of scores is skew because of two very low scores. Restrictions in participation/quality of life After SF36 analysis no significant differences to the values of a comparable (healthy) age group could be identified in any subscale [10]. The return to work status shows that thirteen of fifteen patients (87%) with paid labour before injury had returned to work at follow up. Seven of them (47%) had arranged changes in the kind of work or in the intensity or duration of the work. One patient (7%) changed his job. Two of fifteen patients with paid labour before the injury stopped working post injury and received social security benefits. Table 6 Comparison of RMDQ values in patients with low back pain (mean score, SD)

and operatively treated spinal fractures

Positive items Author Group n

Follow-up (mean and

range in years) Mean SD

Simple low back pain (mean duration 2.3 weeks)

99 - 10.9 4.7 Leclair [20] Low back pain with radiculopathy

(mean duration 28.1 weeks) 97 - 14.2 5.2

Weinstein [41]

Conservatively treated thoracolumbar burst fractures

42 20.2 (11-55) 13.2 -

Kraemer [18]

Thoracolumbar burst fractures (operative and non-operative)

24 3.8 (2.2-7.1) 15.6 6.5

Our study Thoracolumbar burst fractures after USS

19 4.5 (2.6-7.9) 4.0 6.0

Table 7 VAS Spine Score. Mean score, standard deviation and range

in different groups of persons

Group n Mean Median SD range

Uninjured people 136 92.0 94.0 7.5 58-100 Before trauma [15] 53 89.6 95.0 14.9 21-100 At implant removal (7-13 months) [15] 51 58.3 59.0 22.2 13- 97 At follow up (23 months) [15] 53 66.1 70.0 25.0 15-100 Our study at follow up (54 months) 19 79.4 90.5 25.0 17-100


88

Table 8 Correlation between RMDQ, VAS Spine Score and SF36 Light grey= p <0.05, darkgrey= p <0.01

Intermezzo-questionnaires _________________________________________________________________ ______ Specific questionnaires for (low) back pain evaluation are numerous. For example, the Dallas Pain Questionnaire[19], Back Pain Functional Scale [34], Low Back Outcome Score [8;32], Quebec Back Pain Disability Scale [16;17], Million Questionnaire [27], Waddel Questionnaire [38], Oswestry Disability Questionnaire [6] and Roland Morris Disability Questionnaire RMDQ, derived from the Sickness Impact Profile SIP [7;30;31]. These questionnaires measure the complaints in patients with low back pain, but are scarcely used for the evaluation of trauma patients[14;18;28;32]. The Sickness Impact Profile and the SF36 [10] have seldom been used for spinal fracture treatment evaluation, but the latter shows correlation with low back pain scores after spinal surgery [9]. In a literature review in 1995 it was found that Oswestry and RMDQ had an equal validity as Million and Waddel Questionnaires in low back pain research, however the former have been used and evaluated more frequently [1]. The RMDQ was developed to measure and register changes over time during the treatment of low back pain [30;31]. Changes of the RMDQ-score are dependent on the initial RMDQ-score of the individual patient. So the initial score should be taken into account [29;35]. Comparison to the Oswestry Disability Scale showed us that the RMDQ is more reliable than the Oswestry, because Oswestry is associated with a higher frequency of left blank items and multiple response items [33]. Because of these findings we decided to use the RMDQ as back pain specific questionnaire. _______________________________________________________________________

VA

S Sp

ine

Scor

e

RM

DQ

SF36

ph

ysic

al f

unct

ioni

ng

SF36

soc

ial f

unct

ioni

ng

SF36

rol

e re

stri

ctio

n ph

ysic

al p

robl

em

SF36

vita

lity

SF36

pai

n

SF36

gen

eral

hea

lth

VAS Spine Score 1.00 -0.72 0.54 0.51 0.69 0.71 0.66 0.60 RMDQ -0.72 1.00 -0.69 -0.76 -0.83 -0.56 -0.60 -0.68 SF36 physical functioning 0.54 -0.69 1.00 0.57 0.42 0.51 0.30 0.55 SF36 social functioning 0.51 -0.76 0.57 1.00 0.71 0.56 0.81 0.85 SF36 role restriction 0.69 -0.83 0.42 0.71 1.00 0.63 0.64 0.71 SF36 vitality 0.71 -0.56 0.51 0.56 0.63 1.00 0.75 0.82 SF36 pain 0.66 -0.60 0.30 0.81 0.64 0.75 1.00 0.87 SF36 general health 0.60 -0.68 0.55 0.85 0.71 0.82 0.87 1.00


89

Discussion Functional outcome in patients with an operatively treated thoracolumbar burst fracture is a relative infrequent topic of research. In this study we try to develop insight not only in the radiological and functional impairments in these patients, as defined by the ICF of the WHO, but also in the ability to participate in daily living, like return to work and quality of life [2;36]. Restrictions in body function and structure The radiographic findings, showed similar findings in former studies: AWA and RA can be restored by indirect instrumental manipulation via pedicle screws [13;22]. AWA remains almost the same in the course of follow up, even after implant removal. RA decreases to the preoperative value in 24 months. The main part of the decrease is due to loss of intervertebral angle [22]. The effect of the segmental dorsal spondylodesis is a complete loss of range of motion of this segment[21]. At the adjacent segments a loss of 50% of normal range of motion occurs without spondylodesis. This is supposed to be the effect of trauma, operative treatment and immobilisation [21]. A high correlation was seen between RMDQ, VAS Spine Score and the SF36 subscales (Table 8). No correlation was found between preoperative AWA and RA, AWA and RA at 24 months, AWA differences and functional capacity test scores, RMDQ, VAS Spinal Score, or SF36. Three to eight years after operation the functional capacity in our study population is decreased compared to uninjured people. In testing the maximum leg lift almost 50% of the patients perform less than the lowest normal value. In the arm and trunk lift tests, the VO2-max, and the ergometry test, the patients show better scores, but variable percentages of patients score under the lowest normal values (Table 5). Although no patients with neurological deficit were included in the study and no neurological complications occurred, the test results show a decrease in performance and functional capacity especially in leg lift and bicycle load, suggesting a major effect in leg muscle performance. In the future more intensive leg muscle training in the rehabilitation program might help to overcome this effect. Restrictions in activity (disablement) In our series the mean RMDQ score is better than previous reported RMDQ scores in patients with low back pain, radiculopathy or thoracolumbar fractures at 3.8 years and at 20.2 years of follow up (p<0.001) (Table 6, see also intermezzo-questionnaires at page 88). Questionnaire data analysis of conservatively treated patients after a follow up of 11 to 55 years (mean 20 years) in 42 patients with burst fractures in 1987 showed that 88% had returned to their former job, some time after their injury [41]. Other


90

findings in these patients were: 57% never became pain-free, 90% rated some pain at follow-up and 62% rated their pain as very little. The RMDQ score in this conservatively treated group of patients was 13.2 (range 4.6-17.3 items, i.e. 55%, range 19-72%), considered by the author as low disability of the mean patient [41]. In our opinion a mean number of 13.2 positive items in the RMDQ represents a high level of disability. It is in contrast with the low number of positive items in our study. Another study showed that only 33% of patients with burst fractures returned to their previous employment [18]. This miscellaneous group of patients revealed a mean RMDQ score of 15.6 positive items (SD 6.5 items) after a follow up of 3.8 years. Findings in the return to work status will be influenced by the injury and its treatment, but as well as by the intensity and availability of social security in the country and in the studied era. Our RMDQ results of 4.0 positive items and 87% return to work are favourable, compared to findings in the literature. The RMDQ scores are much better than previous reported RMDQ scores in patients with low back pain, radiculopathy or thoracolumbar fractures at 3.8 years and 20.2 years of follow up (p<0.001). Comparison of our results of the VAS Spinal Score with the results of the patients described by Knop et al. [15] show higher values in our series. The difference in mean (79.4 versus 66.1) and median scores (90.5 versus 70) at follow up are relatively large and significant (p<0.05). Our longer follow up period can be an explanation for this difference, as well as a possible differences in injury factors (for example neurological deficit), operative treatment and other factors which are not specified in the Hannover series [15]. No correlation could be identified between radiological outcome and functional capacity tests. This suggests that there is no relation between the quality of reduction and fixation, and the functional outcome. A possible relation might be unrecognised because of relatively uniform AWA, RA, and its changings in the course of follow up. Unexplained is the lack of correlation between reduced functional capacity and restriction in activities. In our opinion the results suggest that the operative treatment of thoracolumbar fractures without anterior fusion can lead to a good functional result. Restrictions in participation/quality of life Grevitt et al. showed that the validity of the SF36 in operative spinal procedures is good [9]. In our series VAS Spinal Scores correlate with RMDQ and all but one SF36 subscales. RMDQ correlates with all SF36 subscales (Table 8). In our study the limited number of patients cause an increased risk of bias towards the results, although the non-respondents do not differ from the respondents in several ways.


91

Conclusions Evaluation of disease specific and non disease specific health status items and clinical tests show that operatively treated patients with thoracolumbar spinal burst fractures (type A) without neurological deficit, who were operatively treated with dorsal transpedicular internal fixation, transpedicular cancellous bone grafting and dorsal spondylodesis, perform like healthy people 3-8 years after injury, considering RMDQ, VAS Spine Score and SF36. These favourable results are obtained without anterior fusion. Only two of fifteen patients stopped working. About 50% of the patients had to change the intensity of their labour or the kind of work after the injury and treatment. In this matter leg (muscle) performance seems a more important factor than arm lift or overall condition (VO2-max). These impairments do not necessarily imply restrictions in activities and participation, but paying more attention to leg muscle strength in training programs seems logical. References 1. Beurskens AJ, de Vet HC, Koke AJ, van der Heijden GJ, Knipschild PG (1995)

Measuring the functional status of patients with low back pain. Assessment of the quality of four disease-specific questionnaires. Spine 20:1017-1028

2. Bickenbach JE, Chatterji S, Badley EM, Ustun TB (1999) Models of disablement, universalism and the international classification of impairments, disabilities and handicaps. Soc Sci Med 48:1173-1187

3. Blauth M, Bastian L, Jeanneret B, Knop C, Moulin P, Müller-Vahl H, Schmidt U, Schratt HE, and Wippermann B (1988) Wirbelsäule. In Tscherne H, Blauth M (eds) Tscherne Unfallchirurgie, vol 3. Springer Berlin Heidelberg New York, pp 241-372

4. Daniaux H (1982) Technik und erste Ergebnisse der transpedikulären Spongiosaplastik bei Kompressionsbrüchen im Lendenwirbelsäulenbereich. Acta Chir Austriaca, 43 [Suppl]:79

5. Dick W (1987) The "fixateur interne" as a versatile implant for spine surgery. Spine 12:882-900

6. Fairbank JC, Couper J, Davies JB, O'Brien JP (1980) The Oswestry low back pain disability questionnaire. Physiotherapy 66:271-273


92

7. Gilson BS, Gilson JS, Bergner M, Bobbit RA, Kressel S, Pollard WE, Vesselago M (1975) The sickness impact profile. Development of an outcome measure of health care. Am J Public Health 65:1304-1310

8. Greenough CG, Fraser RD (1992) Assessment of outcome in patients with low-back pain. Spine 17:36-41

9. Grevitt M, Khazim R, Webb J, Mulholland R, Shepperd J (1997) The short form-36 health survey questionnaire in spine surgery. J Bone Joint Surg Br 79:48-52

10. Jenkinson C, Coulter A, Wright L (1993) Short form 36 (SF36) health survey questionnaire: normative data for adults of working age. BMJ 306:1437-1440

11. Khalil TM, Goldberg ML, Asfour SS, Moty EA, Rosomoff RS, Rosomoff HL (1987) Acceptable maximum effort (AME). A psychophysical measure of strength in back pain patients. Spine 12:372-376

12. Kingma J, ten Vergert E, Werkman HA, ten Duis HJ, Klasen HJ (1994) A Turbo Pascal program to convert ICD-9CM coded injury diagnoses into injury severity scores: ICDTOAIS. Percept Mot Skills 78:915-936

13. Knop C, Blauth M, Bastian L, Lange U, Kesting J, Tscherne H (1997) Frakturen der thorakolumbalen Wirbelsäule. Spätergebnisse nach dorsaler Instrumentierung und ihre Konsequenzen. Unfallchirurg 100:630-639

14. Knop C, Fabian HF, Bastian L, Blauth M (2001) Late results of thoracolumbar fractures after posterior instrumentation and transpedicular bone grafting. Spine 26:88-99

15. Knop C, Oeser M, Bastian L, Lange U, Zdichavsky M, Blauth M (2001) Entwicklung und Validierung des VAS-Wirbelsäulenscores. Unfallchirurg 104:488-497

16. Kopec JA, Esdaile JM, Abrahamowicz M, Abenhaim L, Wood-Dauphinee S, Lamping DL, Williams JI (1995) The Quebec Back Pain Disability Scale. Measurement properties. Spine 20:341-352

17. Kopec JA, Esdaile JM, Abrahamowicz M, Abenhaim L, Wood-Dauphinee S, Lamping DL, Williams JI (1996) The Quebec Back Pain Disability Scale: conceptualization and development. J Clin Epidemiol 49:151-161

18. Kraemer WJ, Schemitsch EH, Lever J, McBroom RJ, McKee MD, Waddell JP (1996) Functional outcome of thoracolumbar burst fractures without neurological deficit. J Orthop Trauma 10:541-544


93

19. Lawlis GF, Cuencas R, Selby D, McCoy CE (1989) The development of the Dallas Pain Questionnaire. An assessment of the impact of spinal pain on behavior. Spine 14:511-516

20. Leclaire R, Blier F, Fortin L, Proulx R (1997) A cross-sectional study comparing the Oswestry and Roland-Morris Functional Disability scales in two populations of patients with low back pain of different levels of severity. Spine 22:68-71




24. Mayer TG, Barnes D, Kishino ND, Nichols G, Gatchel RJ, Mayer H, Mooney V (1988) Progressive isoinertial lifting evaluation. I. A standardized protocol and normative database. Spine 13:993-997

25. Mayer TG, Barnes D, Nichols G, Kishino ND, Coval K, Piel B, Hoshino D, Gatchel RJ (1988) Progressive isoinertial lifting evaluation. II. A comparison with isokinetic lifting in a disabled chronic low-back pain industrial population. Spine 13:998-1002

26. McHorney CA, Ware JE, Raczek AE (1993) The MOS 36-Item Short-Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care 31:247-263

27. Million R, Hall W, Nilsen KH, Baker RD, Jayson MI (1982) Assessment of the progress of the back-pain patient 1981 Volvo Award in Clinical Science. Spine 7:204-212

28. Mumford J, Weinstein JN, Spratt KF, Goel VK (1993) Thoracolumbar burst fractures. The clinical efficacy and outcome of nonoperative management. Spine 18:955-970

29. Riddle DL, Stratford PW, Binkley JM (1998) Sensitivity to change of the Roland-Morris Back Pain Questionnaire: part 2. Phys Ther 78:1197-1207


94

30. Roland M, Morris R (1983) A study of the natural history of back pain. Part I: development of a reliable and sensitive measure of disability in low-back pain. Spine 8:141-144

31. Roland M, Morris R (1983) A study of the natural history of low-back pain. Part II: development of guidelines for trials of treatment in primary care. Spine 8:145-150

32. Shen WJ, Liu TJ, Shen YS (2001) Nonoperative treatment versus posterior fixation for thoracolumbar junction burst fractures without neurologic deficit. Spine 26:1038-1045

33. Stratford PW, Binkley J, Solomon P, Gill C, Finch E (1994) Assessing change over time in patients with low back pain. Phys Ther 74:528-533

34. Stratford PW, Binkley JM, Riddle DL (2000) Development and initial validation of the back pain functional scale. Spine 25:2095-2102

35. Stratford PW, Binkley JM, Riddle DL, Guyatt GH (1998) Sensitivity to change of the Roland-Morris Back Pain Questionnaire: part 1. Phys Ther 78:1186-1196

36. Thuriaux MC (1995) The ICIDH: evolution, status, and prospects. Disabil Rehabil 17:112-118

37. Van der Sluis CK, ten Duis HJ, Geertzen JH (1995) Multiple injuries: an overview of the outcome. J Trauma 38:681-686

38. Waddell G, Main CJ (1984) Assessment of severity in low-back disorders. Spine 9:204-208

39. Ware JE, Sherbourne CD (1992) The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 30:473-483

40. Waters TR, Putz-Anderson V, Garg A, Fine LJ (1993) Revised NIOSH equation for the design and evaluation of manual lifting tasks. Ergonomics 36:749-776

41. Weinstein JN, Collalto P, Lehmann TR (1988) Thoracolumbar "burst" fractures treated conservatively: a long-term follow-up. Spine 13:33-38

THORACOLUMBAR FRACTURES 7 GENERAL DISCUSSION

95

Chapter 7

General discussion, recent, present and future developments Introduction In Chapter 1 (General introduction) we gave an overview of the questions that we answered in chapter 2 to chapter 6. The most important answers and the conclusions that were drawn from these answers will be discussed in this chapter because they imply important consequences for future management of spinal fractures. Conclusions and consequences At least 30% of the B-type lesions are unrecognised without operation [20]. These are interpreted as relative simple A-type fractures in stead. The result of this misinterpretation is a higher chance for relative undertreatment. Research in the field of spinal fractures demands a good classification, but in clinical practise it is important as well. Although we have been using the comprehensive classification since 1994 [21], we often feel the need for better evaluation of the soft tissues like ligaments and discs, so that those lesions can be incorporated in classifying the fracture complex. In our view radiographs and CT-scans with reconstructions no longer are sufficient. MRI seems to be the most suitable diagnostic instrument to supply the missing data, but MRI images registered shortly after the accident are difficult to interpret, because it takes some time before the amount of water in the tissues changes (oedema). We should find new ways to evaluate the MRI findings in the immediate posttraumatic period to improve decision making in the process of diagnostics and treatment. Although some important steps have been made [25-27], further MRI studies will have to be performed. For example endplate


96

comminution and posterior ligament complex lesions as recognised in MRI predict recurrence of the kyphotic deformity [25]. In the near future new generation MRI can probably completely take over the position of CT-scan in spinal fracture diagnostics. It will diminish roentgen exposure, but it will also have major consequences on logistics, because MRI generally is less available than CT, changes in the MRI set up can be necessary before the diagnostic procedure can be performed, and after early procedures a repeated procedure might be necessary to draw definitive conclusions about the condition of the ligaments and discs. After completing the radiological result studies in this thesis we can confirm certain aspects suggested in the literature, like the loss of the reduced regional angle in the course of follow up. We could clearly show that this is a gradual occurring phenomenon, predominantly caused by collapse of the intervertebral space after implant removal, and not influenced by (major) loss of anterior wedge angle of the reduced vertebral body [19]. We should realize that early studies concerning dorsal instrumentation using the Dick internal fixator revealed excellent results [1;24], just because of short follow up. The longer the follow up, the more loss of reduction of regional angle [12;13;19;29]. We showed that reduction of the vertebral body (AWA) combined with transpedicular cancellous bone grafting is effective and lasting, in contrast with other studies [12;14]. This might be the effect of double-sided cancellous bone graft. In the aforementioned studies the cancellous bone graft was intended to act as an interbody fusion technique, after removing the disc the disc space was filled with cancellous bone as well as the reduced vertebral body. CT-scans revealed that the interbody fusion was insufficient [13]. Unfortunately detailed comparison of outcome in different studies in dorsal instrumentation is not possible, because many variations of the technique, the rehabilitation, the follow up, and the way of evaluation are used in different studies. We never performed direct decompression by laminectomy or corporectomy, but others see this as a standard procedure in case of neurological deficit. To evaluate the effectiveness of the operative treatment with indirect reposition and fixation with an internal fixator, unilateral transpedicular bone grafting and dorsal spondylodesis at the disc levels with endplate lesions compared to conservative treatment (six weeks of rest in supine position followed by mobilisation of the patient with a reclination corset) we designed a randomised trial. In this trial patients with A-type fractures without neurological deficit are included. In this trial not only radiological parameters are compared, but also


97

functional aspects, measured by Roland Morris Disability Questionnaire and VAS Spine Score, as well as cost-effectiveness, are evaluated. The spinal canal is reduced partly as an effect of ligamentotaxis during operation. After the immediate postoperative period a further spontaneous remodelling occurs, not caused by ligamentotaxis. This study confirms earlier results in CT-scans in the literature [6;7;10;11;18;28;32]. We showed that the phenomenon can also be visualized in plain lateral radiographs, although not as accurate as in CT-scans. The remodelling process reveals a gradual clearance of the spinal canal in two years in 97% of the patients. Remodelling was used as a plea against operative treatment [11], and it was one of the arguments against direct or open surgical decompression of the spinal canal as well [3;28]. An explanation for the spontaneous remodelling could not be found. Spinal fluid pressure changes, caused by respiration are suggested to be the drive of this process of remodelling [28]. Bone remodelling, including growth and bone formation, however, are the results of all functional influences, so one can imagine that in normal weight bearing the best form and structure of the spinal column includes a more or less normal spinal canal. More research is needed to clear this subject. Two years after the operative treatment and more than a year after implant removal the ROM of the segment of the spondylodesis equals zero. It is concluded that the dorsal spondylodesis is effective [19]. The other temporarily bridged segment regains only about 50% of the normal ROM. The ROM of the segment above the internal fixator diminishes to 50% as well (compared to normal values). So there is considerable cumulative loss of motion. Regain of ROM after trauma, operation and partial immobilization is a dynamic process. In the future we hope to study ROM as one of the functional aspects of the back in more frequent measurements to clear the dynamic aspects without flexion and extension radiographs (Fig.1). A series of tests concerning multiple aspects of functional outcome revealed that most patients with spinal fractures without neurological deficit have regained their abilities after three to eight years. The series of tests revealed restrictions in body function and structure (AWA, RA, functional capacity, arm, trunk and leg lifting tests), restrictions in activities (RMDQ and VAS Spine Score) and restrictions in participation in daily life (SF36 and return to work status). We could only study a small group of patients. Though we can state that the results are remarkable. Most of the patients have returned to their previous work. In some of the patients some kind of adaptation was necessary to permit them to perform their work. In the functional tests patients have poor scores in the leg lifting test. Therefore muscle training of the lower extremities should be encouraged in the rehabilitation period


98

and the effect of this training has to be evaluated. Lifting tests and questionnaires provide many data about the subjective functional status of the individual patient: we recommend a regular use of at least the RMDQ or VAS Spine Score in the follow up in the clinical setting. The VAS Spine Score seems to reflect this status more precisely than the RMDQ, but further evaluation of both methods in spinal fracture treatment is necessary. Only few studies supply data of RMDQ [17;31] in spinal fracture patients, as it is designed to evaluate low back pain treatment. Even less studies supply normal data in the recently developed VAS Spine Score [16]. Whether we compare our data to low back pain studies or spinal fracture studies, the results are very much in favour of our series. Recent research, ongoing studies and new developments New minimal invasive techniques will be introduced in order to improve the operative treatment of spinal fractures. For example, in selected cases thoracoscopic anterior instrumentation in thoracic spinal fractures (T6-T10) will diminish morbidity compared to conventional lateral thoracotomy. This technique and the indications still need further investigation, before it can be accepted as a standard procedure. Combined anterior and dorsal approach of which the anterior approach can be performed in a thoracoscopic technique, with or without diaphragm splitting to reach for example the second lumbar vertebral body can be applied to spinal fractures of the thoracolumbar junction [2;4;8;9]. Mini lumbotomy for anterior approach of lumbar spinal fracture treatment can be an alternative for the thoracoscopic part of the operation, and can be performed in more distal fractures as well (L3-L5) [5]. Computer navigation will help to position the screws in the pedicles and will help to diminish roentgen exposure to patients and surgeons. Generally spoken, in clinical research concerning spinal fractures and its treatment, it is important that we do not only evaluate the radiological results. All (new) procedures will have to be evaluated more thoroughly than we used to do, including evaluation of morbidity by blood loss and complications, late results, functional outcome, quality of life and cost-effectiveness. Very interesting research is going on and preliminary results have been published in the following fields: • Fluoroscopic and computerized planning and peroperative navigation in

spinal fracture treatment in order to perform more precise positioning of pedicle screws to avoid pedicle fractures, neurological complications and loosening of the pedicle screws [23].


99

• Titanium cages to replace a part of the fractured vertebral body in spinal fracture treatment to facilitate reduction, fixation and interbody fusion [15].

• Balloon vertebroplasty with calcium phosphate cement augmentation for direct restoration of traumatic thoracolumbar vertebral fractures [30].

• MRI and 3D-MRI diagnostic evaluation of spinal fractures, including discs and ligaments [22].

We have started the following studies in the field of spinal fractures: • A prospective randomised trial in two centres of A-type spinal fractures

without neurological involvement conservatively or operatively treated (USS) in the Free University Hospital in Amsterdam and the University Hospital Groningen. Evaluation of radiological results, morbidity, quality of life, impairments, costs, etc (Siebenga, Bakker, Patka, Haarman, Leferink, Zimmerman, ten Duis)

• Functional outcome in non operative spinal fracture treatment in 33 patients, comparable to the study described in chapter 6, including bicycle ergometry, lifting tests, RMDQ, VAS Spinal Score, SF-36 (Keizer, van der Sluis, Leferink, ten Duis)

Fig.1 The SpinalMouse® is run paravertebrally from C7 to the rima ani


100

Table 1 Data as stored and supplied by the software of the SpinalMouse®. Patient 3 years after operative treatment of a spinal fracture level L1, type A3.1. See text for explanation

Static test data Range of motion data

Segment Upright Flexion Extension Flex-up Ext-up Flex-ext

T1/2 0 0 6 0 6 -6 T2/3 2 3 8 0 5 -5 T3/4 9 4 7 -5 -2 -3 T4/5 4 2 2 -2 -2 0 T5/6 2 2 1 0 0 0 T6/7 3 1 1 -2 -2 0 T7/8 2 2 1 0 -2 1 T8/9 5 6 0 1 -5 6 T9/10 7 11 5 4 -2 6 T10/11 6 7 6 1 0 1 T11/12 6 16 5 10 0 10 T12/L1 6 12 3 6 -2 9 L1/2 -1 2 6 3 8 -4 L2/3 -6 2 7 8 13 -5 L3/4 -10 1 0 11 10 1 L4/5 -7 1 -4 8 2 5 L5/S1 -6 2 -6 8 0 8

Sac/Hip 4 80 -43 76 -47 124 Thor. Sp 47 53 42 7 -4 11 Lum. Sp -24 20 6 44 30 14 Inclin -3 115 -30 117 -27 145 Lth[mm] 517 612 444 94 -73 167

1

23

1

2

3

Fig.2 Sagittal shape of the back and inclination angles (C7-S3) in flexed (1), upright (2)

and extended (3) position (same patient as Table 1). Black dot = T7


101

• Validation of the SpinalMouse®, a new device to measure segmental range of motion (interobserver error of a sample of measurements) (Post, Leferink: submitted, Fig.1)

• ROM study, non radiographical measurements of segmental ROM, with SpinalMouse® and its relation to subjective functional outcome as measured in RMDQ and VAS Spine Score in three groups of persons: operatively treated patients with spinal fractures after 3 to 5 years, conservatively treated patients with spinal fractures after 3 to 5 years and healthy volunteers (Post, Leferink: submitted)

SpinalMouse® The SpinalMouse® technically seems to be an appropriate device to incorporate in regular follow-up evaluation of spinal fracture patients from one year after the operative treatment. The measurements are very easily taken by rolling the wheels of the computer mouse like device paravertebrally from C7 to S3 (rima ani) (Fig.1). The length of the back, the local inclination to the plumb line and the sagittal shape of the back is measured in upright, flexed and extended position of the back. These data are send to the computer. The data are stored and immediate calculations supply data about ROM per segment (!) and these are available on the computer screen for comparison with normal values (age and sex specific) and for comparison with previous (interval) data of the same individual. The data are provided as numerical tables (Table 1) and graphic illustrations as well, including the comparison with normal values (Fig.2 and Fig.3). It will supply feed back data to direct and control the rehabilitation process. This feed back will also benefit patients regaining confidence in posture and movement of the back. As a first step in the validation of the SpinalMouse® we analysed interobserver errors of inclination measurements and measurements of the length of the back in different positions. In this respect it is important to analyse the calculated results as well, because some systematic errors will disappear in the calculations. For example systematic measurements of C7-S1 in stead of C7-S3 will not result in (large) interobserver errors in the measurements flexion-upright and extension-upright and flexion-extension. Interpretation of the values of the measurements will be subject of a clinical study. The results can easily be printed and given to the patient as an encouragement to do physical training exercises. It is expected that subjective improvement can be demonstrated and visualized. Eventually the value of the device will have to be determined in larger clinical studies.


102

Upright minus Flexion

Upright minus Extension

Extension minus Flexion

T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12 L1 L2 L3 L4 L5

Fig.3 Graphic demonstration of the calculated measurements of segmental ROM: flexion,

extension and flexion-extension, compared to age and gender specific normal values. See text for explanation

Recommendations This thesis evaluates the radiological outcome and the functional outcome of operatively treated patients with spinal fractures of the thoracolumbar transition. We were able to make and explore a database of all 183 patients operated in 8 years in the surgical department in the University Hospital Groningen. Certain aspects, suggested in literature, could definitively be confirmed, for example progressive loss of regional angle and intervertebral angle. The comprehensive classification should be revised, including more aspects of soft tissue lesions, in order to avoid undertreatment of B-type lesions. Spontaneous remodelling of the spinal canal and the (observed) positive effect of transpedicular bone grafting still needs further investigation. Future research should be directed to diagnostics of soft tissue lesions and its predictive value, functional outcome, including ROM, objective and subjective findings in follow up, and computer navigation for safer procedures.


103

References 1. Aebi M, Etter C, Kehl T, Thalgott J (1987) Stabilization of the lower thoracic and


2. Beisse R, Potulski M, Temme C, Buhren V (1998) Das endoskopisch kontrollierte Zwerchfellsplitting. Ein minimal-invasiver Zugang zur ventralen Versorgung thorakolumbaler Frakturen der Wirbelsäule. Unfallchirurg 101:619-627

3. Boerger TO, Limb D, Dickson RA (2000) Does 'canal clearance' affect neurological

outcome after thoracolumbar burst fractures? J Bone Joint Surg Br 82:629-635

4. Bühren V (1998) Thorakoskopische Behandlung bei Frakturen der Brust- und Lendenwirbelsäule. Langenbecks Arch Chir Suppl Kongressbd 115:108-112

5. Bühren V, Beisse R, Potulski M (1997) Minimal-invasive ventrale Spondylodesen bei Verletzungen der Brust- und Lendenwirbelsäule. Chirurg 68:1076-1084

6. Gertzbein SD, Crowe PJ, Fazl M, Schwartz M, Rowed D (1992) Canal clearance in burst fractures using the AO internal fixator. Spine 17:558-560

7. Ha KI, Han SH, Chung M, Yang BK, Youn GH (1996) A clinical study of the natural remodelling of burst fractures of the lumbar spine. Clin Orthop 323:210-214

8. Huang TJ, Hsu RW, Liu HP, Liao YS, Hsu KY, Shih HN (1998) Analysis of techniques for video-assisted thoracoscopic internal fixation of the spine. Arch Orthop Trauma Surg 117:92-95

9. Huang TJ, Hsu RW, Sum CW, Liu HP (1999) Complications in thoracoscopic spinal surgery: a study of 90 consecutive patients. Surg Endosc 13:346-350

10. Johnsson R, Herrlin K, Hagglund G, Stromqvist B (1991) Spinal canal remodelling after thoracolumbar fractures with intraspinal bone fragments. 17 cases followed 1-4 years. Acta Orthop Scand 62:125-127

11. Klerk de LW, Fontijne WP, Stijnen T, Braakman R, Tanghe HL, van Linge B (1998) Spontaneous remodelling of the spinal canal after conservative management of thoracolumbar burst fractures. Spine 23:1057-1060

12. Knop C, Blauth M, Bastian L, Lange U, Kesting J, Tscherne H (1997) Frakturen der thorakolumbalen Wirbelsäule. Spätergebnisse nach dorsaler Instrumentierung und ihre Konsequenzen. Unfallchirurg 100:630-639


104

13. Knop C, Fabian HF, Bastian L, Blauth M (2001) Late results of thoracolumbar fractures after posterior instrumentation and transpedicular bone grafting. Spine 26:88-99

14. Knop C, Fabian HF, Bastian L, Rosenthal H, Lange U, Zdichavsky M, Blauth M (2002) Fate of the transpedicular intervertebral bone graft after posterior stabilisation of thoracolumbar fractures. Eur Spine J online first: http://dx.doi.org/s00586-001-0360

15. Knop C, Lange U, Bastian L, Blauth M (2001) Biomechanische Stabilität mit einem neuen Wirbelkörpersersatzimplantat: Drei-dimensionale Bewegungsanalyse an instrumentierten humanen Wirbelsäulenpräparaten. Unfallchirurg 104:984-997

16. Knop C, Oeser M, Bastian L, Lange U, Zdichavsky M, Blauth M (2001) Entwicklung und Validierung des VAS-Wirbelsäulenscores. Unfallchirurg 104:488-497

17. Kraemer WJ, Schemitsch EH, Lever J, McBroom RJ, McKee MD, Waddell JP (1996) Functional outcome of thoracolumbar burst fractures without neurological deficit. J Orthop Trauma 10:541-544

18. Kuner EH, Schlickewei W, Kuner A, Hauser U (1997) Restoration of the spinal canal by the internal fixator and remodelling. Eur Spine J 6:417-422


20. Leferink VJM, Zimmerman KW, Veldhuis EFM, ten Vergert EM, ten Duis HJ (2002) Classificational problems in ligamentary distraction type vertebral fractures: 30% of all B-type fractures are initially unrecognised. Eur Spine J 11:246-250


22. Martel AL, Heid O, Slomczykowski M, Kerslake R, Nolte LP (1998) Assessment of 3-dimensional magnetic resonance imaging fast low angle shot images for computer assisted spinal surgery. Comput Aided Surg 3:40-44

23. Nolte LP, Slomczykowski MA, Berlemann U, Strauss MJ, Hofstetter R, Schlenzka D,

Laine T, Lund T (2000) A new approach to computer-aided spine surgery: fluoroscopy-based surgical navigation. Eur Spine J 9 Suppl 1 S78-88

24. Olerud S, Karlstrom G, Sjostrom L (1988) Transpedicular fixation of thoracolumbar vertebral fractures. Clin Orthop 227:44-51


105

25. Öner FC, Gils-van APG, Faber JAJ, Dhert WJA, Verbout AJ (2002) Some complications of common treatment schemes of thoracolumbar spine fractures can be predicted with magnetic resonance imaging. Spine 27:629-636

26. Öner FC, van der Rijt RR, Ramos LM, Dhert WJ, Verbout AJ (1998) Changes in the disc space after fractures of the thoracolumbar spine. J Bone Joint Surg Br 80:833-839

27. Öner FC, van Gils AP, Dhert WJ, Verbout AJ (1999) MRI findings of thoracolumbar spine fractures: a categorisation based on MRI examinations of 100 fractures. Skeletal Radiol. 28:433-443

28. Scapinelli R, Candiotto S (1995) Spontaneous remodelling of the spinal canal after burst fractures of the low thoracic and lumbar region. J Spinal Disord 8:486-493

29. Speth MJ, Öner FC, Kadic MA, de Klerk LW, Verbout AJ (1995) Recurrent kyphosis after posterior stabilization of thoracolumbar fractures. 24 cases treated with a Dick internal fixator followed for 1.5-4 years. Acta Orthop Scand 66:406-410

30. Verlaan JJ, van Helden WH, Öner FC, Verbout AJ, Dhert WJ (2002) Balloon vertebroplasty with calcium phosphate cement augmentation for direct restoration of traumatic thoracolumbar vertebral fractures. Spine 27:543-548

31. Weinstein JN, Collalto P, Lehmann TR (1988) Thoracolumbar "burst" fractures treated conservatively: a long-term follow-up. Spine 13:33-38

32. Wessberg P, Wang Y, Irstam L, Nordwall A (2001) The effect of surgery and remodelling on spinal canal measurements after thoracolumbar burst fractures. Eur Spine J 10:55-63


106

THORACOLUMBAR FRACTURES 8 ABSTRACTS

107

8 Abstracts Abstract chapter 1 Some aspects of the ancient history of the treatment of spinal fractures are described. Conservative and operative treatments are discussed, for example the operative techniques of bitibial traction and fixation with Harrington rods. This treatment was performed in our hospital during many years. The epidemiology of trauma patients is described in general and the epidemiology of spinal trauma patients, including the studied group of patients of this study, is described more in detail. ICD-9, ISS and RLOG features are discussed. The incidence of spinal fractures showed only a marginal increase in the last three decades. Operative treatment, reduction and fixation, with the internal fixator, combined with transpedicular cancellous bone grafting and dorsal spondylodesis is the actually performed treatment in our study group, with only minor variants in the years 1988 to 1996. After treatment schemes, as well as follow-up schemes has been unchanged during the years, which gave us the opportunity to study 183 patients with thoracolumbar fractures who were operatively treated in this period. Abstract chapter 2 The clinical records, operation records, X-rays and CT-scans of 160 operatively treated patients with A-type and B-type spinal fractures were evaluated in a retrospective study. The preoperative diagnosis was compared with the postoperative diagnosis. Analysis of characteristics of patients with A-type fractures (without the unrecognised B-type fractures), initially unrecognised B-type (uB) fractures, and B-type fractures (without the unrecognised B-type fractures) was performed. We analysed the age of the patients, the respective


108

fracture levels, neurological deficit, anterior wedge angles (AWA), anterior corporal height (ACH), posterior corporal height (PCH), and the percentage of frontal corporal collapse (FCC). The t-test was used for statistical analysis. The mean age of patients in each group did not show a significant difference. The group of unrecognised B-fractures had a more caudal fracture level than the recognised B-type fractures. The fracture levels of the A-group and the uB-group patients showed no difference using the t-test. The percentage of patients with spinal fractures with neurological deficit is 16% in the A-type fracture group, 12% in the uB-fracture group and 50% in the B-type group. The preoperative classification of patients in the A-group and in the uB-group showed that patients in the uB-group have more than proportional relatively simple preoperative A-fractures. The mean PCH of the uB-group was higher than the PCH of the A-group. Thirty percent of B-type fractures are misdiagnosed when plain X-rays and CT scans with 2D reconstructions are used as the only preoperative diagnostic tools. A large PCH with a normal interspinous distance should raise the suspicion of a B-type lesion. A large AWA does not point to a ligamentary B-type fracture. Abstract chapter 3 In internal posterior fixation of thoracolumbar fractures combined with transpedicular cancellous bone graft and posterior fusion of the intervertebral facet joints at the level of the destroyed end plate it is still uncertain as to whether significant vertebral body collapse and loss of correction of the regional angle (RA) and the intervertebral angle (IVA) occur (after removal of the implants). These questions were investigated in a retrospective study of 183 consecutive patients, 18-65 years old, with a spinal fracture between the 9th thoracic and the 5th lumbar vertebral body (inclusive), treated operatively between 1988 and 1996 (27% had objective neurological deficit, 37% had multiple injuries). According to the Comprehensive Classification, 128 type A, 32 type B and 21 type C-fractures were identified preoperatively. Changes in AWA, IVA and RA were measured preoperatively, and within 1 month, 9 months and 24 months postoperatively. The effect of implant failure was also evaluated. The normality of the distribution was tested using the Kolmogorov-Smirnov (K-S) test. The one-sample runs test and the t-test were used to evaluate angle changes. The reduced vertebral body did not collapse after 9 months, when most of the patients (170) underwent removal of the implants, but significant changes in IVA were found after implant removal. Correction of the RA was statistically significant before implant removal, but the RA 2 years after surgery equalled the preoperative value. Changes at the level of the intervertebral space contributed to the loss in the RA. Broken pedicle screws


109

(10.9% of the patients) resulted in significant changes in the AWA and RA before implant removal. Abstract chapter 4 Although multiple studies concluded operative decompression of a traumatic narrowed spinal canal is not indicated because of spontaneous remodelling, instrumental decompression is frequently used as part of the operative treatment of spinal fractures. To investigate the process of remodelling we studied the diameter of the spinal canal in 95 patients with burst fractures of the thoracolumbar junction (T9-L2). To measure and compare the spinal canal’s diameter we used either CT-scans or radiographs, made preoperatively, postoperatively, after 9 months and after 24 months. In lateral plain radiographs we found that the initial percentage of cases with bony canal narrowing preoperatively of 76.5 is reduced to 18.4% postoperatively, to 8.2% at 9 months, and to 2.4% at 24 months. In CT-scans in a selection of patients the mean residual diameter of the spinal canal was 53% preoperatively and 78% at 24 months. The posterior segmental height increases during operation and decreases in the respective periods after operation. We conclude a significant spontaneous remodelling of the spinal canal follows the initial surgical reduction. Two years after operation bony narrowing of the spinal canal is only recognisable in 2.4% of the patients in plain lateral radiographs. The remodelling of the spinal canal can be seen in plain radiographs, although not as accurate as in CT-scans. Abstract chapter 5 In order to study the effect of dorsal spondylodesis on intervertebral movement in patients treated for thoracolumbar fractures, we measured the sagittal range of motion (ROM) in the segments above and below the fractured vertebral body two years after operation. Between 1991 and 1996, 82 consecutive patients with a fracture of the thoracolumbar spine (T12, L1, L2 and L3) were treated operatively. Eighteen T12, 42 L1, 17 L2 and 5 L3 fractures were included. The range of motion of two segments above and two segments below fracture level was measured on plain flexion and extension radiographs. The data were compared to normal values and to the zero distribution with the K-S test. At all fracture levels the ROM of the segment adjacent to the disturbed endplate of the fractured body was zero. All other evaluated segments showed significant loss of ROM (p<0.05) compared to normal values, except segment L1-L2 in


110

L3 fractures. Dorsal spondylodesis at the level of the disturbed endplate in thoracolumbar spinal fractures is effective. More than 50% loss of motion in the two adjacent levels is equivalent to complete loss of ROM in a second segment. Abstract chapter 6 The aim of this study is to develop insight in the impairments in operatively treated spinal fracture patients, and also in their ability to participate in daily living, return to work and quality of life as defined by the WHO. Nineteen patients operated for a type A fracture of the thoracolumbar spine (T9-L4) between 1993 and 1998 in the University Hospital Groningen, aged between 18 and 60 years, without neurological deficit, were included in the study. Operative treatment consisted of fracture reduction and internal fixation using the Universal Spine System, combined with transpedicular cancellous bone grafting and dorsal spondylodesis. Restrictions in body function and structure are measured in radiographs and in functional capacity tests, like lifting tests and ergometry. Restrictions in activities are studied with the Visual Analogue Scale Spine Score and the Roland Morris Disability Questionnaire. Restrictions in participation/quality of life are analysed with the Short Form 36 and described in the return to work status. The radiological results are comparable to the literature. The reduction of the anterior wedge angle is followed by a gradual partial loss of intervertebral angle and regional angle. The maximum oxygen uptake (VO2-max) is reduced in only 8.3% of the patients. Arm and trunk lift is within the normal range in 87% and 80% of the patients respectively, but only 53% of the patients is able to perform a leg lift within the normal range. A mean RMDQ score of 4.0 positive items (SD 6.0) was found and the mean VAS Spinal Score was 79.4 (SD 25.0), both better than in other series. No significant differences to the values of a comparable (healthy) age group could be identified in any variable of the SF36. A high correlation was seen between RMDQ, VAS Spine Score and the SF36 categories. No correlation was found between AWA and RA, and functional capacity tests or questionnaire scores. 87% of the patients with paid labour before the trauma had returned to work at follow-up. About 50% of the patients had to change the intensity of their labour or the kind of work after the injury and treatment. In this matter leg (muscle) performance seems a more important factor than overall condition (VO2-max).


111

Abstract chapter 7 Important issues of this thesis, with consequences for future management of spinal fractures are:

• Ligamentary lesions in the complex spinal injury are not appreciated enough by Plain X-rays and CT-scans only, although these lesions should have major influence on the choice of treatment. Probably additional MRI investigation, at the right time, will elp to overcome this problem.

• Radiological results of the studied treatment modality confirm earlier data in the literature. The reduced AWA stays about the same, but because of complete loss of IVA the RA finally equals the preoperative value.

• Spontaneous remodelling of the narrowed spinal canal can be shown not only in CT scans, but in plain transverse radiographs as well.

• Dorsal spondylodesis in spinal fracture treatment is effective, as measured in flexion-extension radiographs after 2 years: ROM is zero.

• Additional loss of ROM in the adjacent segments accounts for 50% at each segment. In the near future we expect that we can measure intersegmental ROM with the SpinalMouse®. We have tested this apparatus and publication of the data is prepared.

• Functional tests and back specific questionnaires (RMDQ, VAS Spine Score) supply important information about functional outcome in spinal fracture patients. General condition is very well, but leg muscle performance is bad. This is important for further development of rehabilitation programs.


112

THORACOLUMBAR FRACTURES 9 NEDERLANDSE SAMENVATTINGEN

113

9

Thoracolumbale wervelfracturen Aspecten van de epidemiologie, classificatie, radiologische en functionele resultaten Nederlandse samenvattingen Hoofdstuk 1 Een niet behandelde wervelfractuur betekende vroeger nog al eens de dood, zeker wanneer er ook een dwarslaesie aanwezig was. Gelukkig is de prognose tegenwoordig niet meer zo somber. De behandelingsmethoden voor wervelfracturen zijn de afgelopen decennia sterk veranderd. Het rekken en trekken van weleer heeft plaatsgemaakt voor moderne operatieve technieken. Na een inleiding over de behandelingsmethoden voor wervelfracturen in de oudheid staan we stil bij verschillende aspecten van de conservatieve behandelingen zoals die ook nu nog voor sommige typen wervelbreuken in zwang zijn. Ook de diverse operatieve therapieën komen kort aan bod, waarbij in extenso wordt ingegaan op de toepassing van Harrington staven, een methode die gedurende vele jaren is toegepast in onze kliniek. Voor een beschrijving van de epidemiologische aspecten van de ongevalspatiënten die de basis vormen voor dit onderzoek wordt gebruik gemaakt van het systeem: Registratie Letsels en Ongevallen Groningen (RLOG). In dit systeem zijn alle ongevalspatiënten opgenomen die vanaf 1969 door het aandachtsgebied traumatologie van de chirurgische kliniek van het AZG zijn behandeld. De patiënten met wervelfracturen worden in detail besproken. Hierbij blijkt bijvoorbeeld dat de incidentie van wervelfracturen de laatste 30 jaar nauwelijks is toegenomen.


114

We besluiten dit hoofdstuk met een bespreking van de op dit moment in de Groninger kliniek toegepaste operatieve behandelingsmethodiek: de methode volgens Dick. Tijdens deze ingreep wordt, via een dorsale benadering, een onbloedige repositie van het wervellichaam verricht in combinatie met een inwendige fixatie, een botplastiek via de boogvoetjes en het vastzetten van de facetgewrichten. We gaan in op de voordelen van het door ons gehanteerde sterk protocollaire behandelingsschema voor zowel operatie, nabehandeling in het revalidatiecentrum als poliklinische controle. Het patiëntencollectief dat de basis vormt voor de deelonderzoeken beschreven in de volgende hoofdstukken wordt gevormd door die 183 patiënten die in de periode 1988-1997 vanwege een instabiele fractuur in de thoracolumbale wervelkolom operatief werden behandeld (binnen het aandachtsgebied traumatologie). Hoofdstuk 2 In dit hoofdstuk bespreken we de diverse classificatiesystemen voor wervelfracturen. In onze kliniek wordt de AO- classificatie gebruikt (Arbeitsgesellschaft für Osteosynthesefragen). Toepassing van deze indeling heeft in ons patiëntenmateriaal tot gevolg gehad dat 30% van alle B-letsels (distractieletsels) aanvankelijk zijn gemist wanneer we gebruik maakten van conventionele röntgenfoto’s en CT-scan. Pas peroperatief kon de diagnose B-letsel in deze groep worden gesteld op grond van dorsaal ligamentaire verscheuring. De niet-herkende B-letsels gingen vaker gepaard met een relatief eenvoudig corpusletsel en minder vaak met neurologische uitval dan wel preoperatief herkende B-letsels. Een grote wighoek (AWA) bleek geen voorspeller van dorsaal ligamentair letsel. Echter bij een intacte achterzijde van het corpus moeten we wel bedacht zijn op ligamentaire betrokkenheid. We realiseren ons daarbij dat een intacte achterzijde van het corpus kan fungeren als een draaipunt (fulcrum) bij compressie aan de voorzijde. Het logische gevolg is een distractie moment met letsel aan de achterzijde welke niet altijd op de röntgenfoto’s herkenbaar is. Hoofdstuk 3 In dit hoofdstuk bespreken we de radiologische resultaten van de operatieve behandeling van alle door ons geopereerde wervelfracturen (n=183). Naast het fractuurniveau, de fractuurclassificatie en de aan/afwezigheid van neurologisch letsel, richten we ons op de radiologische veranderingen van het gefractureerde


115

wervellichaam en de aangrenzende tussenwervelruimten zoals deze zich gedurende het gehele behandelingstraject voordoen. We onderscheiden in dit traject de direct posttraumatische situatie, het direct postoperatieve resultaat, de situatie na 9 maanden wanneer het fixatiemateriaal wordt verwijderd en het (eind) resultaat na 2 jaar. Als radiologische parameters hanteren we in het sagittale vlak de wigvormige deformiteit van het wervellichaam (AWA) in graden, de hoek die gevormd wordt door de lijnen die de intervertebrale ruimte van de bovenliggende wervel begrenzen (IVA) en de som van beide genoemde parameters, de regionale hoek (RA). De AWA van 18 graden posttraumatisch wordt gecorrigeerd tot 5.9 graden bij de operatie, een correctie die in de loop van de tijd voor een gering deel weer verloren gaat (tot 6.8 en 7.3 graden respectievelijk). De IVA verandert van 4,5 via 4,9 en 3,7 naar –0,4 graden, een verlies van 4,9 graden. De RA bedraagt posttraumatisch 9,9 om vervolgens via –0.3 en 2.3 terug te gaan naar 9.2 hetgeen betekent een nagenoeg volledige teruggang naar de situatie als vlak na het ongeval. Uit de literatuur blijkt dat die onderzoeken die een korte controle periode hanteren (korter dan 6 maanden) een behoud van de regionale hoek laten zien, maar dat bij een langere controle periode (jaren) het verlies van de RA ook circa 10 graden bedraagt. Breuk van het fixatiemateriaal beï nvloedt de corpusvorm (AWA) wel, maar de IVA niet. Hoofdstuk 4 De vormveranderingen van het wervelkanaal (in het sagittale vlak) als gevolg van het ongeval, de operatieve behandeling en zoals te zien tijdens de herstelperiode bij de 95 patiënten met een A3 wervelfractuur komen in dit hoofdstuk aan de orde. Als parameters hanteren we hier de hoogte van de achterzijde van het gefractureerde wervellichaam (PVH) en de hoogte van de achterste begrenzing van de intervertebrale ruimte (PIV). Tezamen vormen zij de achterste segmentale hoogte (PSH). Het aanspannen van het hier aanwezige ligament (ligamentum longitudinale posterius) en de invloed die dit heeft op de verplaatste botfragmenten wordt ligamentotaxis genoemd. De als gevolg van het ongeval optredende verplaatsing van de fractuurdelen naar dorsaal geeft een vernauwing van de diameter van het wervelkanaal. De PSH neemt als gevolg van de operatie toe van 40.5 tot 43.2 mm om daarna af te nemen tot 38.7 mm na 2 jaar.


116

Bij 75% van de patiënten verdwijnen als gevolg van de operatie de botfragmenten uit het wervelkanaal. Na 24 maanden is bij 96% van de patiënten die initieel een vernauwing van het wervelkanaal hadden, deze niet meer waarneembaar. Een complicerende factor bij de beoordeling van de wijdte van het spinale kanaal is dat niet alle botfragmenten goed kunnen worden geï dentificeerd op de zijdelingse foto. Een CT-scan is daartoe een meer geëigend diagnosticum. Duidelijk is evenwel dat remodellering van het spinale kanaal (teruggang van de vernauwing) doorgaat gedurende de twee jaar durende herstelperiode. Ligamentotaxis draagt hiertoe waarschijnlijk niet aan bij aangezien de PSH in deze periode afneemt. Hoofdstuk 5 Algemeen wordt aangenomen dat de bewegingsmogelijkheid van het rugdeel dat de gefractureerde wervel bevat, als gevolg van het letsel en de uitgevoerde operatieve behandeling, vermindert. De range of motion (ROM) van de segmenten rondom de gebroken wervel kan worden gemeten door het maken van zogenaamde flexie-extensie röntgenopnamen. In dit hoofdstuk wordt beschreven hoe groot na 2 jaar de range of motion is bij 82 patiënten die vanwege een wervelfractuur een operatieve behandeling hebben ondergaan. Voor een goed begrip van de uitkomsten is het van belang te weten dat er bij de operatieve behandeling van het wervelletsel naar wordt gestreefd, om naast het herstel van de normale anatomische verhoudingen, het intervertebrale gewricht aan de craniale zijde van de gebroken wervel definitief te verstijven door het uitvoeren van een zogenaamde achterste spondylodese. Repositie van de fractuur en genoemde spondylodese worden tijdelijk (9 maanden) onderhouden door een fractuur overbruggende fixatie van de craniaal en caudaal (rostraal) gelegen wervel. Na verwijderen van het fixatiemateriaal gaan we er van uit dat het craniaal van de fractuur gelegen intervertebrale gewricht onbeweeglijk is geworden en dat het caudaal gelegen gewricht een deel van zijn oorspronkelijke bewegingsmogelijkheid terugkrijgt. De eerste veronderstelling wordt in dit onderzoek bevestigd. Het tijdelijk gefixeerde distale intervertebrale gewricht herwint 50% van zijn oorspronkelijke mobiliteit, een percentage dat ook geldt voor het (onbeschadigde en niet tijdelijk gefixeerde) intervertebrale gewricht craniaal van de tijdelijk gefixeerde segmenten. De conclusie is dat de spondylodese inderdaad effectief is maar dat het verlies van beweeglijkheid in de aangrenzende gewrichten groter is dan verwacht. Hoofdstuk 6


117

Met het in dit hoofdstuk beschreven onderzoek wordt inzicht verkregen in de beperkingen die patiënten ervaren na de operatieve behandeling van een wervelfractuur, hun conditie en hun mogelijkheden in het dagelijkse leven, o.a ten aanzien van betaald werk en de kwaliteit van leven. Hiertoe werden 19 patiënten onderzocht, allen met een wervelfractuur tussen Th 9 en L4, zonder neurologische uitval. Onderzoek vond plaats 3 tot 8 jaar na ongeval en operatieve behandeling. Met functionele capaciteitstesten, zoals tiltesten en fietsergometrie werden de beperkingen in lichaamsfunctie vastgelegd. Vermindering in dagelijkse activiteiten werden vastgelegd met behulp van vragenlijsten als de Visuele Analoge Wervel Scorelijst (VAS) en de Roland Morris Vragenlijst (RMDQ). Beperkingen in het dagelijkse leven alsmede de kwaliteit van leven werden beoordeeld aan de hand van de SF36 (Rand 36) vragenlijst en door het vastleggen van de arbeidsstatus. Op genoemde parameters scoren wervelpatiënten in het algemeen even hoog als vergelijkbare personen zonder wervelletsel. Zo is de maximale zuurstofconsumptie slechts bij 8,3% van de patiënten onder de norm. Arm- en romptiltesten laten een vergelijkbaar beeld zien. Opvallend is dat bij de tiltesten die de beenfunctie evalueren slechts 53% van de patiënten normale waarden bereiken. De RMDQ bedroeg gemiddeld 4,0, de VAS 79,4; beide scores zijn beter dan in andere series wervelfractuur-patiënten. De SF36 gaf vergelijkbare waarden als bij gezonde proefpersonen. Van de patiënten was 87% weer aan het werk, maar de helft ervan met aanpassingen in duur of zwaarte. De functie en de kracht van de benen lijken, gezien de uitkomsten, het functioneren meer te beperken dan de algemene conditie. Revalidatie zou daar meer op gericht kunnen worden. Hoofdstuk 7 Als belangrijkste bevindingen in dit proefschrift kunnen worden genoemd: • Op het gebied van de fractuurclassificatie is duidelijk geworden dat de

ligamentaire letsels als onderdeel van het letselcomplex onderbelicht worden met het alleen maken van gewone röntgenfoto’s en CT-scans. Deze ligamentaire letsels beï nvloeden de keuze van de in te stellen behandeling echter wel. Indien er onduidelijkheid bestaat of er naast de wervelfractuur ook sprake is van ligamentair letsel ontstaat de kans op onderbehandeling. Met het uitvoeren van MRI-onderzoek, eventueel in plaats van CT-onderzoek, is het waarschijnlijk wel mogelijk deze letsels te betrekken bij de indicatiestelling tot de te volgen operatieve behandelstrategie. Hoe dit wat betreft timing en logistiek het beste kan, zal in de toekomst nader onderzocht moeten worden.


118

• De radiologische resultaten bevestigen voor een deel op eenduidige wijze de gegevens uit de literatuur. In de loop van de behandeling gaat de winst verkregen bij de repositie ten aanzien van de regionale hoek van het geopereerde en gefixeerde wervelsegment weer verloren. Deels nieuw is dat dit verlies niet optreedt in het herstelde wervellichaam maar in de intervertebrale ruimte. Bij het ontwikkelen van nieuwe operatietechnieken zal hiermee rekening moeten worden gehouden.

• Het spontane herstel van de vorm van het wervelkanaal is niet alleen op CT maar ook op gewone röntgenfoto’s te herkennen.

• Ten aanzien van de effectiviteit van het uitvoeren van een achterste spondylodese hebben wij met behulp van flexie-extensie foto’s aangetoond dat het segment van de spondylodese na twee jaar niet meer beweegt.

• De aan de spondylodese grenzende wervelsegmenten hebben na twee jaar een bewegingstraject van 50% van normaal. In de toekomst verwachten wij deze interspinale bewegingstrajecten op niet radiologische wijze standaard te gaan vastleggen. Het hiertoe ontwikkelde apparaat (SpinalMouse®-Idiag) wordt momenteel door ons op zijn bruikbaarheid getest.

• Functionele onderzoeken en vragenlijsten zoals de Roland Morris Disability Questionnaire en de VAS Spine Score geven belangrijke informatie over het functioneren van de wervelfractuur patiënt. Hoewel de algemene conditie na 3 tot 8 jaar goed is blijkt de kracht van de benen (beentiltest) ver onder de norm te zitten. Deze bevinding kan een belangrijk gegeven zijn voor aanpassing van de revalidatie.

THORACOLUMBAR FRACTURES 10 DANKWOORD

119

10 Dankwoord Prof.dr. H.J. ten Duis, promotor en medisch hoofd van het aandachtsgebied traumatologie, beste Henk Jan, als (pro-)motor van dit onderzoek heb je mij begeleid vanaf de opzet in 1997 tot en met de allerlaatste punten en komma’s. Jouw werklust was voor mij vaak een stimulans om een tandje hoger te gaan. Jij was ook de stok achter de deur als ik eens geen zin meer had. Mijn hartelijke dank voor de voortreffelijk toegesneden hulp die ik van je mocht krijgen. Het is heel sterk jouw verdienste dat we klaar zijn. Drs. K.W. Zimmerman, beste Kees, jouw onderzoek is eindelijk afgerond. Behalve voor alle praktische hulp zoals bij het scannen en bewerken van afbeeldingen en voor je enthousiasme bij het verbeteren van de manuscripten ben ik je vooral erkentelijk voor de degelijke opzet van het onderzoek zoals je dat met Eduard Veldhuis in 1987 gestart hebt. Bovendien heb je me wervelfracturen leren opereren, een niet alledaagse operatie voor traumatologen. Dr. E.F.M. Veldhuis, referent, beste Eduard, ook jij hebt aan de wieg gestaan van dit langlopende onderzoek. Na jouw promotie-onderzoek naar de resultaten van de behandeling van wervelfracturen met de Harringtonstaven leek dit een logisch vervolg. Nu, bijna 10 jaar later, staan we hier dus weer! En ook van jou heb ik wervels leren opereren, inclusief middernachtelijk onthaken van rijdende luxatiefracturen. Waarvoor mijn dank. Prof.dr. R. van Schilfgaarde, voorzitter van de beoordelingscommissie van dit proefschrift, mijn opleider in de heelkunde en tot recent hoofd van de afdeling. Beste Reinout, voor alle genoemde aspecten ben ik je veel dank verschuldigd. Het is voor mij een eer dat je zitting wilde nemen in de beoordelingscommissie.


120

Prof.dr. H.J.Th.M. Haarman, lid beoordelingscommissie, hoofd van die andere Nederlandse chirurgische afdeling waar traumatologen wervelfracturen behandelen, namelijk in het VU-ziekenhuis. Bedankt voor uw inzet en voor het constructieve denkwerk in het kader van het langlopende gezamenlijke wervelonderzoek. Prof.dr. A.J. Verbout, lid beoordelingscommissie, hoofd orthopedische afdeling van het AZU. Met uw stimulerende woorden tijdens het NOV congres te Groningen en de uitnodiging tot gezamenlijk verder onderzoek op het gebied van wervelfracturen heeft u de deur geopend voor verdere integratie van traumatologie en orthopedie. Ik dank u hartelijk voor uw inspanningen. Dr. K.W. Wendt, paranimf en collega traumatoloog, beste Klaus, mijn hartelijke dank voor je persoonlijke inzet om ons vak op een goede manier te kunnen uitoefenen. Door onze al negen jaren lange intensieve en plezierige samenwerking was het voor mij al tijden duidelijk dat jij mijn paranimf moest zijn. Drs. H.S. Hofker, paranimf en collega chirurg, beste Sijbrand, ook met jou gaat de uiterst plezierige samenwerking en vriendschap terug tot 1993, toen wij tegelijkertijd in de Groningse kliniek aankwamen, jij als eerstejaars en ik als (zogenaamd ervaren) vierdejaars. Mijn dank voor alle inzet. Laten we zeggen Dr. H.S. Hofker in 2004…. Mr. C.H.M. Houben-Mom, beste Inge, met één goedgeplaatste opmerking heb jij mij in april 2001 met succes weer aan het werk gezet, toen we al extrapolerend uitkwamen op 2010 als jaar van promoveren. Het resultaat van dat werk ligt er nu, anderhalf jaar later. Dr. E.M. ten Vergert, mede-auteur, hoofd van het MTA-bureau, beste Els, dankjewel voor alle hulp bij de statistiek. Kolmogorov en Smirnov, ik wilde eerst niet geloven dat ze bestonden. Jij was steeds verbaasd over hetgeen wij met patiënten doen, maar na jouw uitleg en instructies verbaasde het mij altijd weer dat statistiek niet moeilijk hoeft te zijn. Drs. J.M.M. Nijboer, arts, mede-auteur, beste Annemarie, jaren geleden begon je als student mij met strafwerk te bestoken om op wintersport te mogen: van het één komt het ander: je hebt me erg op gang geholpen en gehouden bij dit onderzoek en en passant hebben we samen 5 artikelen geschreven.


121

Dr. C.K. van der Sluis, revalidatie-arts en mede-auteur, beste Corrie, door jouw ervaring in onderzoek naar outcome bij traumapatienten en door jouw intensieve begeleiding van Evelien hebben wij een bijzonder outcome onderzoek kunnen doen. Het was prettig met je samen te werken. Drs. H.E. Keizer, revalidatie-arts en mede-auteur, beste Evelien, door jouw inzet werd een zeer uitgebreid testprogramma bij wervelpatiënten tot een goed einde gebracht. Door positieve en negatieve persoonlijke omstandigheden is jouw artikel nog even blijven liggen. Ik hoop dat we dat binnenkort nog kunnen schrijven en wil je daar graag behulpzaam bij zijn. Drs. J.K. Oosterhuis, arts, mede-auteur, beste Jurjen, ook jij kwam als student binnen en ging nooit meer weg, mijn dank voor alle hulp, de hand- en spandiensten en de gezellige uren. Dr. J.H. Kingma, hoofd datamanagementbureau AZG, mijn dank voor het in overvloed aanleveren van getallen die nodig waren voor het schrijven van hoofdstuk 1. Drs. R.B. Post, arts, beste Richard. Jouw snelheid van werken heeft tot gevolg dat kort na dit boekje nog twee artikelen over de wervelchirurgie gepubliceerd gaan worden. Dat is een grote stimulans om verder te gaan. Drs. D. Mulder, revalidatie-arts, lid werkgroep wervelfracturen, beste Dieneke, mijn dank voor de vele bijdragen aan de verbetering van de behandeling van onze gezamenlijke patiënten. Dr. H. Haaxma, neuroloog, lid werkgroep wervelfracturen, beste Hannie, ook jou komt dank toe voor de inzet tijdens de gezamenlijke spreekuren van de werkgroep wervelfracturen. Ik ben dankbaar voor de secretariële en administratieve ondersteuning van Riëtte, Marianne, Bernadette, Maaike, Kitty, Jadwiga en Willemien voor typewerk, verzorging briefwisseling en overdrukken, maar met name voor de stimulerende woorden en de gelegenheid tot stoom afblazen. Hoewel een beetje auteur zelf op de hoogte zou moeten zijn van alle mogelijkheden van Word, Excel en Powerpoint was raad en daad van jullie van tijd tot tijd onontbeerlijk. Fijkje, Wilma, Ali, Myrtha en meneer Bosker, voor de laatste keer: zoekgeraakte MD’s en X-foto’s liggen nooit bij mij, behalve soms. Mijn hartelijke dank voor


122

alle hulp bij het steeds weer verzamelen van de benodigde pakketjes dossiers en foto's. Drs. P. Houben, chirurg te Zwolle en prof.dr. P. Robinson, plastisch chirurg, beste Paul, beste Peter, in de fase van de afronding van de teksten heb ik veel hulp ervaren van jullie belangrijke verbeteringen. Drs. M.F. Grapenthin, arts, beste Maren, vele uren van je beperkte en kostbare tijd als co-assistent in Groningen heb je aan mijn boekje besteed. De vriendelijke zelfverzekerdheid waarmee je mij vastberaden en telkens weer duidelijk maakte dat de tekst en de figuren die niet goed waren verder verbeterd moesten worden, overtuigde mij. Duizenden verbeteringen, tot het (bijna) perfect was. Overige collegae, chirurgen en assistenten, mijn excuses voor het voortdurende geroep dat het bijna klaar was. Herman en Willy, lieve papa en mama, ik dank jullie hartelijk voor de mogelijkheden die jullie mij geboden hebben, door mij te laten studeren en vrij te laten bij de vele soms lastige keuzes in een tijd van numerus fixus en beperkte middelen. De kansen die er waren zijn daardoor op zinvolle wijze benut. Dit boek moge daar ook een tastbaar resultaat van zijn. Tineke, dank zij jou heb ik de tijd aan dit onderzoek kunnen besteden die nodig was; de begeleiding en opvoeding van Anouk en Nicole kwam daardoor te vaak op jou neer. Door het samenlopen van te lange werkdagen, een te hoge dienstfrequentie met teveel spoedoperaties en de tijdsbesteding aan onderzoek en artikelen was ik vaak afwezig gedurende avonden en weekends. Toch heb jij gezorgd dat alles thuis op rolletjes liep en dat het voor mij ook nog prettig thuiskomen was. Zonder jouw steun was dit boek er niet gekomen. Anouk en Nicole, hè, hè, de artikelen zijn klaar, het boekje is klaar, alles is geregeld. De (promo)veer kan gestoken worden……

THORACOLUMBAR FRACTURES 11 CURRICULUM VITAE

123

11 Curriculum vitae The author was born in 1959 in Herwen en Aerdt, nowadays called Rijnwaarden, where the river Rhine enters the Netherlands. He attended the Gymnasium in Arnhem and in Zwolle until 1977. For two years he studied econometry, before he started his medical study in 1979 at the University of Groningen. He passed his surgical internships in Deventer (Head: Dr. P. van Elk) and graduated in 1987. Afterwards he worked as a resident in Arnhem (Plastic Surgery: Dr. J.-P.A. Nicolai) and in Enschede (Surgery: Dr. Ph. Engel, Dr. J.W.J.L. Stapert, Dr. I.J. Hoogendam) including the first three years of his registered surgical training. After completing his surgical training in the University Hospital Groningen (Head: Prof.dr. R. van Schilfgaarde) in 1996 he worked as a traumatology fellow in the subdivision of traumatology (Head: Prof.dr. H.J. ten Duis). As a fellow of the Arbeitsgruppe für Osteosynthesefragen he worked in the Medizinische Hochschule Hannover (Head: Prof. Dr. H. Tscherne) in 1997. Afterwards he joined the senior staff traumatology in Groningen, with an additional task in vascular surgery. During this period he started to work on this thesis. The author has been married to Tineke Halverhout since 1989 and they have two daughters, Anouk and Nicole. Address for correspondence V.J.M. Leferink Department of Surgery University Hospital Groningen Postbox 30.001 9700 RB GRONINGEN The Netherlands E-mail to: [email protected]

THORACOLUMBAR FRACTURES 11 CURRICULUM VITAE

124

THORACOLUMBAR FRACTURES 12 LIST OF ABBREVIATIONS

125

12 List of abbreviations ACH AIS AME AO AWA bpm CC cf CID CT D df FCC fem Fig ICD ICF ISS IVA kg K-S L LBM m Max min ml mm

Anterior corporal height Abbreviated Injury Score Acceptable maximum effort Arbeitsgemeinschaft für Osteosynthesefragen Anterior wedge angle Beats per minute Comprehensive classification Cardiac frequency Confidence interval of the difference Computer tomography Dimension Degrees of freedom Frontal corporal collapse Female Figure International classification of diseases International classification of function, disability and health Injury severity score Intervertebral angle Kilogram Kolmogorov-Smirnov Lumbar Lean body mass Month Maximum Minute Milliliter Millimeter

THORACOLUMBAR FRACTURES 12 LIST OF ABBREVIATIONS

126

MOS MRI n NIOSH NPAR p PAR PCH PIH PSH PVH RA RLOG RMDQ ROM SD SF36 SEM SIP SPSS t T uB USS VAS VO2-max Wmax WHO X-ray Z

Medical outcome study Magnetic resonance imaging Number National institute for occupational safety and health Nonparametric Probability Parametric Posterior corporal height Posterior intervertebral height Posterior segmental height Posterior vertebral height Regional angle Registratie letsels en ongevallen Groningen Roland Morris disability questionnaire Range of motion Standard deviation Short form-36 items Standard error of the mean Sickness impact profile Statistical Package for the Social Sciences Time Thoracic Unrecognised B Universal spine system Visual analogue scale Maximum volume of oxygen uptake Maximum lifting weight World Health Organization Radiograph Standard normal value

university of groningen thoracolumbar spinal fractures leferink, … · 2016. 3. 7. · fractures:...

Documents