DOI:10.2214/AJR.07.2701
AJR 2008; 190:1255-1259
© American Roentgen Ray Society
Detection of Bone Graft Failure in Lumbar Spondylodesis: Spatial Resolution with High-Resolution Peripheral Quantitative CT
Peter C. Strohm1,
David Kubosch2,
Thorsten A. Bley2,3,
Christoph M. Sprecher4,
Norbert P. Südkamp1 and
Stefan Milz4
1 Department of Orthopedic and Trauma Surgery, Albert-Ludwigs-University of
Freiburg Medical Center, Hugstetterstr. 55, 79106 Freiburg i Br.,
Germany.
2 Department of Diagnostic Radiology, University of Freiburg Medical Center,
Freiburg im Breisgau, Germany.
3 Present address: Department of Radiology, University of Wisconsin, Madison,
WI.
4 AO Research Institute, AO Foundation, Davos, Switzerland.
Received June 8, 2007;
accepted after revision November 7, 2007.
Address correspondence to P. C. Strohm
(peter.strohm{at}uniklinik-freiburg.de).
P. C. Strohm received financial support from Tutogen Medical for animal
investigation and implants (plates) from Aesculap.
Abstract
OBJECTIVE. In spinal surgery, anterior spondylodesis is often
combined with bone grafting, and graft integration is assessed with CT.
High-resolution peripheral quantitative CT offers a resolution of 82 µm.
The aim of this study was to compare the outcome of anterior spondylodesis as
assessed with three radiologic procedures.
MATERIALS AND METHODS. Monosegmental lumbar spondylodesis with
autologous iliac crest graft or solvent-preserved bovine cancellous bone was
performed on seven sheep. The fused spinal segments were explanted after 24
weeks and examined with clinical 64-MDCT, high-resolution peripheral
quantitative CT, and contact radiography. In 2D views, the area of the disk
space bridged by bone was assessed, and the grafts were examined for
fractures.
RESULTS. In three of seven sheep, clinical CT erroneously showed
stable consolidation, whereas contact radiography revealed a clearly visible
graft fracture, as did high-resolution peripheral quantitative CT. There was a
statistically significant difference (p = 0.038) between bone volume
assessed with clinical CT and that assessed with contact radiography. There
was an almost significant difference (p = 0.053) between volumes
assessed with high-resolution peripheral quantitative CT and clinical
MDCT.
CONCLUSION. High-resolution peripheral quantitative CT, a technique
approved for clinical use, has higher resolution in imaging of bone structure
than does 64-MDCT. Our results show that high-resolution peripheral
quantitative CT is superior to 64-MDCT in assessing osseous implant
integration after anterior spondylodesis. The specimen size limit, however,
prohibits in vivo use of this method in evaluation of the human spine. Our
results suggest that in clinical practice, persisting symptoms despite
radiologic findings of consolidated spondylodesis may be related to graft
failure, which cannot be detected with clinically available methods.
Keywords: bone graft • high-resolution peripheral quantitative CT • spinal surgery • spine • Xtreme CT
Introduction
Anterior stabilization is a procedure frequently performed in spinal
surgery to manage traumatic or degenerative instability. Autologous bone
transplants, such as iliac crest, fibula, and rib grafts, typically are used,
but homogenous and xenogenous transplants also are used in selected cases
[1,
2]. The reference standard for
noninvasive clinical assessment of osseous implant integration is CT
[3,
4]. MRI would theoretically
also serve the purpose and provide valid results
[5], but MRI is not suitable
for visualizing bone because of the lack of water protons. In addition,
osteosynthesis usually is performed with metal materials, such as internal
fixators, plates, and hook-based systems, which also can interfere with the
quality of CT images, increasing the difficulty of evaluation. The performance
of clinical CT is constantly being improved, however; 64-MDCT has become
standard at many hospitals.
Many of the decisions made during the treatment of spinal fusion patients
consequently are based on results obtained from CT investigations, especially
after the fusion operation has been performed. The time point for removal of
an internal fixator is determined on the basis of CT findings. That is, the
decision is based on a diagnosis of osseous integration of the implant and
therefore the assumption of mechanically stable bridging between the two
adjacent vertebral bodies. CT-based diagnosis also is important in cases in
which decisions concerning rehabilitation have to be made or in which there is
a need for an expert medical opinion. In many of these cases, the crucial
parameter is resolution, and the diagnosis has to be made carefully with an
imaging technique that has limited spatial resolution.
In experiments with micro-CT
[6–11]
and high-resolution peripheral quantitative CT
[12], the influence of
resolution on the validity of diagnosis has been studied with regard to the
success of spinal fusion. Contact radiographs of serial sections of vertebrae
embedded in methyl methacrylate have been evaluated as controls. This
procedure can be regarded as almost equivalent to histologic investigation.
The technique yields consecutive 2D views that can easily be compared with the
results of 64-MDCT and high-resolution peripheral quantitative CT
[4,
5]. The purpose of our
investigation was to compare, in imaging of serial sections of an animal model
of spinal fusion, the validity of results obtained with two CT techniques with
that of findings on contact radiography.
Materials and Methods
Anterior spondylodesis with bone graft and anterior angular stable locking
plate fixation (Macs TL, Aesculap) was performed on seven female sheep. The
animal investigation protocol was approved by the local review board and was
given the number G-04/16. In four sheep, bridging was achieved by insertion of
a commercially available bovine cancellous bone block (Tutobone, Tutogen
Medical). In three sheep, autologous tricortical iliac crest blocks were
implanted. In all sheep, spondylodesis was performed between the third and
fourth lumbar vertebrae through an anterolateral approach on the right side.
After the operation, the sheep were kept and cared for in a natural
environment of open and closed areas and sacrificed after 24 weeks.
Immediately after euthanasia, the section of the spine on which the
operation had been per formed was explanted, and the osteosynthesis material
removed. Clinical 64-MDCT was per formed on these specimens. The specimens
were then fixed in 100% methanol, and high-resolution peripheral quantitative
CT was performed. The next step was to embed the complete spinal segments in
methyl methacrylate. After polymer ization of the blocks, 200-µm-thick
serial sections were produced at an interval of 500 µm, and contact
radiographs were obtained on high-resolution radiographic film. The contact
radio graphs were first evaluated macro- scopically for assessment
of bone bridging of the inter vertebral space. After acquisition of the
contact radiographs, serial sections were stained with Giemsa eosin and
evaluated for the presence of artificially produced microfractures. No such
artifacts were detected, but the fracture gaps were filled with connective
tissue.
Bone integration of the grafts was evaluated with the three imaging
procedures. Special attention was paid to the extent of bone bridging across
the former intervertebral space. The first step of the evaluation procedures
was macroscopic assessment of the contact radiographs of the serial sections
(reference standard) and of the two CT data sets with the focus on determining
whether a graft was fractured. The second step of the evaluation was
software-assisted analysis of all CT scans and radiographs 24 weeks after the
operation to determine the absolute volume of the osseous parts of the graft
in the region of the intervertebral space.
Imaging
All conventional CT scans were performed on a 64-MDCT scanner (Somatom
Sensation 64, Siemens Medical Solutions) with a 0.37-second rotation time and
the following parameters: X-ray tube potential, 120 kV; effective tube
current, 680 mA; slice collimation, 64 x 0.6 mm; table feed, 9.2
mm/rotation; pitch, 0.24, allowing nominal isotropic resolution of 400
µm3; and result ing voxel dimension, 400 x 400 x 400
µm. High-resolution peripheral quantitative CT (XtremeCT unit, SCANCO
Medical) was performed with the following parameters: X-ray tube potential, 60
kV; effective tube current, 0.9 mA; 3,072 x 255 detector elements;
pitch, 82 µm; matrix size, 3,072 x 3,072, allowing nominal isotropic
resolution of 55 µm3; and resulting voxel dimension, 82 x
82 x 82 µm. Contact radiographs (Model No. 43855 A, Faxitron X-Ray)
of serial sections of undecalcified vertebrae were obtained with
high-resolution radiographic film (Struk turix-D3, Agfa-Gevaert). Images were
ex posed with the following parameters: accelerator voltage, 20 kV; tube
current, 3.0 mA; and resulting spatial resolution, approximately 1.2 x
1.2 µm [13].
Image Evaluation
The contact radiographs, which had almost the same resolution as histologic
images, were evaluated to determine whether osseous bridging was present in
the region of the intervertebral space. Because the contact radiographs
originated from serial sections cut at constant intervals, volume size was
determined by multiplication of the planimetric values by the interval (i.e.,
thickness) values. Morphometric analysis of the CT images (64-MDCT and
high-resolution peripheral quan titative CT) was performed with OsiriX medical
imaging software (OsiriX) and analysis of the contact radiographs with
Axiovision LE software (Zeiss).
To guarantee precise detection of the outer borders of the graft in the
sagittal sections, horizontal extension of the graft was assessed in the
so-called "live synchronization mode" (simultaneous display of
sagittal view and corresponding horizontal view). The volume was derived from
the surface data combined with knowledge of the slice thickness. Because the
clinical 64-MDCT image was recorded at a slice thickness of 2 mm and the graft
had a total length of 10 mm, the analysis required a minimum of six
sections.
View larger version (107K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —Sheep 6, consistently abnormal results among tested
techniques. 64-MDCT scan (A), high-resolution peripheral quantitative
CT scan (B), and contact radiograph (C) show lysis in region of
graft.
|
|
View larger version (133K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —Sheep 6, consistently abnormal results among tested
techniques. 64-MDCT scan (A), high-resolution peripheral quantitative
CT scan (B), and contact radiograph (C) show lysis in region of
graft.
|
|
View larger version (104K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C —Sheep 6, consistently abnormal results among tested
techniques. 64-MDCT scan (A), high-resolution peripheral quantitative
CT scan (B), and contact radiograph (C) show lysis in region of
graft.
|
|
To allow comparable evaluation modes for all three imaging procedures,
high-resolution peripheral quantitative CT investigations were based on the
same number of sections (slice thickness, 0.82 mm). Calculations for the
contact radiographs also were based on six histologic sections from
corresponding locations. For each 2D view, the corresponding volumes were
calculated by multiplication of the area values with an interval thickness of
2 mm. Statistical analysis was performed with the Mann-Whitney rank sum
test.
Results
In three of the seven sheep, conventional clinical 64-MDCT erroneously
showed stable consolidation, whereas contact radiography showed a clearly
visible but often very narrow fracture line. High-resolution peripheral
quantitative CT likewise revealed graft fracture
(Table 1). One sheep (sheep 6)
had complete graft resorption. The extensive defect was correctly diagnosed
with all three imaging techniques. Analysis showed that 50% of the graft
fractures were not evident on clinical 64-MDCT and were classified as
clinically consolidated with this method. In most of the sheep, sequestrum
formation in the region of the intervertebral space was recorded as a bone
bridge on 64-MDCT, whereas findings on high-resolution peripheral quantitative
CT and contact radiography led to the correct diagnosis (Fig.
1A,
1B,
1C). In two other sheep, the
fracture gap was either large enough (Fig.
2A,
2B,
2C) to see it on 64-MDCT
images, and in one sheep, the graft was not fractured (Fig.
3A,
3B,
3C).
With bone volume calculated on the basis of the contact radiographs as the
reference standard (Table 1),
no significant differences were found in comparison with the values obtained
with high-resolution peripheral quantitative CT (p = 0.383). There
was, however, a statistically significant difference (p = 0.038)
between the values obtained with conventional clinical 64-MDCT and those
obtained with contact radiography. With conventional CT, bone graft volume was
overestimated. Some values recorded were more than twice as high as those
obtained with contact radiography. Thus by a slight margin, an almost
significant difference (p = 0.053) between high-resolution peripheral
quantitative CT and conventional CT was found in this study
(Fig. 4).
Discussion
In clinical routine, CT is often used for detailed imaging of bone
structures. This method has become established in particular for the
visualization of mineralized osseous structures. Progress in technical design
is leading to increasingly high resolutions for clinically approved devices so
that questions can be addressed that were previously the prerogative of
experimental CT. Bauer et al.
[9], for example, performed
comparative investigations in the context of osteoporosis diagnostics to
evaluate clinical CT and micro CT and came to the conclusion that conventional
clinical CT offers adequate characterization of trabecular structure.
Thomsen et al. [10] found
high correlation between micro CT and histologic specimens whereby contact
radiography was regarded as equivalent to histologic investigation
[4]. Otsuki et al.
[14] and Ho and Hutmacher
[15] found that micro CT is a
suitable instrument for the visualization of patterns of bone ingrowth into
various bone substitutes. Nevertheless, distinct differences remain with
regard to the clinical value of the resolutions available for each of the
imaging technologies. High-resolution peripheral quantitative CT is, to our
knowledge, the first clinically approved device with a resolution comparable
with that of micro CT that can be used to scan larger specimens and even human
extremities in vivo. Our literature search yielded only one study
[12] of the use of such a
device in the context of osteoporosis diagnostics.
The aim of our study was to evaluate the influence of different imaging
capabilities by direct comparison of 64-MDCT, high-resolution peripheral
quantitative CT, and contact radiography in a clinically relevant animal
model. For this purpose we analyzed images of graft incorporation of anterior
spondylodesis of the lumber spine in sheep 24 weeks postoperatively. This
temporal interval corresponded approximately to the clinical follow-up
interval for this operation on human patients.
We found no significant difference between the bone volumes obtained with
contact radiographs and those obtained with high-resolution peripheral
quantitative CT. In contrast, the bone volumes obtained with clinical 64-MDCT
differed significantly from the volumes obtained with contact radiographs and
high-resolution peripheral quantitative CT. We concluded that with
conventional clinical 64-MDCT, bone volume tends to be overestimated owing to
large voxel size and the consequently large partial volume effects
[16,
17]. As a result of this
imaging characteristic, smaller mineralized objects (e.g., bone sequestra) are
frequently merged with larger adjacent bone structures.
Our results clearly indicate that the resolution of high-resolution
peripheral quantitative CT is superior to that of conventional clinical
64-MDCT and that in anterior spondylodesis, these differences frequently led
to incorrect qualitative assessment of bone bridging. This finding is
especially important because resolutions that preclude incorrect diagnosis are
not available for in vivo investigation of the human spine. In terms of
clinical routine, there may be a relation between persistent pain and
misleading diagnosis, which according to radiologic criteria suggest
successful anterior spondylodesis. We believe that the success rates of
graft-induced spinal fusion reported in the literature must be viewed
critically in light of our findings. Evaluation of equally subtle osseous
structures is at least possible in the extremities because high-resolution
peripheral quantitative CT allows investigation of regions with a diameter of
125 mm. Our results show that there is clinical demand for the same technology
for spinal in vivo diagnostics.
References
- Briem D, Rueger JM, Linhart W. Osseous integration of autogenous
bone grafts following combined dorso-ventral instrumentation of unstable
thoracolumbar spine fractures [in German].
Unfallchirurg 2003;106
: 195–203[CrossRef][Medline]
- Kossmann T, Ertel W, Platz A, Trentz O. Combined surgery for
fractures of the thoraco-lumbar junction using the inlay-span method [in
German]. Orthopade 1999;28
: 432–440[CrossRef][Medline]
- Moll R, Schindler G, Weckbach A. Evaluation of ventral
stabilization techniques for thoracolumbar fractures by helical computer
tomography [in German]. Rofo 2002;174
: 880–886[Medline]
- Link TM, Vieth V, Stehling C, et al. High-resolution MRI versus
multislice helical CT: which technique depicts the trabecular bone structure
best? Eur Radiol 2003;13
: 663–671[Medline]
- Phan CM, Matsuura M, Bauer JS, et al. Trabecular bone structure of
the calcaneus: comparison of MRI at 3.0 and 1.5 T with micro-CT as the
standard of reference. Radiology 2006;239
: 488–496[Abstract/Free Full Text]
- Laib A, Barou O, Vico L, Lafage-Proust MH, Alexandre C,
Rügsegger P. 3D microcomputed tomography of trabecular and cortical bone
architecture with application to a rat model of immobilization osteoporosis.
Med Biol Eng Comput 2000;38
: 326–332[CrossRef][Medline]
- Laib A, Kumer JL, Majumdar S, Lane NE. The temporal changes of
trabecular architecture in ovariectomized rats assessed by microCT.
Osteoporos Int 2001;12
: 936–941[CrossRef][Medline]
- Mulder L, Koolstra JH, Van Eijden TM. Accuracy of microCT in the
quantitative determination of the degree and distribution of mineralization in
developing bone. Acta Radiol 2004;45
: 769–777[CrossRef][Medline]
- Bauer JS, Issever AS, Fischbeck M, et al. Multislice-CT for
structure analysis of trabecular bone: a comparison with micro-CT and
biomechanical strength [in German]. Rofo2004; 176:709
–718[Medline]
- Thomsen JS, Laib A, Koller B, Prohaska S, Mosekilde L, Gowin W.
Stereological measures of trabecular bone structure: comparison of 3D micro CT
with 2D histologic sections in human proximal tibial bone biopsies.
J Microsc 2005;218
: 171–179[CrossRef][Medline]
- Teo JC, Si-Hoe KM, Keh JE, Teoh SH. Relationship between CT
intensity, microarchitecture and mechanical properties of porcine vertebral
cancellous bone. Clin Biomech (Bristol, Avon)2006; 21:235
–244[CrossRef]
- Sornay-Rendu E, Boutroy S, Munoz F, Delmas PD. Alterations of
cortical and trabecular architecture are associated with fractures in
postmenopausal women, partially independent of decreased BMD measured by DXA:
the OFELY study. J Bone Miner Res 2007;22
: 425–433[CrossRef][Medline]
- El Ghazaly, M. X-ray phase contrast imaging at the Mainz
microton MAMI [thesis]. Mainz, Germany: Johannes Gutenberg
Universität, Institut für Physik, 2005
- Otsuki B, Takemoto M, Fujibayashi S, Neo M, Kokubo T, Nakamura T.
Pore throat size and connectivity determine bone and tissue ingrowth into
porous implants: three-dimensional micro-CT based structural analyses of
porous bioactive titanium implants. Biomaterials2006; 27:5892
–5900[CrossRef][Medline]
- Ho ST, Hutmacher DW. A comparison of micro CT with other techniques
used in the characterization of scaffolds.
Biomaterials 2006;27
:1362
–1376[CrossRef][Medline]
- Siegel MJ, Luker GD. Pediatric applications of helical (spiral) CT.
Radiol Clin North Am 1995;33
: 997–1022[Medline]
- Wiemker R, Rogalla P, Blaffert T, et al. Aspects of computer-aided
detection (CAD) and volumetry of pulmonary nodules using MDCT. Br J
Radiol 2005; 78[spec
no 1]: S46–S56[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
