Multilevel Cervical Fusion Without Plates; Screws Or Autogenous İliac Crest Bone Graft.
Abstract
Objective: This prospective study was performed to evaluate the safety and efficacy of polyetheretherketone (PEEK) cages packed with demineralized bone matrix (DBM) mixed with autologous blood and curettage microchip material for treatment of multilevel cervical
disc disease and spondylosis without the use of plates; screws or autogenous iliac crest bone graft.
Material and methods: Sixteen patients underwent multilevel anterior cervical discectomy and fusion (ACDF) for a total of 42 levels.
Minimum follow-up was 18 months. Neurological outcomes were evaluated using the Japanese Orthopaedic Association (JOA) scoring
system; cervical lordosis and cervical fusion status was assessed on X-ray. Statistical analysis was performed to compare preoperative and
postoperative scores using a dependent t-test (P < 0.05).
Results: Eight patients underwent two-level; six underwent three-level and two underwent four-level operations. The fusion rate was
90.5% and non-fusion rate was 9.5%; but reoperation was not required for these patients in the follow-up period. Cervical lordosis
was preserved and neurological status was improved. No cage migration or cage failure occured.
Conclusion: ACDF using PEEK cages packed with DBM is a safe and efficient method for treatment of multilevel cervical disc disease
and spondylosis. It preserves cervical lordosis and obviates the complications related to iliac crest graft harvest and screw-plate fixation.
Ó 2006 Elsevier Ltd. All rights reserved.
Keywords: Cervical degenerative disc disease; Multilevel fusion; PEEK cage; Demineralized bone matrix
1. Introduction
patients.10–12 Moreover; even with solid fusion; kyphosis
often develops in multilevel discectomies with autogenous
Since the initial description by Robinson and Smith in
iliac crest graft fusion.10;13 Additionally; morbidity due to
1955;1 the technique of anterior cervical arthrodesis has
bone graft harvest remains high and can compromise the
been refined. Although anterior cervical discectomy and
satisfactory clinical result of cervical nerve root and spinal
interbody fusion (ACDF) for treatment of degenerative
cord decompression.14–16
cervical disease is a highly successful procedure;2–6 the suc-
Multilevel cervical discectomy is often combined with
cess rates decline in multilevel discectomies as the number
plate and screw fixation to maintain the spinal curvature;
of levels increase.7;8 Graft collapse with the use of autoge-
and increasing the graft fusion rate. However; plates and
nous bone has been reported in 20–30% of multilevel fusion
screws may cause complications; such as screw breakage;
screw pull out; esophagus perforation and spinal cord or
*
nerve injury.17–21
Corresponding author. Present address: GATA Haydarpasa Egitim
The deficiencies mentioned above have favoured ongo-
Hastanesi Norosirurji Servisi 34668 Uskudar; Istanbul; Turkıye. Tel.: +90
216 542 2661; fax: +90 216 348 7880.
ing development of cage technology.14;15;22 There has been
E-mail address: mehmetnusretdemircan@yahoo.com (M.N. Demir-
a rapid increase in the use of cervical spine interbody fusion
can).
cages in view of their theoretical ability to prevent graft
0967-5868/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jocn.2006.02.026
724
M.N. Demircan et al. / Journal of Clinical Neuroscience 14 (2007) 723–728
collapse; with the potential advantage of indirect foraminal
A radiologist and a spine surgeon independently as-
decompression by restoration and preservation of inverte-
sessed fusion status with no knowledge of the clinical out-
bral height and lordosis. However; in most studies; cages
come. The operative segment was deemed to be fused if
filled with autologous cancellous bone were used. Although
there was no change in position of the fused levels on dy-
this is likely to reduce graft harvesting complications;
namic views (flexion and extension). Additionally; the
donor site pain still remains a common problem.23–25
radiological images were evaluated using the classification
The purposes of this study were to evaluate the safety
of anterior fusion proposed by Vavruch et al.25 In this clas-
and efficacy of polyetheretherketone (PEEK) cages and
sification; Type 1A is defined as bridging bone anterior and
to determine if it is possible to eliminate donor site and
through the disc space; 1B as bridging bone anterior but
plate complications and to achieve good outcomes for mul-
not through the disc space; 2A as bridging bone not ante-
tilevel discectomy and fusion if the cage packed with
rior but through the disc space; and 2B as no bridging bone
demineralized bone matrix (DBM) is used without plates.
at all. The radiological outcomes were classified as ‘non-
To the authors’ knowledge; a similar prospective study
fusion’ if 2B healing was observed; and as ‘fusion’ if 1A;
has not been reported in the literature.
1B or 2A healing was observed at the levels subjected to
surgery.
2. Materials and methods
Lateral X-rays were performed to evaluate the spinal
curve pre- and postoperatively. The Ishihara Curvature In-
In our department; between February 2000 and Sep-
dex (ICI) was used for this evaluation.27 A straight line was
tember 2002; 16 patients (11 women and 5 men) with a
drawn from the posterior border of the dens to the poster-
mean age of 56 years (range; 40–63) underwent multilevel
ior border of C7. Another line was drawn from the poster-
ACDF using PEEK cages (Spine Next; Bordeaux-France
ior border of C4 perpendicular to the first line; in which the
and Eurospine; L’Hay-Les-Roses; France) packed with
intersected length was measured in millimeters as the de-
DBM (Grafton; Osteotech; Eatontown; NJ; USA) mixed
gree of spinal curvature. A positive intersected length indi-
with autologous blood and curettage microchip material.
cates the degree of lordosis. If the intersected length is
Only patients operated for treatment of degenerative cer-
negative; it indicates kyphosis. When the intersected length
vical disc disease and spondylosis were included in this
is zero; the spinal curve is referred to as straight.
study. Patients with trauma; infection and neoplasms were
Statistical analysis was performed to compare preopera-
excluded. Conforming to international ethical standards;
tive and postoperative scores using dependent t-tests using
all patients were given detailed information on the opera-
SPSS V.12.0 (SPSS Inc.; Chicago; IL; USA).
tion; the follow-up protocol and radiological investiga-
tions; and their consent was obtained. Indication for
3. Results
operation was intractable radiculopathy; myelopathy; or
a combination of both due to nerve root or spinal cord
There were five men and 11 women between the ages of
compression and compatible magnetic resonance imaging
40 and 63 years; with a mean age of 56.3. Patients’ demo-
(MRI) findings.
graphic data and number of levels treated are shown in
The operative procedure was performed as described by
Table 1. Eight patients underwent two-level discectomy;
Smith and Robinson.16 The disc; posterior longitudinal lig-
six patients underwent three-level discectomy and two pa-
ament; and osteophytes; including the posterior part of the
tients underwent four-level discectomy. All patients were
uncinate process; were removed under the surgical micro-
followed up for at least 18 months (range; 18–34 months).
scope. Endplate cartilage was also removed with a high-
After surgery; none of the patients suffered neurological
speed drill and curette. Curettage microchip material was
deterioration. There were no complications during the
conserved. Cages were inserted into the disc space after
immediate postoperative period; and X-rays confirmed
packing with DBM mixed with autologous blood and
appropriate positioning of the vertebral cages.
curettage microchip material. During cage placement; cra-
Mean JOA scores were 13.7 ± 1.34 and 16.4 ± 0.97 pre-
nial traction was applied. The wound was closed with re-
and postoperatively. Patients were significantly better
approximation of two anatomic planes over a suction
(p = 0.004) clinically.
drainage system. All patients wore a Philadelphia collar
for 6 weeks after surgery. Most patients received physio-
therapy after removal of the collar.
Table 1
Clinical and radiological follow-up was performed at the
Patients’ demographic data (n = 16)
3rd; 12th; and 18th months postoperatively. In addition to
Mean age (years)
56.3 ± 6.69 (range; 40–63)
standard neurosurgical examination; outcomes were as-
Women
11
sessed with Japanese Orthopedic Association (JOA) scor-
Men
5
ing.26 We evaluated spinal curves; mobility and fusion
Two treated levels
8
status with X-ray. Four planes of X-rays were used; includ-
Three treated levels
6
ing anterior-posterior; neutral and flexion and extension
Four treated levels
2
Total treated levels
42
lateral views.
M.N. Demircan et al. / Journal of Clinical Neuroscience 14 (2007) 723–728
725
Preoperative mean ICI was 10.4 ± 3.72 and postopera-
There were three patients with non-fusion levels: two
tive mean ICI was 10.1 ± 3.14. The difference was insignif-
non-fusion levels in one patient with a four-level operation
icant (p > 0.05); therefore; preoperative lordosis was said to
(C5-6 and C6-7 levels) (Fig. 3b); one non-fusion level in
be preserved at the 18th month after surgery.
one patient with a four-level operation (C6-7 level); and
At the final follow-up; the fusion rate (Types 1A; 1B;
one level in one patient with a three-level operation (C5-6
and 2A) was 90.5% (38/42 levels). The rate of Type 1A fu-
level). In long-term follow-up; imaging showed no cage
sion was 35.7% (15/42 levels) (Fig. 1); the rate of Type 1B
failure or dislodgement. Reoperation for non-fusion was
fusion was 40.5% (17/42 levels) (Fig. 2); the rate of Type
not necessary. In addition; no mobility was seen on dy-
2A fusion was 14.32% (6/42 levels) (Figs. 3a and b); and
namic X-ray films at any operated segments.
the rate of Type 2B non-fusion was 9.5% (4/42 levels).
Fig. 1. X-ray of Type 1A fusion.
Fig. 2. X-ray of Type 1B fusion.
Fig. 3. Type 2A fusion. (a) preoperative MRI. (b) X-ray of non-fusion (Type 2B) at C5-6 and C6-7 levels and Type 2A fusion at C3-4 and C4-5 levels.
726
M.N. Demircan et al. / Journal of Clinical Neuroscience 14 (2007) 723–728
4. Discussion
of the cervical spine; enhance the fusion rate; and correct
spinal curve to physiologic lordosis.6;15 In ACDF; addi-
Anterior cervical discectomy and interbody fusion is an
tional plate fixation has been reported to result in a higher
efficacious procedure used to treat a variety of cervical
fusion rate; lower reoperation rate; and better pain re-
spinal disorders; including spondylosis; myelopathy; herni-
lief.9;12;13;31 However; in their retrospective study; Das
ated discs; trauma; and degenerative disc disease.
et al.13 studied 38 patients who had arthrodesis with cylin-
The success of this procedure relies on thorough decom-
drical titanium cages filled with autologous bone graft har-
pression and development of a solid osseous fusion.2–6;25;28
vested from the operative site and screw-plate fixation; and
Brown et al.29 reviewed serial X-rays after anterior cervical
they reported the rate of pseudoarthrosis was 6–8% for
fusion performed in a total of 139 levels in 98 patients and
one-level and 15–46% for treatment of several levels. Over-
found arthrodesis in 97% of patients who underwent auto-
all; in three- and four-level discectomies the successful fu-
graft procedures. In their series; Savolainen et al.30 found a
sion rate decreases 18–82%; even when a cervical spine
98% fusion rate in patients who underwent procedures with
locking plate is used.37–39 Moreover; plating has complica-
autograft. According to the results obtained from other
tions. Plate complication rate varies from 2.2–24.0%20;34
series; for single-level discectomy with autogenous bone fu-
and includes screw pullout;21;40 screw breakage;21 injury
sion; ACDF can achieve a 92–100% fusion rate20 and 70–
of the laryngeal nerve;8 injury of oesophagus;19 injury of
90% neurologic and symptomatic improvement.5;6
spinal cord or root; injury of vertebral artery; and wound
Although; arthrodesis with autologous iliac crest graft is
infection.21 Additionally; the operative time is usually
considered as the biological and biomechanical standard
longer.
in anterior cervical reconstruction;9;31 the morbidity of
These complications of classical fusion procedures fa-
the iliac bone harvest can often tarnish these re-
voured ongoing development of cage technology. Because
sults.4;15;23–25;28;30;32 Silber et al.15 observed that 26.1% of
of the advantages of these devices; the use of cages in
patients reported pain at the donor site. Summer et al. also
ACDF operations has been increasing in popularity. In
reported chronic pain in the donor site in 25% of 290 pa-
parallel with this; there are several different types of inter-
tients.24 According to Arrington et al.; in addition to the
body fusion cages commercially available.22–25;33 Cage-
minor complications of the donor site (superficial infec-
assisted ACDF has proven to be a safe and effective proce-
tions; hematoma; cosmetic problems; etc.) there were ma-
dure for the treatment of degenerative disc disease. It has
jor complications in 5.8% of cases; requiring therapeutic
been reported that the cage achieves excellent fusion rates
modifications; surgical revision and prolongation of hospi-
ranging from 93.1–100%.3;25;28;32;33;40;41 In our series; the
talization.14 Castro et al.2 reported a donor site complica-
fusion rate was 90.5%; comparable to the related literature.
tion rate of 22% in their series. In addition to the donor
There were three patients with non-fusion. Although these
site problems; the graft complication rate in autogenous
non-fusions were seen in four-level operated patients no
bone graft can be high.3;8;9;28;33;34 Matge reviewed patients
clinical signs or radiographic mobility of pseudoarthrosis
who had undergone autogeneous bone fusion procedures
were observed during the follow-up period.
and found that there were many graft-related complica-
With the use of a cage donor site morbidity was
tions; including migration (2.1–4.6%); kyphosis (3–10%)
avoided.2;25;33 In our study; no cage failure or migration
and pseudoarthosis (1–3%).32
was encountered; even in patients who underwent fusion
In the Cloward procedure; the best results have been re-
at more than two levels. The use of the cage was found
ported for young male patients with soft disc disease; at the
to preserve the spinal lordosis and the height of the foram-
single level.35;36 Multilevel anterior cervical discectomy and
ina.7;28;33;40;42 Bartels et al. reported that the cervical cage
fusion still remains a difficult problem. Autogeneous bone
effectively increased foraminal height even after 1 year;
does not maintain spinal instability in multilevel discec-
which contributed to decompression of the nerve root.42
tomy very well and the graft complication rate in autoge-
The wedge shape of the device may contribute to restora-
nous bone graft in multilevel fusion is higher than at the
tion of lordosis. In accordance with the current study; they
single level.4;8;9;13 Graft collapse with autogenous bone is
showed that the PEEK cage resulted in preservation of the
reported in 20–30% of multilevel fusions.10–12 Moreover;
preoperative lordosis.
it has been reported that even with solid fusion; kyphosis
In our study; we preferred to use PEEK cages for mul-
often develops in multilevel discectomies with autogenous
tilevel fusion; because of the lower reported complication
iliac crest graft fusion.10;17 The literature also reports a
rates.28 PEEK is a semicrystalline polyaromatic linear
consistent rate of 10–12% non-fusion for single-level ante-
polymer that provides a combination of strength; stiffness;
rior discectomy and autogenous bone fusion; 20–27% for
toughness; and environmental resistance.33 In laboratory
two-level; and approximately 30–56% for three-level
studies; this device demonstrated excellent resistance to
fusions.7–9 It is clear that the success rates decline as the
compression.43 It is also biocompatible.44 The cage has
number of levels increase.
been shown to have a stimulatory effect on the protein con-
In the light of these reports; in multilevel ACDF proce-
tent of osteoblasts.45 In one animal study; osteocalcine pro-
dures; augmentation with plate fixation; may seem to be
duction; alkaline phosphatase activity; and the
preferable. Plate fixation may decrease the micromovement
proliferation of fibroblasts were enhanced after the inser-
M.N. Demircan et al. / Journal of Clinical Neuroscience 14 (2007) 723–728
727
tion of a PEEK cage.46 Furthermore; the cage structure
9. Zdeblick TA; Ducker TB. The use of freeze-dried allograft bone for
(two titanium spikes on the upper and bottom frame; in
anterior cervical fusions. Spine 1991;16:726–9.
10. Katsuura A; Hukuda S; Imanaka T; et al. Anterior cervical plate used
addition to the retention teeth on the surface of the upper