Keywords
basilar invagination - craniovertebral junction anomaly - irreducible atlantoaxial
dislocation
Introduction
The craniovertebral junction (CVJ) anomaly treatment underwent paradigm shift in the
management in the past two decades. From transoral decompression and posterior fixations
to only posterior approaches with C1C2 fixations with complete restoration of normal
alignment at the joints have become the norm now.[1]
[2]
[3]
[4]
[5] Goel et al pioneered and popularized the technique of atlantoaxial fixations (C1C2
fixation), minimizing the need for transoral decompression.[3]
[4] Continuing further with the modifications in the surgical procedures, Chandra et
al developed the novel technique of distraction, compression, and extension reduction
(DCER) for the complex anomalies associated with basilar invagination (IB) and atlantoaxial
dislocation (AAD) further obviating the need for transoral decompression.[5]
[6] The DCER technique involved occiput to C2 fusion and the further maneuvers for achieving
proper alignment at the C1C2 joint. However, the occipitocervical fusions have the
inherent disadvantages of restricting the neck movements to a major extent. The authors
in this article described a novel technique involving C1C2 fixations and subsequent
manipulations of distractions and compression to achieve alignment at the DCER to
achieve.
Materials and Methods
The study population included the patients with CVJ anomaly who had the C1C2 fixation
with joint distraction and manipulations to achieve alignment of the C1C2 joint. Patients
with CVJ anomalies having BI and the fixed AAD were included in the study.
Study Design
This is a retrospective observational study.
Preoperative Clinical Features and Evaluation
Preoperative evaluation included clinical assessment in the pre- and postoperative
period as per Nurick's grading.[7]
Each patient had magnetic resonance imaging (MRI) of the CVJ and computed tomography
(CT) of the CVJ in flexion and extension with reconstruction in coronal and sagittal
planes and the CT angio of the vertebral artery to enable the better assessment of
the CVJ anatomy.
Especially size of the lateral masses of C1 and the pedicles of C2 measurements and
also to look for any anomalous course of the vertebral artery, if exists, were noted
preoperatively.
The traditional indices such as Wackenheim clival canal line (normal being 0.96 +
2.2 mm below this line), Chamberlain line (normal being 2.36 + 2.6 mm below this line),
McRae line (5.86 + 1.6 mm below this line), and modified Ranawat line (normal 29.76
+ 2.6 mm) were used as criteria for diagnosing BI.[8]
[9]
[10]
Presurgical Planning
All patients underwent C1C2 fixation and C1C2 joint opening followed by insertion
of the metallic cages and the manipulations of the C1 and C2 to achieve reduction.
Operative Technique
Patients were intubated using the fiberoptic laryngoscope. They were positioned prone
with head in neutral position over the horseshoe head rest. Cervical traction of 5
to 7 kg was applied in all the cases using Gardner Well Tongs. Traction ensured the
distraction at the C1C2 joint rendering the C1 lateral mass and the C1C2 joint access
easier. It also helps in minimizing the venous congestion at the CVJ.
Midline skin incision from below inion up till C3 was given, and subsequent muscle
detachment and dissection was done to expose the C1 posterior arch and C2 lamina up
till the facets laterally.
As an initial step in cases of the occipitalized C1 arch and the foramen magnum rim
compressing the cervicomedullary junction, excision of the C1 arch in addition to
the rim of the foramen magnum, if necessary, was done. This facilitated the visualization
the C1 lateral masses for screw placement in an otherwise crowded foramen magnum as
a result of the BI. This decompression also prevents any inadvertent injury to the
already compromised cervicomedullary junction during manipulations at the C1C2 joint
subsequently.
The authors avoided resecting the C2 ganglion, by dealing with the bleeding around
using bipolar coagulation of venous plexus around it and retraction of the ganglion
to gain working space around the C1C2 joint.
The posterior surface of the C1 Lateral mass was exposed and the midpoint was chosen
as the entry point for the screw. A 3.5 mm diameter and 18 to 22 mm length were inserted
15 degrees medial and 15 degrees cranial under fluoroscopic guidance until the screw
tip reached the anterior cortex of the lateral mass.
C2 screw of similar size as that used in C1 was inserted with the entry point 5 mm
rostral and 1 mm medial to the inferomedial aspect of the inferior articulating surface
and directed 25 to 30 degrees medially and 10 to 15 degrees cranially.
The C1C2 joints were initially probed with a blunt dissector, followed by periosteal
elevator or a chisel with its broad tip parallel to the articular surfaces during
insertion, subsequently rotating it by 90 degrees, thereby wide opening the joint.
The articular surfaces are then scraped with curettes. In cases of vertically aligned
joints, engagement of the rod is done on one side and distracted to get partial correction
of the rotatory subluxation, which can help in engaging the cage snugly between the
joint surfaces of other side followed by rod placement. Even after the placement of
the rod, the joint can be accessed in an oblique direction easily for inserting a
cage. These simple tricks help engage cages in the C1C2 joints with vertical alignment.
Then the cages filled with bone chips are impacted into the disc space. We use cuboid-shaped
cervical cages of 5 to 9 mm (manufactured by General Surgical Company [India] Private
Ltd.) based on the degree of BI. This achieves the reduction in the BI.
Slight lordosis is given to the rods before placement in the screw heads. Compression
of the screw heads is done with the compressor tool to close the gap between the screw
heads almost ending them touching each other. This maneuver brings in posterior translation
of the C1 arch and anterior translation of the odontoid peg, resulting in the reduction
in the AAD.
Remaining bone chips are placed around the C1C2 joint space and closure is done.
Case 1
A young 21-year-old man presented to our services with paraesthesia all over the body
below the neck level and spasticity and resulting difficulty in walking for 6 months.
The patient had occipitocervical fusion (OCF) for BI with fixed AAD at another neurosurgical
center. Neurologic examination revealed spastic quadriparesis, short neck, and other
features of high cervical myelopathy.
X-ray, MRI, and CT of the CVJ showed features of BI with fixed AAD ([Fig. 1a]). OCF construct removal and foramen magnum decompression were performed followed
by C1 lateral mass and C2 pedicle screw placement. C1C2 joint opening and cage placements
and subsequent compression of the C1C2 screws over the rod achieved the reduction
in the BI and AAD ([Fig. 1b–e]). The patient reported improvement in symptoms from postoperative day 1, and at
6 months follow-up his Nurick's grade improved from grade 4 to 2.
Fig. 1 Case1: (a) MRI sagittal section showing BI with AAD and anterior obliteration of the CSF column
at CMJ. (b) Postoperative CT showing reduced BI and CSF column anteriorly visible at CMJ. (c, d) Preoperative images of CT sagittal section at facet level showing C1C2 joint alignment
and (e) showing corrected alignment at C1C2 after surgery.
Case 2
A 42-year-old male patient presented with torticollis since childhood had a trivial
trauma as a result of fall and fell unconscious for 10 minutes. Since then he started
having neck pain, tingling, and paraesthesia all over the body. He gradually developed
spastic quadriparesis and required assistance for walking. He improved Nurick's grade
from 4 to 2 with in the first week of surgery.
X-ray, MRI, and CT of the CVJ demonstrated a BI with irreducible AAD ([Fig. 2a]).
Fig. 2 Case 2: (a) CT sagittal sequence showing BI with AAD. (b) Postoperative CT showing reduction in BI and AAD.
The patient underwent foramen magnum decompression and following that a C1C2 screw
placement on the right side and C1C3 screw placement on the left side as the C2 pedicle
was thin, and there was injury to the vertebral artery during the attempted screw
placement. Distraction and compression after cage placement resulted in normal alignment
at the CVJ ([Fig. 2b]).
Case 3
A young 15-year-old male patient had BI with AAD presented with high cervical compressive
myelopathy symptoms had C1C2 fixation at another center presented to us with no improvement
in symptoms after surgery. He was reoperated, C1 screw readjustment and cage placement
were done with distraction and compression for realignment ([Fig. 3a, b]). The patient reported improvement of symptoms the next day. His preoperative Nurick's
grade was 4 and postoperatively it improved to 2.
Fig. 3 Case 3: (a) CT sagittal sequence with AAD and BI with Os odontoideum. (b) Postoperative images of CT showing reduction in BI and AAD with foramen magnum drilled.
Case 4
A 40-year-old man with neck pain and compressive myelopathy symptoms at the CVJ with
BI and AAD had restoration of the alignment at the CVJ with C1 and C2 fixation, distraction,
and compression. The patient reported improvement in symptoms on postoperative day
1. ([Fig. 4a, b]). He improved from Nurick's grade 5 to 2 postoperatively.
Fig. 4 Case4: (a) BI with AAD on CT sagittal section. (b) Patient's postoperative X-ray of CVJ showing cages in the C1C2 joint, and it demonstrates
the screw heads almost touching each other after compression of the screw heads leading
to reduction in the AAD.
Discussion
The CVJ anomaly treatment has undergone phenomenal change with pioneering work by
Goel et al.[3]
[4]
[11]
[12] The entire focus is now on atlantoaxial screw placement with joint manipulations.[5]
[6]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
The complex CVJ anomalies with BI and irreducible or fixed AAD have dislocation in
sagittal and coronal alignment with some rotational component in addition.[12]
[13] Salunke et al proposed that correction in one plane of the joint can well result
in the corrections in the other planes.[13]
Because the paradigm shift that has happened after pioneering work by Goel et al,
many modifications and additions for the techniques have come up over time. Technique
by Goel et al advocates C1C2 screw placement and opening of the joint with chisel
and manipulating it to place spacer or strut bone graft to achieve the alignment at
the C1C2 joint. The spacer size inserted in the joint space depends on the extent
of the possible distraction.
Chandra et al proposed the DCER involves the distraction of joint space with spacers
(in the occipito-C1/C2 joint space) followed by compression between the occiput and
C2, leading to extension at the O-C1/C2 joint, which in turn leads to reduction in
AAD and complete remodeling of the joint. The technique of distraction compression
and reduction reduces the anteroposterior dislocation along with the vertical dislocation.
This uses the spacer as a fulcrum and C2 laminar screws as long lever.
This compilation of cases by the authors is unique in the sense that the technique
used is essentially C1C2 fixation and the principles of DCER for reduction. In cases
with atrophic C1 lateral mass on one side resulting in torticollis, cages of different
sizes were used. Even in cases in which the C1C2 joint line is vertical, the engagement
of the cage in the joint is possible after the joint opening and engagement of the
rod in the screw heads on one side, which results in partial restoration of the alignment
in all three dimensions. The authors believe that the foramen magnum decompression
in cases of occipitalized C1 arch impinging posteriorly on the cervicomedullary junction
prior to all manipulations is advisable to prevent the inadvertent injury during manipulation.
Slight lordosis of the rod combined with compression of the screw heads to achieve
near approximation results in the attainment of the proper alignment at C1C2 joint.
Engagement of the rod in the screw heads itself corrects the rotational alignment
to an extent with the rest of the correction happening when the compression of the
screw heads is done over the rod. This alignment restoration happens much like what
happens in scoliosis correction of the spine when the engagement of the rod is done
in the screw heads following osteotomies rendering the vertebral segments mobile.
Therefore, the authors believe that opening of the joint and ability to engage a cage
in the joint with manipulations and maneuvers using a chisel as described by Goel[12] and adding a compression of the screw heads of C1C2 as described by Chandra et al[5] over a lordotic rod do the entire job for restoration of the alignment at CVJ.
The recent reports by Chandra et al and Saluke et al advocated the drilling of joints
to insert cages.[6]
[17]
[18]
[19] Though this is a small series, the authors demonstrated achieving the alignment
without the need for drilling the joint. In addition, drilling the joint is technically
more cumbersome and also theoretically it carries the prospect of subsidence of the
cage into the lateral mass of the C1 keeping in view the fact that the strong cortical
surface is rendered weak by drilling.
The authors also have considerable experience in using this technique of placing spacers
in the C1C2 joint in cases with partially reducible AAD to achieve a good alignment
at the CVJ.
Though the number of patients included in the study was less, the authors could arrive
at the aforementioned conclusions because of the long experience of treating these
complicated anomalies involving manipulation of the C1C2 joint.
Conclusion
The treatment of CVJ anomalies involving BI with fixed AAD has changed with the C1C2
joint realignment techniques. A combination of the C1C2 screw placement, followed
by joint opening, subsequent distraction, and compression of the screw heads can achieve
the alignment at the CVJ. These maneuvers have the potential to correct the alignment
in all planes of the C1C2 joint, including the rotational component.
