Endoscopic Submandibular Retropharyngeal Approach to the Craniocervical Junction and Clivus: an Anatomic Study

Accepted Manuscript Endoscopic submandibular retropharyngeal approach to the craniocervical junction and clivus: an anatomical study Henri Salle, George de Albuquerque Cavalcanti Mendes, Clément Gantois, Justine Lerat, Nouman Aldahak, François Caire PII:

S1878-8750(17)31070-7

DOI:

10.1016/j.wneu.2017.06.162

Reference:

WNEU 6033

To appear in:

World Neurosurgery

Received Date: 9 April 2017 Revised Date:

25 June 2017

Accepted Date: 28 June 2017

Please cite this article as: Salle H, Cavalcanti Mendes GdA, Gantois C, Lerat J, Aldahak N, Caire F, Endoscopic submandibular retropharyngeal approach to the craniocervical junction and clivus: an anatomical study, World Neurosurgery (2017), doi: 10.1016/j.wneu.2017.06.162. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Endoscopic submandibular retropharyngeal approach to the craniocervical junction and clivus: an anatomical study. Henri Salle1, George de Albuquerque Cavalcanti Mendes3, Clément Gantois1, Justine

1

RI PT

Lerat2, Nouman Aldahak4, François Caire1, 5

Department of Neurosurgery, Limoges University Hospital, France

2

Department of ENT surgery, Limoges University Hospital, France

3

Santacoop, Santa Efigênia, Belo Horizonte, Brasil

4

SC

Department of Neurosurgery, Lariboisière Hospital, Paris, France.

Univ. Limoges, CNRS, XLIM, UMR 7252, F-87000 Limoges, France

Key words:

M AN U

5

Upper cervical spine, craniocervical junction, clivus, skull base, chordoma, endoscopic surgery, retropharyngeal approach.

Abbreviations used in this paper

ARPA: anterior retropharyngeal approach

AC C

EP

VA: vertebral artery

TE D

CCJ: craniocervical junction

Corresponding author:

François Caire, MD, PhD Service de neurochirurgie CHU de Limoges, Hôpital Dupuytren 2 avenue Martin Luther King 87000 Limoges, France [email protected] Phone: +33 5 55 05 65 21 (sec.) Fax: +33 5 55 05 65 27 1

ACCEPTED MANUSCRIPT SUMMARY (244 words)

Introduction: Surgery of the craniocervical junction and clivus is technically demanding. For many

RI PT

years, we have used the submandibular retropharyngeal approach for surgery of the upper cervical spine, especially Hangman’s fracture. We hypothesized that

submandibular gland resection could offer a significant cranial enlargement of the

SC

operative field, up to the clivus. Our aim in this work was to assess the feasibility of endoscope-assisted retropharyngeal approach to the craniocervical junction and

M AN U

clivus. Methods:

Eight anatomical specimens were used, including four silicon-injected specimens. We performed a submandibular retropharyngeal approach with gland resection, and

TE D

then we exposed the craniocervical junction and clivus. We drilled the C2 vertebral body, odontoid process, C1 anterior arch and the clivus. We noted 8 anatomical landmarks that were easily identified on each anatomical specimen. These

EP

measurements were designed to quantify the exposure of the clivus and craniocervical junction after bone resection.

AC C

Results:

Submandibular approach was feasible in all specimens. The main dimensions of the area of dural exposure after bone drilling were as follows: mean width between C1 lateral masses: 19 mm (range 17-20 mm); at the tip of clival window: 18mm (16-20 mm); distance between the C3 vertebra and the tip of the window within the clivus: 57 mm (range 55-60 mm). Conclusions:

2

ACCEPTED MANUSCRIPT Endoscopic submandibular retropharyngeal approach provides a simple and straightforward access to the CCJ. It provides also a very convenient exposure of the

AC C

EP

TE D

M AN U

SC

RI PT

clivus. This technique could be added to the techniques used for this difficult surgery.

3

ACCEPTED MANUSCRIPT INTRODUCTION

Surgery of the craniocervical junction is challenging for neurosurgeons because it is a relatively rare condition and because its approach is technically demanding.

RI PT

Standard Smith-Robinson anterior cervical approach is not convenient because the access is too lateral and too tangential to the upper cervical spine. Its upper limit is indeed approximately the lateral horn of the hyoid bone. It does not allow an

SC

orthogonal access to the cervical spine above C4.

Transoral surgery, initially described by Fang and Ong1 provides an orthogonal and

M AN U

direct access to C1 and C2, but the lateral exposure is limited, and the rate of complications, especially infectious, is significant. Extended transoral approaches have been proposed in order to enlarge the exposure caudally2, but glossotomy and mandibulotomy are fraught with risks and technical difficulties.

TE D

That is the reason why a retropharyngeal approach could be an interesting alternative. Southwick and Robinson3 gave the first description of this technique. Their followers, in line with McAfee et al.4, mainly used this approach to access the

EP

upper cervical spine. In fact, our own experience with a retropharyngeal approach was based on C2C3 fusion procedures for Hangman’s fracture with C2C3 instability.

AC C

Nevertheless, several authors in the past had reported a limited but relevant experience with surgery of the clivus using a retropharyngeal approach5-8. In our experience, we had observed that submandibular gland resection could offer a significant cranial enlargement of the operative field, as recently suggested by Skaf et al.9. Moreover, the use of an endoscope could also make the procedure more convenient and improve the visualization in the depth of the surgical field. Based on these considerations, we proposed to assess in a cadaveric study the feasibility and

4

ACCEPTED MANUSCRIPT the reliability of a retroparyngeal approach to the craniocervical junction under endoscopic control, and to quantify the surgical exposure that could be reached in

RI PT

practice.

MATERIAL AND METHODS

SC

Dissections:

Eight anatomical specimens (5 women, 3 men) without obvious cervical pathology,

M AN U

including four specimens injected, were dissected. The surgical procedure was conducted using an endoscope (Karl Storz), and standard micro neurosurgical instruments. Bone drilling was performed using a Midas Rex Legend high-speed drill

Surgical procedure:

TE D

(Medtronic Inc.).

The head was positioned in slight extension and rotated towards the left by

EP

approximately 15 to 30°. We performed a U-shaped in cision arching 4 to 5 cm below the mandible (figure 1), in order to avoid injury to the marginal mandibular branch of

AC C

the facial nerve (figure 2). Platysma muscle was incised, and then the muscle flap was rolled superiorly. The carotid artery was exposed. The superior thyroid and lingual arteries were dissected, and ligated if necessary. At this stage, the submandibular gland was exposed. The facial vein, which is located at the upper and outer part of the submandibular gland, was ligated. Then we removed the submandibular gland. Mobilization and upward reflexion of the gland exposed the hypoglossal nerve medial to the tendon of the digastric muscle. The

5

ACCEPTED MANUSCRIPT posterior border of the mylohyoid muscle was identified and retracted anteriorly. This mobilization exposed the Wharton’s duct, which was ligated. The lingual nerve was exposed and mobilized upward after section of the fibers entering the submandibular gland. This procedure allowed the complete mobilization and removal of the

RI PT

submandibular gland (figure 3).

At this stage, we observed the submandibular triangle formed by the lower edge of the mandible, and the anterior and the posterior belly of the digastric muscle joined at

SC

their apex by digastric tendon. The hypoglossal nerve was identified. The mylohyoid muscle was retracted anteriorly to give a direct access to the retropharyngeal space.

M AN U

The retropharyngeal space was gently dissected. A retractor was introduced under the lower edge of the mandible, and then a rigid 30° endoscope ( figure 4). We identified the anterior tubercle of C1 (figure 5), indicating the midline. The longus colli muscles were detached from the vertebrae, and the C2-C3 intervertebral disc

TE D

was exposed. At this stage, the working area extended to the anterior arch of C1. The longus colli muscles were detached and reflected laterally, thus exposing C2 (figure 6) and C1 (figure 7). The mucosa was detached to the sphenooccipital

EP

synchondrosis and base of the vomer. Surgical swab was tailored and introduced to protect the mucosa. Under endoscopic control, the C2C3 intervetebral disc was

AC C

removed.

The C2 vertebral body was carefully hollowed out until only a transparent posterior cortical rim remained (figure 8). The dura was then exposed using a small rongeur (figure 9). The anterior arch of C1 was drilled (figure 10) as far as the lateral masses and the C0-C1 and C1-C2 joints. The alar and apical odontoid ligaments were resected, and the odontoid process was removed en bloc. The transverse atlantal ligament was removed using a small rongeur. The dura was thus exposed between

6

ACCEPTED MANUSCRIPT C2C3 intervertebral disc and the foramen magnum (figure 11). The anterior aspect of the foramen magnum was opened, and the clivus was carefully drilled (figure 12).

Anatomical measurements

RI PT

We noted 8 anatomical landmarks that were easily identified on each anatomical specimen (figure 13). We measured the following distances: (1) width of the lower plate of the C2 vertebra between points 1 and 3, (2) width of the area of resection of

SC

the C1 anterior arch between point 4 and 5, (3) width of the window created within the clivus between points 7 and 8, (4) distance from the upper plate of C3 to the

M AN U

anterior border of the foramen magnum between point 2 and 6, (5) distance from the anterior border of the foramen magnum and the tip of the clival window between points 6 and 8. These measurements were designed to quantify the exposure of the

EP

RESULTS

TE D

clivus and craniocervical junction after bone resection.

Retropharyngeal approach to the CCJ and complications encountered

AC C

Using a U-shaped incision arching 4 to 5 cm below the mandible, there was no obvious injury to the mandibular branch of the facial nerve. The key step of the procedure was the identification of the submandibular triangle, which defines the surgical corridor. The submandibular gland resection was easily achieved without obvious nerve or vascular injury. We identified the hypoglossal nerve, but the nerve was not obviously exposed during the surgical procedure. Resection of the styloid process was not necessary.

7

ACCEPTED MANUSCRIPT The access to C2 vertebral body was simple and direct in all cases. We obtained an orthogonal access to the C2C3 intervertebral disc space. We decided after the first surgery to mark the midline with a marker pen on the anterior border of the foramen magnum, anterior tubercle of C1 (figure 5) and

head movement during the procedure.

RI PT

anterior wall of C2 or C3, in order to keep satisfactory visual landmarks in case of

For the first two specimens, there was a lesion in the mucosa when detaching it from

SC

the clivus. For the following procedures, the dissection was performed carefully and we protected the mucosa with a surgical swab after dissection.

Anatomical measurements

M AN U

Bone drilling was performed without adverse event.

The measurements designed to quantify the bone resection performed within the

TE D

clivus and craniocervical junction are summarized in the table 1. The dimensions of the area of dural exposure after bone drilling were as follows: -

Mean width at the level of C2C3 disc space: 17 mm (range 15-19 mm);

EP

between the C1 lateral masses: 19 mm (range 17-20 mm); at the tip of clival window: 18mm (16-20 mm); Length between C3 vertebra and the tip of the window within the clivus

AC C

-

(distance 2-8): 57 mm (range 55-60 mm).

DISCUSSION

8

ACCEPTED MANUSCRIPT Conventional approaches to the craniocervical junction Conventional approaches to the craniocervical junction can be classified in posterolateral, antero-lateral and anterior. The posterolateral approach10, also called the “far lateral approach”, is commonly

RI PT

used for the surgery of intradural lesions: schwannomas, meningiomas, vascular pathology. The patient is in ¾ prone position, so that the line of sight reaches the anterior aspect of the craniocervical junction. Prone position can also be used. The

SC

skin incision has the shape of a hockey stick. One disadvantage of the skin incision is that the muscle flap retracted laterally may obscure the angle of approach. If we go

M AN U

down toward the tip of the mastoid, we can retract the skin not only laterally but more downward. Muscle dissection is performed in one layer with subperiostal dissection on the occipital bone.

We may use additional drilling (transcondylar or supracondylar approach), depending

TE D

on the pathology, the location and the attachment of the lesion. Indeed, when we have to deal with a tumor, the pathology creates the working space and has already retracted the brainstem. The line of sight to reach the anterior margin of the foramen

EP

magnum goes through the tumor volume and we don’t need to drill the condyle, except for vascular lesions. However, for vascular lesions such as PICA or vertebral

AC C

artery aneurysm or vertebra-basilar junction aneurysm, it can be interesting because we cannot retract the brainstem and the line of sight may go through the condyle to expose the aneurysm11-13. Anterolateral approach reaches the craniocervical junction and the upper cervical spine between the carotid artery and the vertebral artery. This approach is mainly used for extradural lesions of the anterior aspect of the craniocervical junction, the odontoid process, the lateral mass of C1 and C2, or the C2 body: chordomas,

9

ACCEPTED MANUSCRIPT metastasis and primary bone tumors11. The patient is in supine position, rotation of the head 30 to 45° with a little extension. When ro tating the head, we must be careful with the position of C1-C2 vertebral artery, because the C1-C2 and the C0-C1 segments of the vertebral artery become parallel on both side of the posterior arch of

RI PT

C1. The insertion of nuchal muscles is cut and muscles are retracted posteriorly. One difficult step is to identify the XI nerve. First, we dissect between sternocleidomastoid muscle and jugular vein, and then we follow the posterior edge of digastric muscle to

SC

identify the XI nerve. We have to free the nerve and retract it with the fat. At this stage, we expose the suboccipital triangle with the transverse process of C1. Then,

M AN U

we detach all the muscles tied to the transverse process of C1: the superior, inferior oblique, and elevator scapular muscle. We retract these muscles medially to expose the anterior arch of C1; above it lies the C0-C1 segment of the vertebral artery and below it the C1-C2 segment of the vertebral artery. There is usually no need for

TE D

transposition of the VA, because the VA is already invaded by the tumor. If we really need to transpose the VA, it should be done with gentle manipulation, and there is a high risk of dissection.

EP

The anterior surgery of the upper cervical spine and CCJ is usually performed via a transoral approach, which is historically well known by neurosurgeons, even if 14

AC C

technically demanding

. Several extended approaches have been proposed and

are still performed: mandibulotomy, mandibuloglossotomy, palatotomy

2

. They

provide an excellent exposure of the CCJ, C1 and C2 vertebrae, but the rate of postoperative morbidity is high. There is a real risk of infection, especially after placement of bone graft, due to the communication between the oral cavity and the spine.

10

ACCEPTED MANUSCRIPT When using a standard transcervical Smith-Robinson approach, the working angle to reach C1 and C2 is too acute to be comfortable15 - approximately 15°- and the retropharyngeal approach is at last the only alternative to the transoral approach.

RI PT

Retropharyngeal approach to the craniocervical junction and upper cervical spine

The first occurrence of the retropharyngeal approach to the craniocervical junction or

SC

the upper cervical spine in the literature remains uncertain due to the variety of keywords that can be used for search. The first description probably arose in 19573.

M AN U

In 1962, White and Albin 16 ligated the basilar artery via an anterior surgical approach through the clivus in a monkey. A careful search led us to read a paper published in 1966 by the San Francisco team 5, which remained unnoticed. The authors described in detail, with nice drawings, the removal of a clivus chordoma using a “transcervical

TE D

transclival approach” which was actually retropharyngeal.

In contrast, the papers describing and naming a few years later the “extrapharyngeal” or “retro-pharyngeal” approach are well known

4,17

. Since then, a very

anatomical work

EP

limited number of papers were dedicated to this surgical approach, based either on 15,18,19

or on clinical cases

20-25

. Table 2 presents a brief summary of

AC C

procedures identified in the literature as retropharyngeal approaches. Most cases are C2C3 fusions for Hangman’s fracture. The use of retropharyngeal approach for other indications is very limited. In our team, the retropharyngeal approach was initially used to expose the C2C3 intervertebral disk, because the classical transcervical approach provides an oblique field of view that makes discectomy and fusion difficult. A retropharyngeal approach, in contrast, provides an excellent exposure orthogonal to the vertebral bodies. We

11

ACCEPTED MANUSCRIPT use it in daily practice, for example in anterior fusion in case of Hangman’s fracture with C2C3 instability. Several authors reported recently similar experience

21,26-28

,

with good results. In this case, a “standard” retropharyngeal approach is usually performed: the submandibular gland is retracted upward, the tendon of the digastric

RI PT

muscle is either sectioned or dissected from the hyoid bone and elevated, and the hypoglossal nerve is gently retracted superiorly 21,23. This simple and quick dissection indeed provides a good exposure between C2 and C4 vertebrae. The upper limit of

SC

the operative field is usually the body of C2, but several authors report their experience with anterior fusion between C1 and C3 or C4. The anterior arch of C1 is

or a titanium mesh cage 20-23,25.

M AN U

then used to anchor the upper extremity of a locking plate after positioning of a graft

Resection of the submandibular gland

advantages.

TE D

Resection of the submandibular gland, proposed by Skaf et al.9, offers two main

The first one is to improve the surgical exposure. In our experience, the access to the

EP

anterior arch of C1 with a “conventional” retropharyngeal approach is often oblique and somewhat difficult. For this reason, we tried to improve the surgical exposure in

AC C

the direction of the craniocervical junction. We observed that this problem could be easily addressed by submandibular gland resection, as demonstrated by Skaf et al. in 2007 9. Mobilization and upward reflection of the gland exposes the hypoglossal nerve medial to the tendon of the digastric muscle. The posterior border of the mylohyoid muscle is identified and retracted anteriorly. Wharton’s duct is ligated and divided. The lingual nerve is identified, attached to the submaxillary gland superiorly. The parasympathetic fibers entering the gland are ligated and transected, allowing

12

ACCEPTED MANUSCRIPT the lingual nerve to retract into the floor of the mouth. Submandibular gland resection is quick and simple, and can be performed if necessary with the help of an ENT surgeon. The surgical corridor is then completely modified: it is limited cranially by the mandible and caudally by the body of digastric muscle. The operative field is

RI PT

considerably extended cranially, up to the clivus (Figures 14 and 15), providing an orthogonal access to C1. The resection of C1 and odontoid process is therefore considerably facilitated, as well as C1 plating. Based on our own observations, a

SC

retropharyngeal approach with submandibular gland resection allows simple and quick exposure and bone drilling between C3 and the foramen magnum in all cases.

M AN U

The midline has to be identified, because of head rotation. The use of an endoscope offers a high quality imaging in the depth of the operative field. It is therefore possible to remove extradural tumors invading C1 and C2, as well as intradural tumors of the midline. Moreover, when combined with a sub-digastric approach, the operative field

TE D

extends at least to the fourth cervical vertebra (figure 15).

The second advantage of submandibular gland resection is to decrease the risk of neurapraxia to the hypoglossal nerve. Without gland resection, the retraction of the

EP

hypoglossal nerve, even if “gentle”, is difficult to control and often prolonged, leading to a serious risk of postoperative palsy

21,25

. After gland resection, the surgical

AC C

corridor is enlarged cranially, the hypoglossal nerve is identified and should be easily spared. There is therefore theoretically no risk of hypoglossal nerve injury.

Retropharyngeal approach to the clivus Most papers regarding retropharyngeal approach deal with upper cervical spine surgery, but submandibular gland resection also makes the surgery of the clivus and craniocervical junction possible. Our own results show that:

13

ACCEPTED MANUSCRIPT -

retropharyngeal approach with submandibular gland resection provides a large exposure up to the vomer.

-

Under control of endoscope, we can drill the clivus, thus creating a bone defect from the anterior rim of the foramen magnum up to the vomer,

-

RI PT

approximately 2 cm in width.

The window created by bone drilling is limited laterally by the internal carotid artery and the groove of the inferior petrosal sinus (petro-occipital fissure). The main difficulty encountered in this cadaveric study was the risk of injury of

SC

-

the pharyngeal mucosa, when peeling it off from the clivus, thus creating a

M AN U

communication between working area and nasal cavities. This risk was previously reported by Stevenson et al 5. Agrawal et al.

18

compared in a cadaveric study the working areas of clival exposure

obtained with transoral and retropharyngeal (without submandibular gland resection)

TE D

approaches: respectively 546 ± 72 mm2 and 874 ± 75 mm2, i.e. a 62% improvement. They concluded that the gain provided by this approach consisted in the change in working angle, in working distance to the surgical target, and in working area for both

EP

ipsilaterally and contralaterally-seated lesions. Finally, the only obstacle is the lack of familiarity of neurosurgons with the retropharyngeal space, which is better known by

AC C

ENT surgeons.

To our knowledge, only two cases of clivus chordomas resected via a retropharyngeal approach were reported in 1966 and 1970

5,6

. Stevenson et al.

performed a preoperative tracheostomy, in order to maintain tight closure of the jaws and to obtain a better retraction of the larynx and pharynx. This procedure was never repeated, to our knowledge. Moreover, they removed the muscles (hyoglossus and constrictor pharynges medius) from the lateral hyoid bone, which was resected

14

ACCEPTED MANUSCRIPT between the great and lesser horn. This technical nuance, in our experience, is of interest in the case of a “conventional” retropharyngeal approach below the digastric muscle, because it improves the medial exposure, but useless when the gland is resected.

described:

pterygomaxillary

transmandibular

29

,

rhinoseptal

34

30,31

,

RI PT

Removal of clival tumors is known to be difficult. Many approaches have been transoral

32,33

,

, and of course endocospic endonasal approach

35,36

unilateral which is

SC

probably the most widely used today. The variety of techniques used shows that none is ideal. We think that retropharyngeal approach with gland resection could

M AN U

have a place in our therapeutic arsenal. Figure 15 illustrates the field of view that can be obtained via this approach, easily exposing the area comprised between the sphenooccipital synchondrosis and the third cervical vertebra.

Moreover, the

distance between the skin and the target is shorter than with other techniques: if all

TE D

roads lead to the clivus, retropharyngeal approach is probably one of the most simple.

The feasibility of such anterior transclival surgery is also illustrated by two papers

artery aneurysm

EP

published in 1967 reporting successful clipping of a vertebral artery 8

7

and a basilar

via the approach described one year earlier by Stevenson et al 5.

AC C

In both cases, the authors described a window created in the clivus by bone drilling (repectively 2x2 cm and 1x3 cm), followed by midline incision of the dura.

Endoscope-assisted retropharyngeal approach This work shows that the use of a conventional rigid endoscope is feasible. Retropharyngeal approach to the CCJ can be performed with the help of a surgical microscope, as exposed by Stevenson et al. in 1966 5. Nevertheless, it requires the

15

ACCEPTED MANUSCRIPT placement of a retractor below the mandible, and there is probably a real risk of pharyngeal edema when the surgery is prolonged. In the paper by Stevensons et al., we can note that the patient underwent a tracheostomy, clearly visible on the postoperative photograph provided by the authors. In endoscopic procedure, the

RI PT

need for soft tissue retraction is reduced, as the surgical corridor is limited to the introduction of endoscope and surgical instruments. The risk of vascular and nervous injury should therefore be limited.

SC

It provides also a good vision in the depth of the operative field, even for the ispilateral side if we use a 30° endoscope. The us e of an endoscope is probably not

M AN U

necessary for C1 anterior arch, odontoid process and C2 vertebral body resection, but useful for clival exposure and resection, and certainly for dural opening. There is to date no example in the literature of endoscopic retropharyngeal approach to the CCJ. Baird et al.

15

described an endoscopic transcervical approach, which

TE D

was actually a classical Smith-Robinson approach, almost tangential to the odontoid process and unable to provide a correct access to the clivus. In our experience, the main landmark during endoscope-assisted procedure is the

EP

tubercle of atlas: it is easy to identify during initial exposure, and it remains a simple landmark for midline position during further drilling. Nevertheless, the tubercle of

AC C

atlas is sometimes very prominent, and has to be drilled in order to expose the foramen magnum.

In the operative room, the head of the patient is fixed using a Mayfield clamp. An endoscope-holding arm is also usefull.

Complications of ARPA

16

ACCEPTED MANUSCRIPT The risk of infection of the operative site is not precisely known, due to the limited number of patients that were operated upon using a retropharyngeal approach

25

.

The most common side effect seems to be transient difficulties in swallowing, up to two months after the surgery 9 due to a prolonged retraction of the pharynx during the

also been reported

RI PT

surgical procedure. Transient palsy of the mandibular branch of the facial nerve has 37

. We think that (1) the lower aspect of skin incision has to be

located five centimeters below the mandible in order to make sure that the facial

SC

nerve will not be severed, and (2) resecting the submandibular gland instead of retracting it should minimize the risk of facial nerve injury 9. Postoperative palsy of

M AN U

the hypoglossal nerve is always possible when reaching the CCJ by an anterior approach, either transcervical or retropharyngeal

4,21

. Cases of posteoperative XII

palsy are probably not systematically reported in the literature. We assume with Skaf et al

9

that submandibular gland resection should significantly reduce the risk of

TE D

hypoglossal injury and neurapraxia, by avoiding the need for upward retraction of the nerve.

Regarding surgery of the clivus, Stevenson et al.

5

reported no postoperative

Wissinger et al.

EP

complication in a patient with preoperative signs of severe brainstem compression. 8

reported postoperative hypoglossal palsy due to a voluntary nerve

AC C

cutting that is quite obviously not necessary. Dural opening and clival defect have to be closed carefully: Fox et al.

7

reported the occurrence of a postoperative CSF

collection within the retropharyngeal space that required second surgery and transient shunt placement.

Biomechanical issues

17

ACCEPTED MANUSCRIPT There is a lack of robust data regarding the biomechanics of craniocervical junction after surgical procedures involving this region. A retropharyngeal approach can be used for the resection of tumors invading C2, followed by grafting and fusion between C1 and C3 or even C4

20,23

, but there is a

RI PT

risk of failure of the construct. The biomechanical value of such a fusion is unclear, as well as is the place for associated posterior fusion. A retropharyngeal approach also allows direct anterior transarticular atantoaxial fusion 24.

without opening of the foramen magnum

7,8

SC

For clival surgery, drilling can be limited to a window within the body of the clivus and without risk of postoperative

M AN U

instability of the CCJ. But clival drilling can also be associated with the resection of the anterior arch of C1 and odontoid process, as in our own cadaveric study. In this case, occipitocervical or atlantoaxail fusion could be discussed… but Stevenson et al. in a similar situation decided to cancel the fusion that was planned, because

TE D

cinefluorographic studies revealed no significant instability 5. A halo vest was also not used.

For these reasons, we think that the question of the indications and techniques for

EP

fusion after CCJ surgery has to be addressed in future works.

AC C

Conclusion

Submandibular retropharyngeal approach provides a simple and straightforward access to the upper cervical spine, orthogonal to the C2C3 intervertebral disc space. We think that the resection of submandibular gland is of interest for clival exposure, because it provides a comfortable original working corridor above the digastric muscle and hypoglossal nerve, thus limiting the risk of nervous injury.

18

ACCEPTED MANUSCRIPT This approach is suitable for the surgery of midline lesions straddling the craniocervical junction: clival chordomas involving the upper cervical spine, foramen magnum meningiomas, or basilar invagination in rheumatoid arthritis. These indications for retropharyngeal approach should be developed in the coming years,

RI PT

probably with the help of endoscopy. Technical nuances have to be evaluated, and neurosurgeons have to learn this simple approach, maybe with the help of ENT surgeons. Lastly, biomechanical data is mandatory in order to clearly establish the

SC

indications of occipitocervical or atlantoaxial fusion after surgery of the CCJ and

M AN U

clivus.

Acknowledgments:

We thank Mr. Tyler Clay for manuscript correction.

TE D

Funding:

This work was supported by grant 53-2014 from “Fondation des Gueules Cassées” (Union des Blessés de la Face et de la Tête). The funding source had no

EP

involvement in the conduct of the research or preparation of the article.

AC C

Conflict of interest: None.

19

ACCEPTED MANUSCRIPT Table 1. Distances measured between the anatomical landmarks (in mm). 1-3: width of the lower plate of C2 vertebra. 4-5: width of the area of resection of C1 anterior arch. 7-8: width of the window created within the clivus. 2-6: distance from the lower plate of C2 to the anterior border of the foramen magnum. 6-8: distance from the

RI PT

anterior border of the foramen magnum to the tip of the clival window. Note the progressive enlargement of the extent of clival resection with the experience of the

SC

authors.

Table 2.

M AN U

Surgical procedures identified in the literature as based on retropharyngeal

AC C

EP

TE D

approach. Anatomical studies were excluded.

20

ACCEPTED MANUSCRIPT ILLUSTRATIONS

Figure 1. U-shaped skin incision. Skin flap is rolled upward.

(arrows) and the submandibular gland (*).

RI PT

Figure 2. Muscle flap is rolled upward. See the mandibular branch of the facial nerve

SC

Figure 3. Exposure of the surgical corridor before (a) and after resection of the

M AN U

submandibular gland (*).

Figure 4. Exposition of the surgical corridor after submandibular gland resection, during the instruction of the endoscope. The submandibular triangle is limited at its top by the mandible, and below by the anterior and posterior bellies of the digastric

TE D

muscle.

Figure 5. After introduction of the endoscope within the repropharyngeal space, we

EP

note the position of the midline with a surgical marker (here on the anterior arch of the atlas), before detaching the longus colli muscles. The bottle-shaped silhouette of

AC C

the midline at C1C2 level is clearly visible.

Figure 6. Subperiosteal exposure of C2 anterior wall (*). Longus colli muscle is reflected laterally (**). The pen marks on C2C3 disc (bottom) and on the C1 tubercle (top) are clearly seen.

21

ACCEPTED MANUSCRIPT Figure 7. Subperiosteal exposure is prolonged from C2 (*) to the anterior arch of the atlas (**). The pen marks still indicate the midline (arrows).

Figure 8. Drilling of the C2 vertebral body. The odontoid process (*) and the anterior

RI PT

arch of C1 (**) are intact.

Figure 9. Exposure of the dura (*) and left lateral recess after removal of the vertebral

SC

body. Dotted line indicates the base of the odontoid process, and ** the C1 anterior

M AN U

arch.

Figure 10. The anterior arch of the atlas is progressively drilled, thus exposing the odontoid process.

TE D

Figure 11. After the resection of the transverse ligament of atlas, the dura is exposed from C3 (**) up to the foramen magnum (*).

EP

Figure 12. Exposure of the dura above the anterior arch of C1, after drilling the clivus.

AC C

Figure 13. Schematic illustration representing 9 standardized points used to quantify the surgical exposure. Points 1 to 3 mark the lower plate of C2 vertebral body. Points 4 and 5 were positioned at the lateral aspects of C1 anterior arch resection. Point 6 was the lower tip of the clivus on the midline, and points 7 to 9 the limits of clival resection. Dotted line illustrates the lines of bone resection.

22

ACCEPTED MANUSCRIPT Figure 14. Schematic representation of the area drilled at the end of the procedure. The basilar artery is represented by transparency through the dura.

Figure 15. Drawing illustrating the working areas obtained with ARPA (left) and with

RI PT

conventional approaches (right). Yellow: submandibular retrophryngeal approach. Extension to C3 and C4 is of course possible (dotted line) when working under the digastric muscle. Blue: endoscopic endonasal approach. Red: transoral approach.

AC C

EP

TE D

M AN U

SC

Green: endoscopic transcervical approach as described by Baird et al. 16.

23

ACCEPTED MANUSCRIPT REFERENCES

1. Fang HSY, Ong GB. Direct Anterior Approach to the Upper Cervical Spine. J Bone Joint Surg Am, 1962 ; 44 (8): 1588 -1604.

RI PT

2. Youssef AS, Sloan AE. Extended transoral approaches: surgical technique and analysis. Neurosurgery. 2010 ;66(3 Suppl):126-34.

3. Southwick WO, Robinson RA. Surgical approaches to the vertebral bodies in

SC

the cervical and lumbar regions. J Bone Joint Surg Am. 1957 J;39-A(3):63144.

M AN U

4. McAfee PC, Bohlman HH, Riley LH Jr, Robinson RA, Southwick WO, Nachlas NE. The anterior retropharyngeal approach to the upper part of the cervical spine. J Bone Joint Surg Am. 1987 ;69(9):1371-83.

5. Stevenson GC, Stoney RJ, Perkins RK, Adams JE. A transcervical transclival

TE D

approach to the ventral surface of the brain stem for removal of a clivus chordoma. J Neurosurg. 1966 ;24(2):544-51. 6. Bartal AD, Heilbronn YD. Transcervical removal of a clivus chordoma in a 2-

EP

year-old child. Reversal of quadriplegia and bulbar paralysis. Acta Neurochir (Wien). 1970; 23(2):127-33.

AC C

7. Fox JL. Obliteration of midline vertebral artery aneurysm via basilar craniectomy. J Neurosurg. 1967 ;26(4):406-12.

8. Wissinger JP, Danoff D, Wisiol ES, French LA. Repair of an aneurysm of the basilar artery by a transclival approach. Case report. J Neurosurg. 1967 ; 26(4):417-9.

24

ACCEPTED MANUSCRIPT 9. Skaf GS, Sabbagh AS, Hadi U. The advantages of submandibular gland resection in anterior retropharyngeal approach to the upper cervical spine. Eur Spine J. 2007 ;16(4):469-77. 10. Karam YR, Menezes AH, Traynelis VC. Posterolateral approaches to the

RI PT

craniovertebral junction. Neurosurgery. 2010; 66(3 Suppl):135-40.

11. Bruneau M, George B. Foramen magnum meningiomas: detailed surgical approaches and technical aspects at Lariboisière Hospital and review of the

SC

literature. Neurosurg Rev. 2008 ; 31(1):19-32.

12. George B(1), Dematons C, Cophignon J. Lateral approach to the anterior

M AN U

portion of the foramen magnum. Application to surgical removal of 14 benign tumors: technical note. Surg Neurol. 1988 ; 29(6):484-90. 13. George B, Lot G, Boissonnet H. Meningioma of the foramen magnum: a series of 40 cases. Surg Neurol. 1997 ; 47(4):371-9.

TE D

14. Hsu W(1), Wolinsky JP, Gokaslan ZL, Sciubba DM. Transoral approaches to the cervical spine. Neurosurgery. 2010 ; 66(3 Suppl):119-25. 15. Baird CJ, Conway JE, Sciubba DM, Prevedello DM, Quiñones-Hinojosa A,

EP

Kassam AB. Radiographic and anatomic basis of endoscopic anterior craniocervical decompression: a comparison of endonasal, transoral, and

AC C

transcervical approaches. Neurosurgery. 2009 ; 65(6 Suppl):158-63.

16. White RJ, Albin MS. The technique and results of ligation of the basilar artery in monkeys. J Surg Res. 1962 ; 2:15-8.

17. de Andrade JR, Macnab I. Anterior occipito-cervical fusion using an extrapharyngeal exposure. J Bone Joint Surg Am. 1969 ; 51(8):1621-6. 18. Agrawal A, Cavalcanti DD, Garcia-Gonzalez U, Chang SW, Crawford NR, Sonntag VK, Spetzler RF, Preul MC. Comparison of extraoral and transoral

25

ACCEPTED MANUSCRIPT approaches to the craniocervical junction: morphometric and quantitative analysis. World Neurosurg. 2010 ; 74(1):178-88. 19. Singh H, Harrop J, Schiffmacher P, Rosen M, Evans J. Ventral surgical approaches to craniovertebral junction chordomas. Neurosurgery. 2010 ; 66(3

RI PT

Suppl):96-103.

20. Yang X, Wu Z, Xiao J, Teng H, Feng D, Huang W, Chen H, Wang X, Yuan W, Jia L. Sequentially staged resection and 2-column reconstruction for C2

SC

tumors through a combined anterior retropharyngeal-posterior approach:

Operative):ons184-93

M AN U

surgical technique and results in 11 patients. Neurosurgery. 2011 ; 69(2 Suppl

21. Park SH, Sung JK, Lee SH, Park J, Hwang JH, Hwang SK. High anterior cervical approach to the upper cervical spine. Surg Neurol. 2007 ; 68(5):51924

TE D

22. Behari S, Banerji D, Trivedi P, Jain VK, Chhabra DK. Anterior retropharyngeal approach to the cervical spine. Neurol India. 2001; 49(4):342-9. 23. Vender JR(1), Harrison SJ, McDonnell DE. Fusion and instrumentation at C1-

9.

EP

3 via the high anterior cervical approach. J Neurosurg. 2000 ; 92(1 Suppl):24-

AC C

24. Koller H, Kammermeier V, Ulbricht D, Assuncao A, Karolus S, van den Berg B, Holz U. Anterior retropharyngeal fixation C1-2 for stabilization of atlantoaxial

instabilities: study of feasibility, technical description and

preliminary results. Eur Spine J. 2006 ; 15(9):1326-38. 25. Hodges SD(1), Humphreys SC, Brown TW Jr, Eck JC, Covington LA. Complications of the anterior retropharyngeal approach in cervical spine

26

ACCEPTED MANUSCRIPT surgery: a technique and outcomes review. J South Orthop Assoc. 2000 ; 9(3):169-74. 26. Hur H, Lee JK, Jang JW, Kim TS, Kim SH. Is it feasible to treat unstable

approach? Eur Spine J. 2014 ; 23(8):1641-7.

RI PT

hangman's fracture via the primary standard anterior retropharyngeal

27. Li Z(1), Li F, Hou S, Zhao Y, Mao N, Hou T, Tang J. Anterior discectomy/corpectomy and fusion with internal fixation for the treatment of

Spine. 2015 Apr;22(4):387-93.

SC

unstable hangman's fractures: a retrospective study of 38 cases. J Neurosurg

M AN U

28. Xu H(1), Zhao J, Yuan J, Wang C. Anterior discectomy and fusion with internal fixation for unstable hangman's fracture. Int Orthop. 2010 Feb;34(1):85-8. 29. Crumley RL, Gutin PH. Surgical access for clivus chordoma. The University of California, San Francisco, experience. Arch Otolaryngol Head Neck Surg.

TE D

1989 ; 115(3):295-300.

30. Guiot G, Rougerie J, Bouche J. The rhinoseptal route for the removal of clivus chordomas. Johns Hopkins Med J. 1968 ; 122(6):329-35.

EP

31. Pellerin P, Ghestem M, Lesoin F, Autrique P. [Approach to the clivus via maxillary bipartition. Report of a case of chordoma]. Ann Chir Plast Esthet.

AC C

1990 ; 35(6):475-9.

32. Makhmudov UB(1), Tcherekayev VA, Tanyashin SV. Transoral approach to tumors of the clivus: report of two cases. J Craniofac Surg. 1992 ; 3(1):35-8.

33. Crockard HA(1), Sen CN. The transoral approach for the management of intradural lesions at the craniovertebral junction: review of 7 cases. Neurosurgery. 1991 ; 28(1):88-97.

27

ACCEPTED MANUSCRIPT 34. Fournier HD, Hue AS, Laccourreye L. [Extensive lower clivus chordomas. Removal using the unilateral transmandibular approach]. Neurochirurgie. 2008 ; 54(2):63-71. 35. Vellutini Ede A, Balsalobre L, Hermann DR, Stamm AC. The endoscopic

Neurosurg. 2014; 82(6 Suppl):S106-15.

RI PT

endonasal approach for extradural and intradural clivus lesions. World

36. Stippler M, Gardner PA, Snyderman CH, Carrau RL, Prevedello DM, Kassam

SC

AB.Endoscopic endonasal approach for clival chordomas. Neurosurgery. 2009; 64(2):268-77.

M AN U

37. Laus M, Pignatti G, Malaguti MC, Alfonso C, Zappoli FA, Giunti A. Anterior extraoral surgery to the upper cervical spine. Spine (Phila Pa 1976). 1996 ;

AC C

EP

TE D

21(14):1687-93.

28

ACCEPTED MANUSCRIPT TABLE 1 Distance between anatomical landmarks (in mm) 4-5 7-8 2-6 6-8 19 18 39 16 20 20 40 17 18 19 38 17 17 17 39 19 21 16 37 19 18 19 36 24 19 18 34 25 20 17 36 23

1-3 17 15 19 18 16 17 16 18

2-8 55 57 55 58 58 60 59 59

RI PT

Specimen 1 2 3 4 5 6 7 8

AC C

EP

TE D

M AN U

SC

Table 1. Distances measured between the anatomical landmarks (in mm). 1-3: width of the lower plate of C2 vertebra. 4-5: width of the area of resection of C1 anterior arch. 7-8: width of the window created within the clivus. 2-6: distance from the lower plate of C2 to the anterior border of the foramen magnum. 6-8: distance from the anterior border of the foramen magnum to the tip of the clival window. Note the progressive enlargement of the extent of clival resection with the experience of the authors.

ACCEPTED MANUSCRIPT TABLE 2 Reference

Indications (number of patients)

Submand. gland Approach-related complications resection

Stevenson, 1966 Chordoma (1)

no

Trachostomy

McAfee, 1987

no

XII nerve palsy with dysphagia (2)

no

Instrumentation failure (1)

no

Dysphagia (3), infection (1),

Tumors (10) Injuries (4)

Vender, 2000

Foramen magnum meningioma (2) Rheumatoid arthritis (2) C2 fracture (2) Rickets (1)

Hodges, 2000

Disc herniation (8) Various spinal cord compressions (6)

Behari, 2001

C2 tumor (1)

Tracheostomy (1)

no

Respiratory insufficiency (2)

Degenerative (2) C2 fractures (4) Jefferson fracture (1) C1C2 tumor (1) Park, 2007

Hangman's fractures (11) C2 epidural hematoma (1) C2 chordoma (1)

Skaf, 2007

TE D

C3C4 tumor (1) OPLL (1)

Pharyngeal fistula (1),

M AN U

OPLL (2) Koller, 2006

SC

Others (2)

RI PT

Osteomyelitis (1)

Degenerative (3)

no

Instrumentation failure (1)

no

Dysphagia (3)

yes

Troubles swallowing (2)

C2 fracture (2)

Rheumatoid arthritis (1)

Hangman's fractures (28)

no

No complication reported

Yang, 2011

C2 tumors (11)

no

Troubles swallowing (2)

Hur, 2014

Hangman's fractures (17)

no

Dysphagia (1)

Li, 2015

Hangman's fractures (38)

no

Neurological (3), hematoma (2),

AC C

EP

Xu, 2010

Hoarsness (2)

Table 2. Surgical procedures identified in the literature as based on retropharyngeal approach. Anatomical studies were excluded.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

-

Retropharyngeal approach provides a direct access to the upper cervical spine.

-

Submandibular gland resection offers cranial enlargement of the operative

RI PT

field. -

The use of a conventional rigid endoscope is feasible.

-

It provides a large exposure, between the C2C3 intervertebral disk and the

EP

TE D

M AN U

This new approach could be useful for surgery of the CCJ and clivus.

AC C

-

SC

vomer.

1

ACCEPTED MANUSCRIPT

Article Title : Endoscopic submandibular retropharyngeal approach to the craniocervical junction and

RI PT

clivus: an anatomical study.

Authors :

Henri Salle, George de Albuquerque Cavalcanti Mendes, Clément Gantois, Justine

SC

Lerat, François Caire.

M AN U

I certify that there is no actual or potential conflict of interest in relation to this article.

Name François Caire

AC C

EP

TE D

Date 30 mars 2017