Académique Documents
Professionnel Documents
Culture Documents
39 (2006) 943–958
For more than 100 years, surgical decompression of the orbit has been
used to treat the severe proptosis and optic neuropathy associated with
Graves’ disease. Although decompression techniques involving removal
each of the four walls of the orbit have been described [1–4], the Walsh-
Ogura decompression [5] described in the 1950s was favored by most otolar-
yngologists. This operation uses the familiar Caldwell-Luc approach to
remove the inferior and medial orbital walls, allowing the enlarged orbital
fat and muscles to decompress into the ethmoid and maxillary sinus cavities.
Soon after the introduction of transnasal endoscopic sinus surgery in the
mid 1980s, surgeons began to experiment with endoscopic orbital surgery.
Endoscopic orbital decompression was first described by Kennedy [6] and
Michel [7] in the early 1990s. Enhanced visualization of key anatomic land-
marks allowed for safe and thorough decompression of the entire medial or-
bital wall, as well as the medial portion of the orbital floor. This improved
visualization is most notable in the region of the orbital apex, a critical area
of decompression in patients with optic neuropathy. These advantages have
allowed the endoscopic approach to replace the Walsh-Ogura procedure as
the technique of choice for orbital decompression.
The benefits of endoscopic instrumentation for orbital decompression
can be similarly applied when operating in the vicinity of the optic nerve.
Decompression of the optic nerve involves complete removal of the bone
0030-6665/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.otc.2006.06.003 oto.theclinics.com
944 PLETCHER et al
that forms the medial wall of the optic canal. Although the indications for
and surgical techniques of orbital decompression are fairly well established,
controversy exists regarding both the indications and extent of surgery nec-
essary for optic nerve decompression.
Technique
The patient is positioned in the supine position, and topical vasoconstric-
tion is achieved with topical oxymetazoline (0.05%) pledgets. The eyes
are maintained within surgical field, protected with scleral shields. Image-
guidance systems may be used at the surgeons discretion. Local injection
of lidocaine 1% with 1:100,000 epinephrine is administered along the lateral
nasal wall in the region of the maxillary line (a bony eminence that extends
from the anterior attachment of the middle turbinate to the root of the
inferior turbinate).
Surgery begins with an incision just posterior to the maxillary line. The
uncinate process is medialized and removed, exposing the natural ostium
of the maxillary sinus. With orbital decompression it is important to widely
open the maxillary sinus to achieve adequate access to the orbital floor and
prevent blockage of the ostium from orbital fat, which protrudes following
decompression. Enlargement is performed primarily in a posterior direction
as extension of the antrostomy too far anteriorly risks of damage to the
Fig. 1. Bone removed during endoscopic orbital decompression includes the medial orbital wall
and the medial aspect of the orbital floor.
946 PLETCHER et al
Fig. 2. Once the medial orbital wall has been exposed, Blakesly forceps are used to remove frag-
ments of the lamina papyracea.
ENDOSCOPIC ORBITAL AND OPTIC NERVE DECOMPRESSION 947
and optic nerve decompression into standard orbital decompression are un-
clear, and may lead to inadvertent injury to the nerve.
Removal of the orbital floor can be technically challenging, depending on
its thickness. Only that portion of the floor that is medial to the infraorbital
nerve is removed. A spoon curette is used to engage the orbital floor at its
medial extent and down-fracture the bone (Fig. 3). The bone of the orbital
floor is thicker than that of the medial orbital wall, and significant force may
be required for this maneuver. If the spoon curette is not sturdy enough for
this portion of the procedure the heavier mastoid curette may be used. The
bone may fracture in one large piece, typically with a natural cleavage plane
at the canal of the infraorbital nerve, or, more frequently, it fractures into
several small pieces. A 30-degree endoscope and angled forceps may facili-
tate bone removal while preserving the infraorbital canal as the lateral limit
of dissection.
Once the lamina papyracea and medial orbital floor have been removed,
the periorbita is fully exposed. A sickle knife is then used to open this fascial
layer. Care must be taken to avoid ‘‘burying’’ the tip of the sickle knife and
potentially injuring the underlying orbital contents such as the medial rectus
muscle. The periorbital incision should be initiated at the posterior limit of
decompression (just anterior to the sphenoid face) and brought anteriorly so
that prolapsing fat does not obscure visualization. Parallel incisions are per-
formed along the ethmoid roof and orbital floor. To minimize the risk of
postoperative diplopia, a 10 mm-wide sling of fascia overlying the medial
rectus muscle may be preserved while the remainder of the periorbita is re-
moved using angled Blakesley forceps (Fig. 4) [10]. In patients with optic
neuropathy, the fascial sling technique is not used to allow maximal
Fig. 3. A spoon curette is used to down-fracture the medial portion of the orbital floor.
948 PLETCHER et al
Fig. 4. Following bony decompression, a sickle knife is used to incise the periorbita. In patients
without optic neuropathy, a sling of fascia may be preserved overlying the medial rectus muscle
to minimize postoperative diplopia.
Fig. 5. Following decompression, orbital fat protrudes into the ethmoid and maxillary spaces
(A). With the ortbital sling technique, a strip of periorbita remains to support the medial rectus
muscle and minimize postoperative diplopia (B).
Results
The goals of orbital decompression vary depending upon the indication
for the procedure. In patients with compressive optic neuropathy,
950 PLETCHER et al
restoration of visual deficits is the key outcome, while in patients with cor-
neal exposure or severe proptosis, ocular recession may be the primary end
point. The reported incidence of improvement following endoscopic orbital
decompression for Graves’ orbitopathy ranges from 22% to 89% [6,12,13].
This wide variation in results reflects the diverse patient populations and
definitions of improvement. Postoperative deterioration of visual acuity oc-
curs in less than 5% of patients [7,12,13]. Ocular recession as a result of en-
doscopic decompression alone averages 3.5 mm (range 2–12 mm). The
addition of concurrent lateral decompression to the endoscopic procedure
provides an additional 2 mm of globe recession [13].
Complications
Diplopia is not an uncommon occurrence following orbital decompres-
sion with 15% to 63% of postoperative patients reporting new-onset diplo-
pia or worsening of preexisting symptoms [7,11,13–16]. This complication is
believed to be a result of a change in the vector of pull of the extraocular
muscles. Decompressive surgery rarely alleviates preexisting diplopia. Pa-
tients who have diplopia following decompressive surgery often require stra-
bismus surgery for correction. All patients should be informed of the
possibility of postoperative double vision, as well as the potential need for
further surgical intervention if this persists.
Several methods to decrease postoperative diplopia have been reported.
Multiple authors have described the preservation of a strut of inferomedial
bone between the decompressed floor and medial wall [12,17]. When this
strut is maintained, however, it is technically difficult to remove the orbital
floor through a purely endoscopic technique. The maintenance of a facial
sling in the region of the medial rectus has also been demonstrated to de-
crease the incidence of postoperative diplopia [10]. This technique provides
similar support as the medial strut technique, but allows for endoscopic
access to decompress the medial orbital floor. The concept of a balanced de-
compression (concurrent medial and lateral decompression) has also been
suggested as a means to decrease postoperative diplopia [14,18,19]. When
operating for compressive optic neuropathy, techniques designed to limit
diplopia may also limit the extent of decompression, and postoperative dip-
lopia is often accepted as a concession to improved visual acuity.
Postoperative bleeding following decompression is best managed through
endoscopic identification and direct cauterization of the bleeding site. Nasal
packing is generally not used to avoid pressure on the exposed orbital apex
and optic nerve. Postoperative infection is minimized through the use of
postoperative antibiotics with staphylococcal coverage. A large maxillary
antrostomy and limited bone removal in the frontal recess region minimize
the risk of developing postoperative sinusitis. Epiphora may develop if the
maxillary antrostomy is extended too far anteriorly with transection of
the nasolacrimal duct. This complication is treated with an endoscopic
ENDOSCOPIC ORBITAL AND OPTIC NERVE DECOMPRESSION 951
Surgical anatomy
The optic nerve may be divided into three segments: the intraorbital, the
intracanalicular, and the intracranial segment. Optic nerve decompression
aims to relieve compressive forces within the intracanalicular portion of
the nerve. The canal of the optic nerve is formed by the two struts of the
lesser wing of the sphenoid and carries both the optic nerve and the ophthal-
mic artery. At the orbital apex is the fibrous annulus of Zinn.
Pathophysiology
Compression on the optic nerve may result from neoplastic, inflamma-
tory, or traumatic processes. Initial theories of vascular compromise from
952 PLETCHER et al
external compression with resultant injury to the optic nerve have been
largely discarded. Manual compression resulting in conduction block and
focal demyelination from compression are the favored pathophysiologic ex-
planations for compressive optic neuropathy. Rapid recovery following de-
compression results from relief of the manual compression block while
a delayed recovery may occur over a period of weeks to months as a result
of remyelination [25].
Traumatic optic neuropathy may be divided into two categories: indirect
and direct. Direct TON results from a penetrating injury with projectiles or
other sharp objects. In these cases the intraorbital portion of the optic nerve
is generally injured and decompression is not recommended. Indirect TON
results from blunt trauma to the head. This injury can occur with or without
associated fractures of the orbital canal. The pathophysiology of visual loss
in these cases is frequently unclear, and may include intraneural edema, he-
matoma, shearing injury to the microvasculature or axons, altered cerebro-
spinal fluid circulation, and interruption of direct axonal transport. If there
is evidence of complete disruption of the optic nerve, decompression is not
indicated, as the optic nerve will not recover from such injury. In cases
where edema, hematoma, or moderate bony compression is suggested, how-
ever, decompression may be considered.
Technique
Patients are prepared for surgery in a similar manner to those undergoing
orbital decompression. A standard sphenoethmoidectomy is performed. The
sphenoid face is widely opened and the bulge of the optic canal is identified
along the lateral wall of the sphenoid sinus, superior to the carotid artery. In
some patients, the optic canal may be initially identified in a posterior eth-
moid or Onodi cell, which can be seen on preoperative CT scan [26]. Iden-
tification and opening of the Onodi cell is important to provide adequate
surgical exposure and allow full access to the optic canal.
Following exposure of the medial orbital wall, a spoon curette is used to
fracture the lamina papyracea approximately 1 cm anterior to the optic ca-
nal. The lamina is then carefully removed in a posterior direction to expose
the annulus of Zinn and the optic canal. Care must be taken to avoid pen-
etration of the periorbita, as subsequent herniation of orbital fat will ob-
scure the surgical field. As the optic canal is approached, the thin lamina
will be replaced with the thick bone of the lesser wing of the sphenoid.
This bone must be thinned before removal. A long-handled drill with a dia-
mond burr is used to methodically thin the medial wall of the optic canal
(Fig. 6). While drilling, care must be taken to prevent contact of the drill
bit with the prominence of the carotid artery, which is located just inferior
and posterior to the optic nerve. Care should also be taken to avoid excess
generation of heat while drilling this bone as thermal damage to the optic
nerve may result. After the bone is appropriately thinned, a microcurette
954 PLETCHER et al
Fig. 6. Endoscopic view of the right posterior nasal cavity following wide sphenoidotomy. The
lamina papyracea has been removed to reveal periobita near the orbital apex. A diamond burr is
used to thin the bone of the optic canal.
is used to fracture the thinned bone in a medial direction, away from the op-
tic nerve (Fig. 7). Bone fragments are then removed from the decompressed
nerve using blakesly forceps (Fig. 8), with resultant medial decompression of
the optic nerve (Fig. 9).
Controversy exists regarding the length of optic canal that should be de-
compressed as well as the necessity for decompression of the optic sheath.
With compressive optic neuropathy secondary to neoplasms, the extent of
decompression is dictated by the size and location of the neoplasm. For
Fig. 8. Blakesly forceps are used to remove bone fragments and expose the underlying optic
nerve.
Fig. 9. The decompressed optic nerve is visible along the lateral wall of the sphenoid sinus.
956 PLETCHER et al
Results
Most patients who undergo optic nerve decompression for compressive
nontraumatic optic neuropathy will have significant improvement in visual
acuity. An immediate improvement in visual acuity is frequently observed,
probably from relief of a mechanical conduction block. Further improve-
ment may occur over a period of weeks to months as remyelination of the
nerve leads to more efficient conduction [25].
The efficacy of optic nerve decompression in traumatic optic neuropathy
is unclear. Much of the difficulty in determining the success of surgical inter-
vention arises from the relatively high rate of spontaneous recovery from
TON. Thus, well-controlled studies with significant power are required to
delineate the efficacy of surgical intervention. Unfortunately, such studies
have not been possible due to the rarity of this condition.
In cases of nontraumatic, compressive optic neuropathy, however, the
natural course of disease is not one of spontaneous resolution. Thus, it is
reasonable to conclude that visual improvement following surgical decom-
pression is a direct result of the procedure. In our recent experience of endo-
scopic optic nerve decompressions for nontraumatic optic neuropathy,
improved visual acuity was noted following 8 of 10 decompressions. One pa-
tient required multiple decompressions due to progression of her fibrous
dysplasia with recurrent impingement upon the optic nerve [20].
Complications
The risk of CSF leak, meningitis, and visual loss with optic nerve decom-
pression appears to be higher than with standard endoscopic sinus surgery
or orbital decompression. Although several studies with 20 to 45 patients re-
port no complications, the IONTS and another recent study reported sev-
eral cases of CSF leak, some with associated meningitis and visual
decompensation [14,23]. In the IONTS, it is unclear whether the complica-
tions occurred in patients who underwent endoscopic decompression repre-
senting less than 40% of patients in their series, or an external approach.
Summary
With excellent visualization of the orbital apex and optic canal, the endo-
scopic transnasal approach is well suited for both orbital and optic nerve
decompression. This operation is an advanced endoscopic technique, and
should be performed only by surgeons experienced in endoscopic nasal
surgery. Although the indications and expected results for orbital
ENDOSCOPIC ORBITAL AND OPTIC NERVE DECOMPRESSION 957
References
[1] Kronlein R. Zur Pathologie und operativen Behandlung der Desmoid Cysten der Orbita.
Beitr Klin Chir 1889;4:149–63.
[2] Sewall E. Operative control of progressive exophthalmos. Arch Otolaryngol Head Neck
Surg 1936;24:621–4.
[3] Hirsch O. Surgical decompression of exophthalmos. Arch Otolaryngol Head Neck Surg
1950;51:325–31.
[4] Naffziger HC. Progressive exophthalmos. Ann R Coll Surg Engl 1954;15:1–24.
[5] Walsh TE, Ogura JH. Transantral orbital decompression for malignant exophthalmos.
Laryngoscope 1957;67:544–68.
[6] Kennedy DW, Goodstein ML, Miller NR, et al. Endoscopic transnasal orbital decompres-
sion. Arch Otolaryngol Head Neck Surg 1990;116:275–82.
[7] Michel O, Bresgen K, Russmann W, et al. Endoscopically-controlled endonasal orbital de-
compression in malignant exophthalmos. Laryngorhinootologie 1991;70:656–62.
[8] Gorman CA, Garrity JA, Fatourechi V, et al. A prospective, randomized, double-blind,
placebo-controlled study of orbital radiotherapy for Graves’ ophthalmopathy. Ophthalmol-
ogy 2001;108:1523–34.
[9] Mourits MP, van Kempen-Harteveld ML, Garcia MB, et al. Radiotherapy for Graves’ or-
bitopathy: randomised placebo-controlled study. Lancet 2000;355:1505–9.
[10] Metson R, Samaha M. Reduction of diplopia following endoscopic orbital decompression:
the orbital sling technique. Laryngoscope 2002;112:1753–7.
[11] Metson R, Shore JW, Gliklich RE, et al. Endoscopic orbital decompression under local an-
esthesia. Otolaryngol Head Neck Surg 1995;113:661–7.
[12] Schaefer SD, Soliemanzadeh P, Della Rocca DA, et al. Endoscopic and transconjunctival
orbital decompression for thyroid-related orbital apex compression. Laryngoscope 2003;
113:508–13.
[13] Metson R, Dallow RL, Shore JW. Endoscopic orbital decompression. Laryngoscope 1994;
104:950–7.
[14] Shepard KG, Levin PS, Terris DJ. Balanced orbital decompression for Graves’ ophthalm-
opathy. Laryngoscope 1998;108:1648–53.
[15] Wright ED, Davidson J, Codere F, et al. Endoscopic orbital decompression with preserva-
tion of an inferomedial bony strut: minimization of postoperative diplopia. J Otolaryngol
1999;28:252–6.
[16] Eloy P, Trussart C, Jouzdani E, et al. Transnasal endoscopic orbital decompression and
Graves’ ophtalmopathy. Acta Otorhinolaryngol Belg 2000;54:165–74.
[17] Goldberg RA, Shorr N, Cohen MS. The medical orbital strut in the prevention of postdecom-
pression dystopia in dysthyroid ophthalmopathy. Ophthal Plast Reconstr Surg 1992;8:32–4.
[18] Unal M, Leri F, Konuk O, et al. Balanced orbital decompression combined with fat removal
in Graves’ ophthalmopathy: do we really need to remove the third wall? Ophthal Plast Re-
constr Surg 2003;19:112–8.
[19] Graham SM, Brown CL, Carter KD, et al. Medial and lateral orbital wall surgery for bal-
anced decompression in thyroid eye disease. Laryngoscope 2003;113:1206–9.
[20] Pletcher SD, Metson R. Endoscopic optic nerve decompression for non-traumatic optic neu-
ropathy. American Rhinologic Society combined otolaryngologic spring meeting, Chicago,
Il, 2006.
[21] Cook MW, Levin LA, Joseph MP, et al. Traumatic optic neuropathy. A meta-analysis. Arch
Otolaryngol Head Neck Surg 1996;122:389–92.
[22] Kountakis SE, Maillard AA, El-Harazi SM, et al. Endoscopic optic nerve decompression for
traumatic blindness. Otolaryngol Head Neck Surg 2000;123:34–7.
958 PLETCHER et al
[23] Rajiniganth MG, Gupta AK, Gupta A, et al. Traumatic optic neuropathy: visual outcome
following combined therapy protocol. Arch Otolaryngol Head Neck Surg 2003;129:1203–6.
[24] Li KK, Teknos TN, Lai A, et al. Traumatic optic neuropathy: result in 45 consecutive sur-
gically treated patients. Otolaryngol Head Neck Surg 1999;120:5–11.
[25] McDonald WI. The symptomatology of tumours of the anterior visual pathways. Can
J Neurol Sci 1982;9:381–90.
[26] Allmond L, Murr AH. Clinical problem solving: radiology. Radiology quiz case 1: opacified
Onodi cell. Arch Otolaryngol Head Neck Surg 2002;128:596, 598–599.
[27] Luxenberger W, Stammberger H, Jebeles JA, et al. Endoscopic optic nerve decompression:
the Graz experience. Laryngoscope 1998;108:873–82.