Orbital Decompression for Graves Disease

Updated: Oct 19, 2018
  • Author: Michael Mercandetti, MD, MBA, FACS; Chief Editor: Arlen D Meyers, MD, MBA  more...
  • Print



Graves disease, originally called Graves-Basedow disease, was first described as the triad of hyperthyroidism, goiter, and exophthalmos in 1835. It is relatively common; it may occur in people of any age but is most common during the third to fifth decades of life. The female-to-male ratio varies from 4:1 [1] to 2.5–6:1 [2] to 8:1. The distribution is bimodal, with the peak incidence in the fifth and seventh decades of life. [3] Men tend to develop more severe orbitopathy. [4, 5]

Severe ophthalmopathy is an uncommon but problematic manifestation of Graves disease. Only 5–6% of patients with Graves disease develop problems severe enough to warrant surgical decompression on a functional basis. This does not include patients who seek cosmetic decompression.

Approximately half of all patients with Graves hyperthyroidism develop ophthalmopathy. In its most involved forms, the ophthalmopathy can result in severe corneal problems that necessitate decompression. Optic neuropathy due to compression of the optic nerve is another indication for decompression.

Orbitopathy associated with Graves disease may severely compromise a patient’s vision. The condition may cause diplopia, decreased ocular motility, exposure keratitis, optic neuropathy, and poor cosmesis. Surgical management is both an alternative and adjunctive treatment to medical therapy, which most often involves corticosteroids, external beam radiotherapy, or both.

Various approaches have been described since surgical decompression of the orbit for thyroid ophthalmopathy was initially advocated in 1911. [6, 7, 8, 9] Orbital decompression most often involves the removal of the bones that comprise the orbit. Decompression of a lesser nature can be accomplished with the removal of extraconal and intraconal fat. This fat decompression can be combined with bone removal for more extensive decompression.

The advent of advanced endoscopic techniques has enabled surgeons to decompress the orbit endoscopically, allowing effective decompression with less morbidity. This is primarily a result of circumventing a gingival incision and a low incidence of cranial nerve V2 hypesthesia. Although an endoscopic approach to the medial and inferior walls can be performed in isolation, a balanced approach, incorporating a lateral decompression with repositioning of the lower lid, can be required. [10, 11]

See also the following:

Pathophysiology and etiology of Graves ophthalmopathy

Orbital findings result from an increase in the volume of orbital tissues secondary to inflammation, edema, and congestion. Most notable is the enlargement of the extraocular muscles secondary to infiltration with inflammatory cells, deposition of immune complexes, and an increase in glycosaminoglycans, particularly hyaluronic acid, which is hydrophilic. In some cases, this is ultimately followed by fibrosis. These alterations are modulated by circulating antibodies. Smoking has been shown to increase venous congestion in the orbit via reduction of flow in the superior ophthalmic vein separate from extraocular muscle involvment, which were comparable between smokers and non smokers. [5]

These autoimmune dysfunctions are considered to be mediated by both humoral and cellular dysfunctions; however, this is debated. Regardless, the processes show a predilection for the orbital tissues, the extraocular muscles, and periorbital structures. Both orbitopathy and thyroid dysfunction are more common in family members of patients who have thyroid-related orbitopathy.

Severe cases of ophthalmopathy may be associated with lid edema, chemosis (edema of the conjunctiva), and diminished ocular motility. Eventual sequelae may include corneal exposure with subsequent ulceration, diplopia due to extraocular muscle restriction, or fibrosis and optic nerve compression with resulting visual field deficits, including blindness.

Patients with Graves orbitopathy usually have hyperthyroidism but can have euthyroidism or hypothyroidism. A patient may present with signs of orbitopathy before being diagnosed with a thyroid disorder. At some point, 90% of patients with orbitopathy have hyperthyroidism. More than 50% of patients develop orbitopathy after they have had hyperthyroidism. The degree of control of the thyroid disorder does not exactly correlate with the extent of the orbitopathy.

The orbit’s inflammatory response to the thyroid dysfunction consists of an increase in both B and T lymphocytes, accompanied by edema. These processes can cause scarring. In addition, fibroblastic activity with the resultant production of collagen and glycosaminoglycans also increases. Increased edema and increases in the thickness of the extraocular muscles and fat cells are also found. Some early fibroblasts convert into fat cells.

If a patient has hyperthyroidism, cigarette smoking increases the incidence and severity of thyroid-related orbitopathy. The antigens that specifically cause this are unclear.

Clinical presentation of Graves ophthalmopathy

Graves-related exophthalmos usually occurs in adult life. The condition most commonly affects middle-aged women. It is more severe in older men, usually with greater exophthalmos. Graves orbitopathy is the most common cause of unilateral and bilateral proptosis. Most commonly it occurs in patients who are hyperthyroid, but also occurs in patients with hypothyroidism, euthyroidism, and Hashimoto's thyroiditis. [12]

Ethnic variations in the anatomy of the orbits can be found; therefore, the degree of proptosis with its concomitant problems can vary greatly, regardless of race or gender. A person with a tighter orbit is more at risk for compressive optic neuropathy from the disease process and may not show as much proptosis as a person with a much less tight orbit who is more proptotic but does not as readily experience compressive optic neuropathy.

Retraction of the upper or lower lid (Dalrymple sign) may develop. This is due to hyperstimulation of the sympathetically innervated Mueller muscle in the upper lid and its analog, the inferior tarsal muscle, in the lower lid. When the patient tries to look downward, the upper lid can hold back and lag behind the movement of the globe (von Graefe sign; also termed lid lag).

Unilateral or bilateral proptosis (34–93% of patients), tearing (epiphora), chemosis (edema of the conjunctiva), and hyperemia or injection of the conjunctiva (an increase in the size of the blood vessels) may develop.

A prominence of the blood vessels that overlie the insertion of the rectus muscles may develop. Edema of the lids and of the malar pad areas may develop. Corneal problems may vary from dryness to perforation when a full-thickness opening in the cornea is present.

Patients can present in various stages. Patients with very mild cases may have low-grade orbital inflammation with some orbital discomfort, tearing, and chemosis. Other patients may have mild proptosis with or without lid retraction. Still other patients present with marked inflammation, tremendous discomfort, corneal problems, or optic nerve compression. Patients may have diplopia.

In burnt-out Graves orbitopathy, many of the same signs persist, except that the inflammation is gone and the eyes and orbits appear quiet. However, the proptosis, lid retraction, and diplopia still remain.

Patients with hyperthyroidism can present with symptoms and signs of hyperstimulation of their sympathetic nervous system. These symptoms and signs include the following: weight loss, sweating, tremulousness, edginess, heart palpitations, an increase in appetite, feeling warm, and feeling "revved up."

Nonsurgical treatment of Graves ophthalmopathy

Severe orbital manifestations of Graves disease early in the disease course often respond to high-dose corticosteroid therapy. If this is unsuccessful, external beam radiation therapy at a dose of 20 Gy may be considered. [13] When orbital findings have been present for a long time, they are less likely to respond to medical management because of fibrosis of the involved tissues.

Steroid treatment combined with radiation treatment has also proven effective. [14]

In patients for whom nonsurgical modalities are unsuccessful or those who are not considered candidates, surgical decompression may be considered. Medical treatment can also be used in conjunction with surgical decompression.

Newer treatment modalities include the use of biologics. In one trial, rituximab showed benefits compared to steroids, and in another trial, it showed no benefit against placebo. [15, 16, 6, 17] One trial reported increased adverse events with the use of rituximab compared to saline. [18]  Compared to adalimumab, tocilizumab, and etanercept, rituximab reduced B cells that are CD20 positive. [19]  However, rituximab may have a role in steroid-resistant cases. [20] Teprotumumab, a monoclonal antibody that is an IGF-1 inhibitor, is showing promise in offsetting inflammation in Graves orbitopathy and decreasing proptosis. [21, 22]

Surgical options

Various surgical approaches can be used for decompression. Otolaryngologists commonly perform decompression via a transantral approach to the medial and inferior orbital walls. An endoscopic approach to the medial and inferior walls is currently used. Oculoplastic specialists often use a transcutaneous or transconjunctival lower-lid approach. Often, the lower lid must be separated from its periosteal attachment and subsequently repaired. Transcaruncular approaches can also be used to access the medial orbital wall.

The lateral wall can be surgically approached in different ways. A team approach that involves the services of an otolaryngologist and an oculoplastic specialist is common. If the decompression requires the removal of the frontal bone, the services of a neurosurgeon may be required.

Ultimately, the approach or approaches used are tailored to the severity of the problem and therefore the degree of decompression desired and the anatomy of the patient. These decisions can be influenced by the cosmesis achieved, especially in patients who are undergoing decompressions to improve their appearance. A balanced approach in which the inferior wall and medial wall decompression are combined with lateral decompression may be used.

Medial wall decompression with preservation of the medial strut between the ethmoid cavity and the inferior wall, balanced with lateral wall decompression and with or without inferior wall decompression, may provide effective reduction of exophthalmos without a high risk for new-onset postoperative diplopia, especially in patients who are having surgery for cosmetic indications.

In the authors’ view, this combined approach to surgery for patients with Graves ophthalmopathy, incorporating both an otolaryngologist with endoscopic experience and an oculoplastic surgeon, provides for ideal surgical management.

Anatomic Considerations

The orbit, which protects, supports, and maximizes the function of the eye, is shaped like a quadrilateral pyramid, with its base in plane with the orbital rim. Seven bones conjoin to form the orbital structure, as shown in the image below.

This image of the right orbit shows the 7 bones th This image of the right orbit shows the 7 bones that contribute to its structure.

The orbital process of the frontal bone and the lesser wing of the sphenoid form the orbital roof. The orbital plate of the maxilla joins the orbital plate of the zygoma and the orbital plate of the palatine bones to form the floor. Medially, the orbital wall consists of the frontal process of the maxilla, the lacrimal bone, the sphenoid, and the thin lamina papyracea of the ethmoid. The lateral wall is formed by the lesser and greater wings of the sphenoid and the zygoma.

For the transcutaneous or transconjunctival lower-lid approach, the inferior oblique muscle must be avoided. In fact, with either approach, during a bony decompression, the orbital space is not entered until the bone has been removed. Opening the periorbita earlier can allow the fat to prolapse, interfering with the dissection. The infraorbital nerve is to be avoided during any dissection to avoid postoperative paresthesia.

Medially, the dissection and bone removal is accomplished but stays below the ethmoidal arteries to avoid compromising the anterior or posterior ethmoidal arteries (and thereby causing hemorrhages) and to avoid disrupting the cribriform plate.

Posteromedially, care must be exercised when the sphenoid sinus is entered. If the carotid artery is violated, the result can be catastrophic.

Additionally, in more posterior dissections, damaging or resecting the optic nerve must be avoided during the dissection.

The relevant anatomy for endoscopic orbital decompression is similar to that for endoscopic ethmoidectomy and maxillary antrostomy. Specifically, the middle turbinate insertion at the skull base represents the medial limit of dissection, the fovea ethmoidalis is the superior limit, and the lamina papyracea is the lateral limit. The lamina extends from the nasolacrimal system anteriorly to the annulus of Zinn posteriorly. Cranial nerve V2 runs in the maxillary roof and can be dehiscent. For more information about the relevant anatomy, see Orbit Anatomy.

In all cases, the globe must not be unduly pressed on. The amount and duration of pressure must be monitored to prevent inadvertently compression of the ophthalmic artery.


Historically, the indications for surgical decompression of the orbit have included exophthalmos accompanied by corneal exposure and disfigurement and increased orbital pressure produced by swelling of extraocular muscles, which can lead to compressive optic neuropathy and visual loss. Recent advances in the techniques for orbital decompression have decreased the morbidity of the procedure, and a need for cosmetic decompressions is also an indication.

Emergency decompression can be warranted in the most severe cases of compression and visual loss. However, this situation is usually concomitantly treated with immunosuppressive agents, most commonly systemic steroids. Radiation therapy can be used in the acute setting in addition to immunosuppressives; it may also be used for patients who are not surgical candidates.


Patients who are unable to undergo a surgical procedure or who are unwilling to accept the potential complications of surgery are not candidates for orbital decompression. Other relative contraindications include a history of chronic sinusitis, immunocompromise, bleeding disorders, and atretic sinuses.



Preoperative Evaluation

All patients undergo a complete head and neck examination. A full ophthalmologic workup should be obtained if not already completed. Even in cosmetic candidates, subtle abnormalities may surface only after the procedure. Visual acuity, visual fields, conjunctival and corneal appearance, Hertel measurements, extraocular motility, and symptoms of diplopia are recorded. Obtain preoperative and postoperative photography.

A thyroid function test must be performed in patients with Graves disease to determine if they are euthyroid before the initiation of general anesthesia. Other than that, routine laboratory studies are sufficient for preoperative clearance. Patients in whom a thyroid disorder has not been diagnosed should undergo appropriate testing of thyroid function. Bleeding may occur during surgery; however, patients are not routinely typed and crossed.

Although patients with Graves orbitopathy have less than a 1% chance of having myasthenia gravis, 5% of patients with myasthenia gravis have hyperthyroidism. Therefore, an appropriate workup, including acetylcholine receptor antibody testing, might be warranted in selected cases. If the clinical suspicion of myasthenia gravis is high and the antibody testing results are negative, further testing might be indicated. This might include Tensilon testing and/or single nerve fiber stimulation studies (single fiber electromyography [SFEMG]).

The complete head and neck examination and ophthalmologic evaluation should be followed by preoperative computed tomography (CT) scanning (coronal and axial preferred) of the orbits and paranasal sinuses. Magnetic resonance imaging (MRI) is sufficient, but the bony landmarks are visualized more easily on CT imaging. Chest CT might be indicated in patients who also have myasthenia gravis to evaluate for thymoma.

Preoperative diagnostic endoscopy is appropriate to ensure that no significant obstruction is present (eg, septal deviation, concha bullosa) and to ensure the absence of abnormalities such as nasal polyposis.


Prepare the patient for endoscopic sinus surgery under monitored anesthesia care or general anesthesia (the authors’ preferred approach). After the administration of parenteral antibiotics and intravenous steroids (eg, dexamethasone, 8–12 mg) and topical (oxymetazoline 0.05%) and local (1% lidocaine, 1:100,000 epinephrine) anesthesia, vasoconstriction is achieved, with particular attention to the middle turbinate and the lateral nasal wall.



The technique for endoscopic medial wall decompression, originally described by Kennedy et al [23] and then Metson et al, [24] is summarized below. A brief description of the lateral orbital wall decompression, as described by Goldberg et al [25] and other groups, is also included.

Medial Wall Decompression

Endoscopic approach

Perform a septoplasty first if it is needed for exposure. Next, perform a total anterior and posterior endoscopic ethmoidectomy with sphenoidotomy (to ensure the most posterior segment of the accessible medial orbital wall has been identified) and extended maxillary antrostomy (to provide inferomedial decompression and as prophylaxis against maxillary sinusitis secondary to ostial narrowing as a result of postoperative edema or scarring).

Use a Freer elevator or a similar instrument to fracture the medial orbital wall (lamina papyracea). Then gently remove the lamina, taking care not to violate the periorbita, because early release of orbital fat has a tendency to obscure the view. This task may be tedious and requires careful use of several instruments, including a Freer elevator, a nerve hook for anterior bone removal, curettes, and Wilde-Blakesley or similar forceps.

In cosmetic cases, try to preserve a strut of bone between the medial and inferior wall decompression to minimize the risk of diplopia. If inferior decompression is to be performed, use a transconjunctival approach.

When the bone of the medial wall has been completely removed (and usually before the lateral wall is addressed), incise the periorbita to liberate the orbital fat and allow decompression. Make 2-4 horizontal periorbital incisions, starting posteriorly and extending anteriorly.

Limiting the depth of the incision is critical to prevent injury to the rectus muscles (especially the medial rectus) and to minimize bleeding. This is accomplished by using a No. 12 blade on a No. 7 Bard-Parker handle. The blade is protected with a sterile bandage (Steri-Strip), except for the distal 2–3 mm. This technique ensures a sharp cutting instrument each time.

Beginning with the inferior incision and sequentially moving to the most superior incision is best. This sequence minimizes the propensity of the orbital fat to obscure the subsequent incisions. For ease of incision and to assist with decompression of the orbital contents into the ethmoid vault, which may require additional effort in an irradiated orbit, apply gentle pressure to the globe during this maneuver. Finally, thin strands of periorbita may persist as bands between the horizontal incisions and should be carefully teased free with a nerve hook or similar instrument.

Transcutaneous approach

Alternative approaches to the medial wall include a transcutaneous incision (such as those used in an external ethmoidectomy approach and a Lynch-type incision). This incision can be enlarged inferiorly with removal of the medial canthal tendon. The floor of the orbit can also be accessed with this approach. The trochlea, through which the superior oblique tendon passes, should be avoided.

Transcaruncular approach

Still another approach is the transcaruncular approach. [1, 26] This approach does not require a cutaneous incision, nor does it require detachment of the medical canthal tendon. In this technique, the entry incision is made medially and posteriorly to the caruncle. This plane provides access to the posterior lacrimal crest and the more posterior ethmoids.

The medial wall can also be approached with a transcutaneous or transconjunctival approach by continuing the dissection superomedially. Take care to avoid the lacrimal system and the origin of the inferior oblique muscle.

Lateral wall decompression

The lateral wall of the orbit may be approached through an upper eyelid incision, lateral incision, lateral canthotomy incision, or vertical incision through the conjunctiva; it may also be approached beneath a coronal flap. An eyelid crease incision may be preferred because it heals well, does not violate the conjunctiva or lateral canthal angle, and allows excellent and rapid access to the orbit.

Carry the eyelid crease incision along a laugh line over the orbital rim, never behind the eyebrow. Use a cutting cautery or a laser incisional device to cut through the orbicularis muscle. Leave the orbital septum intact. Laterally, proceed with dissection onto the periosteum of the lateral orbital wall. Using retractors or traction sutures, expose the entire orbital rim. Score the periosteum at the orbital rim and proceed with dissection in the subperiosteal space. Control bleeding from the zygomaticotemporal, zygomaticofacial, or lacrimal foramina with bone wax or cautery.

Continue the dissection to the inferior orbital fissure inferiorly, the frontosphenoidal and frontozygomatic sutures superiorly, and past the zygomaticosphenoidal suture. Use a high-speed drill to break through the thin anterior wall of the lateral orbit, and remove the anterior bone with a drill or rongeurs, exposing the temporalis muscle. Remove the bone anteriorly, leaving only a thin bony rim.

Posteriorly, extend the dissection into the sphenoid bone, stopping at diploic bleeding or when the dura overlying the temporal lobe is exposed. Use a guarded sickle blade to score the periorbita with 2 vertical strips above and below the level of the lateral rectus muscle. A Freer elevator or forceps may be used to tease out orbital fat.

Orbital floor decompression

The orbital floor can be approached in various ways. A direct inferior approach can be via a Caldwell-Luc incision, which is performed transorally in the Ogura approach. [1] Alternatively, the maxillary sinus can be entered endoscopically and the bone of the orbital floor removed. Avoid damaging the infraorbital nerve.

Nonendoscopic alternatives would consist of a transconjunctival lower eyelid approach or a transcutaneous lower eyelid approach, with or without release of the lateral canthal tendon.

Orbital roof decompression

Orbital roof decompression is reserved for situations in which other approaches have been tried or an extensive amount of decompression is required. The roof can often be thinned out while avoiding the lacrimal gland. The roof is often accessed via a coronal approach.

Other techniques

In fat decompression, a standard transcutaneous or transconjunctival approach is used. The extraconal orbital fat is removed with meticulous hemostasis. Accessing intraconal fat involves a deeper dissection. [27] Li et al reported satisfactory results with fat-removal orbital decompression, which they termed FROD. [28]

The lateral orbital wall can be advanced to provide additional decompression. The anterior lateral wall can be resected, advanced, and plated into its new forward position. Valgus rotations, in which the anterior lateral wall is left intact and the posterior part is removed, can be performed.

The removal of the bony walls of the orbit can also be accompanied with the placement and fixation of an inferior orbital rim implant made of porous polyethylene accompanied by midface elevation, which in a small study decreased the average Hertel exophthalmometry reading by 5.4 mm. [29]



Postoperative care

Nasal packing is not used, and the patient is instructed to avoid nose blowing for at least 2 weeks after surgery.

Carefully evaluate patients in the early postoperative period to ensure maintenance of preoperative level of vision. Examine patients every 1-2 weeks after surgery to ensure proper wound healing. Then, examine them approximately 3–6 months after surgery, when the final result should be attained; postoperative photography may be done if desired.

Expected outcomes

Endoscopically assisted orbital decompression was first described by Kennedy et al in 1990 (13 orbits), who suggested that it was both safe and efficacious when performed by otolaryngologists trained in endoscopic techniques. [30] In 1995, Metson et al reported on 29 cases of endoscopic orbital decompression performed with local anesthesia, confirming the benefit of this less invasive method for medial wall decompression. [31] They advocate operating on a single orbit at a time when bilateral decompression is necessary.

A more recently introduced philosophy is the concept of balanced decompression. In 1995, Goldberg et al suggested that decompression should be performed both medially and laterally to achieve a symmetric and balanced anatomic result. Currently, the method for lateral decompression extends posteriorly into the sphenoid bone, giving a greater decompression than removal of the anterior wall alone.

The authors favor the concept of balanced decompression with the newer endoscopic approach and with an effort to provide a customized surgical management dictated in part by the indications for surgery.

Therefore, in this series, among patients who underwent decompression for cosmesis via an endoscopic approach, the incidence of new-onset postoperative diplopia was zero. This is not always the case. This was attributed, in part, to the balanced approach to decompression (ie, both medial and lateral walls are decompressed, allowing a balanced release of orbital tissues) and preservation of the medial strut between the ethmoid cavity and the inferior orbital decompression.

Goldberg et al have reported that preservation of this strut, in addition to balancing the decompression, maintains globe position, minimizing the risk of postoperative diplopia.

Shepard et al collected retrospective data from a 3-year study period (January 1994 to December 1996) at an academic practice. [32] The surgical procedures were customized to address the indications (ie, visual changes, corneal exposure, diplopia, disfigurement). In cases of optic nerve compression, care was taken to extend the surgical decompression as far posteriorly as possible. All endoscopic decompressions were performed by a board-certified otolaryngologist; lateral wall decompressions were performed by an oculoplastic ophthalmologist.

Seven women and 4 men underwent surgical management for Graves ophthalmopathy; a total of 18 orbits were decompressed. Medial wall and lateral decompressions were performed in all cases. Inferior wall decompression was accomplished in 11 orbits. Six additional procedures were performed (2 septoplasties, 2 turbinate trims, 2 orbital rim implants).

Five patients (7 orbits) underwent surgical decompression for visual changes (3 afferent pupillary defects, 2 cases of color vision loss). All 5 patients had been treated with oral steroids, and 2 patients had received previous orbital radiation.

The mean improvement in exophthalmos was 4.5 mm (by Hertel measurement). All had improved vision postoperatively. The incidence of newly onset postoperative diplopia was 40% (2 of 5). Three patients had persistent preoperative diplopia. Both patients with iatrogenic diplopia underwent successful strabismus surgery and treatment with prism glasses. No other complications were noted.

The remaining 6 patients (11 orbits) underwent decompression for cosmetic improvement and/or exposure keratitis. One patient had undergone previous radiation for Hodgkin disease, but no history of orbital radiation was present in this group. The mean decompression for this group was 4.7 mm as determined by Hertel measurement. Because surgical procedures were performed for cosmesis rather than threatened vision in this group, the medial strut between the ethmoid cavity and the orbital floor was preserved.

None of the patients who underwent cosmetic decompression developed new postoperative diplopia. One patient developed mild postoperative sinusitis that responded to antibiotics and decongestants.

No other complications were noted. Significant cosmetic improvement was achieved in this group of patients (see the first image below). Preoperative and postoperative computed tomography (CT) scans demonstrated expansion of the bony orbit (see the second image below). A second patient who also had significant exposure keratitis required orbital rim implants in addition to decompression (see the third image below).

Frontal views of patient, taken (A) preoperatively Frontal views of patient, taken (A) preoperatively and (B) 6 months postoperatively.
Preoperative coronal noncontrast bone window compu Preoperative coronal noncontrast bone window computed tomography (CT) scans (A); patient had incidental evidence of chronic sinusitis, which improved radiologically. Postoperative coronal noncontrast bone window CT scans (B) demonstrate typical decompression.
Frontal view of patient with exposure keratitis pr Frontal view of patient with exposure keratitis preoperatively (A). Lateral view of patient with exposure keratitis preoperatively (B). Frontal view of patient 4 months postoperatively (C). Lateral view of patient 4 months postoperatively (D).

A more recent study of the combined endoscopic transnasal (medial) and transconjunctival (inferior/lateral) approach reported similar results. Postoperatively, no patients involved in the study reported new diplopia after the procedure, whereas two patients reported worsening of their existing diplopia (3.8%). Mean reduction in proptosis across all patients was 3.2 mm. [33]

Concurrent decompression surgery and Muller muscle recession was reported by Ben Simon et al, with reasonable results. [34] However, in a subsequent article, Goldberg (who was 1 of the authors of the Ben Simon article), delineated a more traditional staged approach that progressed from decompression to muscle surgery, lid surgery, and then soft tissue procedures. [25]  Other authors have advocated combined surgery of decompression with eyelid surgery. [35]

The timing of decompression surgery is still debated, especially if the optic nerve is not acutely compromised. Baldeschi et al reported that early decompression for rehabilitative surgery, compared with later surgery, had a higher risk of causing diplopia and did not improve outcomes. [36] On the basis of a retrospective review over a 10-year period, Baldeschi et al also reported that radiation treatment did not adversely affect decompression results. [37] The group they studied had surgery for aesthetic reasons and no diplopia preoperatively.


Complications of traditional methods of orbital decompression vary with the approach. Complications include diplopia, blindness, epiphora, brain injury, cerebral spinal fluid (CSF) leak, oral-antral fistula, nasolacrimal duct obstruction, and scarring. The endoscopic approach includes all of those risks except for oral-antral fistula.

Optic nerve injury has been reported during orbital decompression for Graves disease, but it is rare and may be less likely with endoscopic control, because of the improved visualization. Complications can include the new onset of postoperative diplopia among patients who undergo decompression for visual loss.

Goh and McNab reported a 30% increase in postdecompression diplopia over an 11-year period. [38]

Baldeschi et al reported apparent reactivation of Graves orbitopathy after orbital decompression. [39] .

Bodoridis and Bunce in their review reported that strong randomized data were lacking for the best decompression approach to use, but the ystated that based on the available data, two-wall decompression, the medial wall and lateral wall, with or without fat removal, was considered the most effective. [40]

Newer treatment modalities are being developed, and their efficacy or lack thereof needs to be determined. [41, 42]