An Orbital Decompression for Severe Dysthyroid Exophthalmos

An Orbital Decompression for Severe Dysthyroid Exophthalmos JOHN S. KENNERDELL, MD, JOSEPH C. MAROON, MD

Abstract: A procedure has been developed for maximizing decompression in patients with nonspecific orbital inflammatory disease or dysthyroid orbitopathy that involves removal of parts of all four orbital walls through a lateral orbitotomy using a 30-mm incision combined with a canthotomy and an inferior cul-de-sac incision. The procedure has been performed on seven orbits of four patients who had greater than 30 mm of forward placement of the eyes as measured by Hertel exophthalmometry and/or a 10-mm difference in the forward placement of the two eyes, indications considered for this procedure . The procedure has resulted in 10-17 mm of reduction in proptosis in these patients, which is consistently greater than that obtained by any other procedure. A complication of spinal fluid leak resulted in the recommendation that the procedure not be performed in patients over 65 whose dura is considered too thin to prevent spontaneous leakage and minimal surgical trauma. In all circumstances, this procedure must be performed by an experienced neuro-ophthalmic orbital surgical team . [Key words: corneal exposure, dysthyroid exophthalmos, dysthyroid optic neuropathy, four-walled orbital decompression, inferior canthotomy, lateral orbitotomy.] Ophthalmology 89:467-472, 1982

Dysthyroid exophthalmos requiring orbital decompression occurs in a small number of patients with this disorder. The authors believe orbital decompression is indicated for corneal exposure, secondary to severe exophthalmos; apical orbital compressive optic neuropathy that fails to reverse with high dose corticosteroids and/or radiation therapy ; and severe exophthalmos with significant cosmetic disfigurement, resulting in psychologic decompensation. Several surgical techniques have been advocated for orbital decompression. In 1889 Kronlein described a lateral orbitotomy procedure for removal of orbital tumors and later advocated its use for orbital decompression/ and Dollinger revised this technique for From the Eye and Ear Hospital . Pittsburgh . Pennsylvania. Presented at the Eighty-sixth Annual Meeting of the American Academy of Ophthalmology . Atlanta . Georgia . November 1-6. 1981 Reprint requests to John S. Kennerdell . MD. Eye and Ear Hospital, 230 Lothrop Street. Pittsburgh. PA 15213. 0161-6420/8210500/0467/$00.80

© American Academy of Ophthalmology

palliation of the optic neuropathy associated with dysthyroid orbitopathy in 1911. 2 In 1930 Hirsch and Urbanik described a Caldwell-Luc approach to removal of the orbital floor for dysthyroid exophthalmos. 3 The following year Naffziger used a technique for removing the orbital roof through a craniotomy and later included part of the superior lateral orbital wall with satisfactory results. 4 In 1936, Sewell advocated a fronto-ethmoidectomy from an anterior approach to decompress the orbit5 and in 1950, Hirsch recommended a trans antral removal of the orbital floor.6 The most widely used procedure today is that developed in 1957 by Ogura and Walsh, who have had success with a combined transantral excision of the bony floor and medial wall of the orbit. 7- 9 In 1978, McCord and Moses advocated an anterior approach to expose the floor and medial wall from a fornix-based conjunctival incision combined with an inferior lateral inferior canthotomy. This also allowed them to add removal of part of the lateral wall. tO In 1979, Cooper and Trokel described a three-wall orbital decompression through a Kronlein lateral orbitotomy as modified by Berke. l l 467

OPHTHALMOLOGY • MAY 1982 • VOLUME 89 • NUMBER 5

removal 0:1 .:Jeet/on of/otero!

orhitolrim

Fig 1. Artist's step-wise illustration of important steps in the four-wall orbital decompression.

These procedures have had varying success in orbital decompression. The reduction in exophthalmos increases usually with the amount of orbital bony wall removed and opening of the periorbita, allowing the fat to prolapse. For those reasons, when patients with marked proptosis, that being defined as over 30 mm by exophthalmometry in one or both eyes or over a

468

lO-mm difference between the forward placement of the eyes, are encountered, the procedures that had been described are inadequate for the amount of decompression desired in these unusual circumstances. Therefore, a decompression procedure was developed that allows for removal of a portion of all four bony walls combined with several periorbital incisions to

KENNERDELL AND MAROON • ORBITAL DECOMPRESSION

HERTEL MEASUREMENTS IN 7 DECOMPRESSED ORBITS

1. G.D. 2.

D.O.

3.

B.B.

4.

F.D.

Pre-op

Post-op

0.D.38 O.S. 38 O.D. 31 O.S. 18

0.D.26 O.S. 23 O.D. 21 O.S. 18

Va O.D. 38 (20/80) O.S. 38 (20/70) 0.D.32 O.S. 32

O.D. O.S. O.D. O.S.

Va 24 (20/40) 22 (20 / 30) 22 21

Fig 2. Pre- and post-operative Hertel measurements in these four patients.

allow fatty prolapse and has consistently obtained greater decompressions than by other less radical procedures. The added removal of the orbital roof dictates that the procedure must be done by an experienced neurosurgicaVorbital surgical team.

PROCEDURE The patient is operated under general anesthesia in a supine position with the head turned to the opposite side. The subcutaneous tissue and temporalis muscle over the lateral wall of the orbit is infiltrated with 10 cc of 11200,000 epinephrine in 1% Xylocaine solution to promote hemostasis. The side of the head is prepared with Betadine soap and tincture of Betadine and appropriately draped. There is no need to remove any of

the hair from the side of the head. A straight line, 30-mm incision is made extending from the lateral canthus toward the top of the ear. The incision is sharply carried down to the temporalis fascia and the periosteum of the zygoma. The subcutaneous tissue is undermined over the periosteum of the lateral zygoma to the level of the superior and inferior orbital rims. The periosteum of the zygoma is incised along the lateral orbital rim using a cutting cautery with a linear incision extended back into the temporalis muscle in the line of the skin incision, creating a T-shaped opening in the periosteum. A periosteal elevator raises the periosteum from the zygoma and the temporalis muscle belly from the posterior aspect of the zygoma. The periorbita is then elevated in a similar fashion, and a Stryker saw is used to make bony cuts through the zygoma angled posteriorly through the superior and inferior orbital margins. The lateral orbital rim is then broken away with a large rongeur. The posterior lateral orbital wall is removed with a rongeur, and the sphenoid wing is partially removed with a Hall drill down to the level of the dura ofthe temporal lobe. The superior periorbita is then dissected free of the roof of the orbit with a periosteal elevator and the lateral posterior roof is removed with rongeurs. The CT scan is in the operating room so that the sinuses can be viewed, and the superior decompression is not carried into the frontal sinuses. Next, an inferior canthotomy is done, and an incision is carried through the inferior conjunctival culde-sac down to the inferior orbital rim, medially as far as the lacrimal punctum. The inferior periorbita is then elevated and the inferior orbital neurovascular bundle is identified. A Hall drill is used to make a small opening in the orbital floor, and then rongeurs remove parts

Figs 3A-C. A, top left, preoperative appearance. B, top right, early postoperative appearance after left decompression. C, left, postoperative appearance following bilateral decompression.

469

OPHTHALMOLOGY. MAY 1982 • VOLUME 89 • NUMBER 5

Figs 4A-C. A, top left, preoperative appearance with unilateral exophthalmos. B, top right , postoperative appearance following decompression . C, left, close-up appearance of postoperative skin incision.

of the orbital floor lateral and medial to the neurovascular bundle and as far posteriorly as deemed safe. Retraction is used to elevate the globe and the medial orbital wall is removed posteriorly to the level of which visibility allows, and the posterior inferior lamina papyracea is manually broken inward toward the midline by the index finger of the surgeon (Fig 1). This completes the removal of parts of all four bony walls. The periorbita is incised between the extraocular muscles allowing the fat to protrude. The inferior conjunctiva is closed with a running 6-0 Dexon® suture. The canthi are approximated with 4-0 Dexon, which is also used to close the lateral periorbita to the periosteum of the zygoma, over the replaced lateral orbital rim, to close the temporalis fascia, and to approximate the wound edges. The wound is closed with a running 6-0 mild chromic catgut suture. Steri-strips are placed over the wound to add reinforcement. This combination eliminates the need for suture removal. Maxitrol ointment is applied to the eye and to the wound, and the eye is covered with a soft dressing that provides an evenly distributed pressure over the orbit and skin incision. The bandage is removed the following day and the patient is discharged on the third postoperative day. 470

Ten milligrams of Decadron IV is given during the procedure. After operation 20 mg of prednisone is given four times per day PO for three days, this is then reduced to 10 mg four times per day for two days and 5 mg four times per day for one day, at which time it is discontinued, Warm compresses are used on the first, second, and third postoperative days.

DISCUSSION The single-walled decompressions as advocated by Kronlein, Naffziger, Hirsch, and Sewell do not provide consistently significant orbital decompressions, although it has been reported that there is a considerable amount of retrodisplacement of the globe in isolated cases. A two-wall decompression as advocated by Walsh and Ogura is the most popular technique today to decompress patients with dysthyroid orbitopathy, but it provides approximately 3 -7 mm of decompression, which is not adequate for these unusual patients with marked proptosis. 7 - 9 Moses and McCord and Cooper and Trokel have advocated multiple wall decompression up to three walls and have

KENNERDELL AND MAROON • ORBITAL DECOMPRESSION

- --

.. - ....:w.

.-,.

-s I

I

I

Figs SA-B. A, left top and bottom, preoperative visual fields of elderly man with optic neuropathy. B, right top and bottom, post-decompression improvement in visual fields .

had good results. 1o •1I The two- or three-wall decompressions are the most commonly performed by the authors, and a three-wall decompression provides 6-10 mm ofretroplacement ofthe globe, which is also adequate for these patients with marked proptosis. The four-wall decompression in the seven orbits that are being reported, has provided 10-17 mm of decompression, which is considered adequate (Fig 2). One patient, a 26-year-old man, had repeated corneal exposure and lid entrapment due to his severe dysthyroid exophthalmos prior to decompression (Fig 3). The second patient, a 21-year-old girl, had chronic unilateral proptosis associated with dysthyroid orbitopathy but with fatty infiltration of inflammatory tissue. This might have represented nonspecific orbital inflammation in a dysthyroid setting (Fig 4). The third patient, a 71-year-old man, had progressive bilateral optic neuropathy associated with the severe exophthalmos due to apical compressive optic neuropathy and this improved after operation as evidenced by the visual field improvement (Fig 5). The fourth patient, a 48-year~0Id woman, decompensated psychologically because of her severe exophthalmos and

became stable and productive after operation as the result of her decompression. The complications included inadvertent laceration of the left lower canaliculus that was repaired primarily. A spinal fluid leak in the left orbit of the elderly patient was thought to be due to the thin incompetent nature of the dura in patients over 65, which has prompted the recommendation that the fourth wall (root) not be included in patients over that age. There was no evidence of increased ocular motility dysfunction following the four-wall orbital decompression in these patients, but it is recognized as a common complication and is just as likely in this operation as other procedures. It is the author's opinion that the procedure must be performed by a skilled and experienced neuroophthalmic orbital surgical team, with the ophthalmologist being familiar with orbital disease and surgery and the neurosurgeon having considerable experience in cranio-orbital junction surgery. Under these circumstances, it is safe and provides an added dimension in reduction of proptosis in patients with severe disorder. 471

OPHTHALMOLOGY. MAY 1982 • VOLUME 89 • NUMBER 5 5. Sewall EC. Operative control of progressive exophthalmos . Arch Otolaryngol 1936; 2 4:6 21-4 . 6 Hirsch O. Surgical decompress ion of malignant exophthalmos . Arch Otolaryngol 1950; 51:325-34. 7. Ogura J. Decompression of the orbit. J Otolaryngol Soc Aust

REFERENCES 1. Kronlein RU. Zur Pathologie und operativen Behandlung der Dermoidcysten der Orbita. Beitr Klin Chir 1889; 4:149-63. 2. Dollinger J. Die Druckentlastung. der Augenhohle durch Entfernung der iiusseren Orbitalwand bei hochgradigern Exophthalmus (Morbus Basedowii) und konsekutiver Hornhauterkrankullg . Dtsch Med Wochenschr 1911 ; 37: 1888-90. 3. Hirsch 0, Urvanek. Behandiung eines excessiven exophthalmos (Basedow) durch Entferrung von Orbitalfett von der Kieferhohle aus. Monatschr Ohrenh Laryngorhinol 1930;

1971; 3:241-3. 8. Ogura JH, Lucente FE Surgical results of orbital decompression of mal ignant exophthalmos. Laryngoscope 1974 ;

84:637 - 44 . 9 Ogura JH , Thawley SE . Orbital decompression of exophthalmos. Otolaryngol Clin North Am 1980; 13(1):29-38. 10. McCord CD Jr, Moses JL. Exposure of the inferior orbit with forni x incision and lateral canthotomy. Ophthalmic Surg 1979;

10(6):53-63.

64 :212-3.

11 . Trokel SL, Cooper WC. Orbital decompression: effect on motility and globe position . Ophthalmology 1979; 86:2064-70.

4. Naffziger HC. Progressive exophthalmos following thyroidectomy; its pathology and treatment. Ann Surg 1931; 94:582-6.

Discussion

by

Bartley R. Frueh, MD Doctors Kennerdell and Maroon have developed a new procedure for treating compressive optic neuropathy of Graves' disease . They have shown that a four-wall decompression can be done transorbitally and that their method gives a significantly greater reduction of proptosis than any of the previous methods reported. More importantly, it decompresses the orbital apex better than any procedure currently in use and similarly to that produced by the Naffziger procedure . 1 . I would certainly agree with their stressing that this procedure cannot be approached lightly, and that when it is to be performed , it be done by an experienced orbital surgeon with the help of an experienced neurosurgeon . They mentioned that the risks include a cerebral spinal fluid leak. Other potential risks to consider must include pulsation of the globe, induction of extraocular muscle imbalance, and when opening the superior periosteum blindly , damage to the ophthalmic division of the fifth cranial nerve . I wonder how much their three-wall decompression, the removal of the floor, medial wall, and the lateral wall all of the way back to the orbital apex , with exposure of the temporal lobe in that area, would accomplish vis a vis a four-wall decompression. Such a three-wall decompression would be technically easier and still allow for decompression of the orbital apex. Perhaps some day we will have the data to answer this question. I would like to differ rather strongly with the authors on their indications for this procedure. I think that there is only one indication: a patient wit~ compressive optic neuropathy of sufficient severity that it is not relieved by the more conventional orbital floor and medial wall decompression. Since at the present time we cannot predict which patients will give an adequate response to a two-wall decompression, I would think that this operation should be done as a two-stage procedure: the lateral wall and orbital roof should be removed only as a secondary procedure when the floor and medial wall have not given a sufficient response. In a recent series of 16 eyes with compressive optic neuropathy at the University of Michigan, 11 responded optimally to a transantral removal of the orbital floor of the medial wall. Quite possiFrom the University of Michigan , Medical Center, Department of Ophthalmology, Ann Arbor, Michigan.

472

bly, the other five eyes would have benefited from removing the entire lateral wall and perhaps the roof also. Perhaps with further research we shall be able to predict which patients are most likely to require the massive decompression through the use of CT scans, measurement of orbital compressibility, lid distensibility, and orbital pressure generated by the lids. With regard to the use of decompression for patients with corneal exposure, in my experience severe corneal exposure has most commonly been found in patients with compressive optic neuropathy who required a decompression anyway . In those patients who had corneal exposure to a significant degree but no compressive optic neuropathy, a recession of the upper eyelid treated the corneal problem well and safely. I do not think this procedure should be done for cosmetic purposes. There are much safer methods for improving the appearance in patients with marked exophthalmos than a four-wall decompression. Recession of the upper eyelid and elevation of the lower eyelid can be done to produce a normal palpebral fissure width and thus provide not only comfort , but a more neady normal appearance. At the most, a two-wall decompression could be done first. To illustrate these points , a 32-year-old woman has Hertel readings of 27 on the right and 26 on the left. After recessing her upper lids, and elevating her lower lids with a scleral graft, she has a nearly normal appearance , although we have not changed her exopthalmos. For that rare group of patients with Hertel readings of greater than 10 mm of difference , and I personally have never seen one, at the most, I would recommend a two-wall decompression, and that only if the patient had no fusion before operation. This would decrease the difference between the two sides to 4 - 6 mm. Recession of the upper lid and elevation of the lower lid could then be performed on one side to create symmetrical palpebral fissures . In summary , while disagreeing with some of their indications for this operation, I think the authors have come up with an ingenious method for decompressing the orbital apex. REFERENCE 1. Naffziger HC. Progressive exophthalmos Clssociated with disorders of the thyroid gland . Ann Surg 1938; 108:529-44.