Conjunctival flaps

Conjunctival

Flaps

Anne Marie Alino, MD,’ Henry D. Perry, MD,le3 Eric D. Donnenfeld, MD,‘s5 Elsa K. Rahn, MD’

Anastasios

J. Kanellopoulos,

MD,4

Purpose: The authors reviewed their experience with total conjunctival flaps (TCF) and partial conjunctival flaps (PCF) for the past 5 years in 61 patients. Methods: Forty-eight patients had TCF and 13 had PCF. Diagnoses for surgery included severe bullous keratopathy for chronic graft failure (not candidates for keratoplasty) (19), herpes zoster ophthalmicus (7), chronic ulcerative keratitis (14), neurotrophic keratitis (2), and herpes simplex keratitis (9). Results: There were seven complications. Four flap retractions occurred in the TCF group, requiring resuturing in two. Three complications occurred in the PCF group. One patient had two flap retractions and recurrent ulceration, requiring tarsorrhaphy. One patient with PCF suffered a perforation after flap retraction, necessitating penetrating keratoplasty. Conclusion: The authors believe conjunctival flaps are underused and should be considered seriously for bullous keratopathy, neurotrophic keratitis, recalcitrant keratitis, and persistent nonhealing epithelial defects. Ophthalmology

1998; 105: I 120-

1123

The conjunctival flap is a well-proven, time-honored treatment for numerous disparate cornea1diseasessharing in common a persistently compromisedocular surface.‘-” The purpose of a conjunctival flap is to restore the integrity of a chronically compromised cornea1 surface, typically damaged becauseof neurotrophic or neuroparalytic disease,severe dry eye, or bullous keratopathy. In doing so, the flaps prevent progressive cornea1 ulceration and secondary infection, as well as control pain, eliminate frequent medications, improve cosmesis,and provide an alternative to invasive surgery or enucleation. The purpose of this article is to stimulate additional interest in conjunctival flaps, which, we believe, currently are underusedand yet represent a viable option for patients with severe ocular surface disease. This study is a retrospective analysis of 6 1 consecutive conjunctival flaps performed over a 5 ‘&-year period. Pa-

Originally Revision

received: February accepted: November

’ Department of Ophthalmology, Meadow, New York. * Department and Harvard

of Ophthalmology, Medical School,

’ North Shore University York University Hospital, 4 Department York, New

9, 1996. 20, 1997. Nassau County

Medical

’ Department of Ophthalmology, tal, New York, New York.

York,

and New

New York

Eye and Ear Infirmary,

Manhattan

Eye, Ear and Throat

Presented as a poster at the Annual Meeting of the American of Ophthalmology, Atlanta, Georgia, November, 1995. Supported Illinois.

East

Massachusetts Eye and Ear Infirmary Boston, Massachusetts.

Hospital, Manhasset, New New York, New York.

of Ophthalmology, York.

Center,

in part by the Lions Club International

Foundation,

New Hospi-

Academy Oakbrook,

Reprint requests to Henry D. Perry, MD, Lions Eye Bank for Long Island, North Shore Umverslty Hospital, 300 Community Drive, Manhasset. NY I 1030.

1120

tient population, underlying disease,and outcomes were examined and tabulated.

Patients and Methods Reportsof all patientswho underwenta Gundersen(full) or partialconjunctival flap (PCF) from June 1989 through November 1994 were reviewed. During this period, 844 transplants were performed. The ratio of transplants to total conjunctival flaps was approximately 1: 18. We reviewed the records of New York Eye and Ear Infirmary for the year 1996 and found 45 I keratoplasties were performed and 6 Gundersen flaps were performed, for a ratio of approximately 1:75. All study subjects were patients from a private cornea1 referral practice (HDP, EDD). The conjunctival flaps were fashioned in a technique similar to the original 1958 description by Gundersen.’ In evaluating patients with pain (e.g., bullous keratopathy, herpes simplex, herpes zoster), a drop of anesthetic was placed over their corneas. If this reduced their pain significantly, a conjunctival flap then was performed for pain relief, and in patients who had no light perception (NLP), a conjunctival flap was performed in lieu of enucleation. In constructing a conjunctival flap, the cornea1 epithelium was debrided carefully with a Bard Parker blade to remove all epithelium and to prevent cyst development under the flap. After the debridement, a 6-O silk suture was passed through the superior aspect of the cornea to fully infraduct the eye into the surgical position in an effort to gain the greatest exposure of the furthest recess of the superior fornix, thus allowing 12 mm of conjunctiva to be dissected. However, even in patients with significant prior surgery, such as previous filtration surgery or cataract surgery, the conjunctiva in the area 12 mm peripheral to the limbus was relatively virgin in nature and permitted good surgical planes until one was within 2 to 3 mm of the surgical limbus. The key to the success of the conjunctival flap surgery was the performance of a thin flap of conjunctiva without any associated Tenon’s capsule. The most critical portion of the procedure is the initial incision, in which great care is necessary

to ensurethat only conjunctivaltissueis incised.The conjuncti-

Alino et al * Conjunctival

Figure 1. Gundersen flap m a patlent wth severe Down’s syndrome peralstent epxhehal defect secondary to herpes aunplex keratltls.

and

val tissue then was brought over the entire cornea to ensure that there was no tension on the flap at any point. A 360” peritomy was performed to increase the mobility of the conjunctiva and allow the conjunctival flap to have no tension on it. It was then secured into place using interrupted 10-O nylon sutures, which usually were removed between 7 and 10 days after surgery (Fig 1). If there was a significant area of stromal loss, it was treated with lamellar keratoplasty, if necessary, or, in rare cases, even penetrating keratoplasty in addition to the conjunctival flap. Partial conjunctival flaps were performed vertically if possible. Eight of the 13 partial flaps either were temporal or nasal, permitting a vertical flap. In the five remaining patients, the flap was horizontal. Of the three flap retractions that occurred, two were in patients with horizontal flaps. The partial flaps were fashioned by debriding the epithelium, so that the conjunctiva would be in contact with Bowman’s membrane. The conjunctiva-only flap then was undermined at the limbus, and a parallel incision was made, usually approximately 5 mm, in a vertical or horizontal arc. The flap was without tension and sutured directly to the cornea with 10-O nylon sutures, which were buried in the cornea. The sutures at the limbus were left exposed.

Results Sixty-one patients underwent conjunctival flap surgery. Fortyeight patients received a Gundersen (full) conjunctival flap, and 13 received a partial flap. The mean patient age was 72 years, with the youngest being 38 years and the oldest being 93 years. There were 32 women and 29 men. Follow-up ranged from 1 month to 6 years, with a mean of 17 months. Preoperative visual acuity ranged from 20/50 to NLP. Postoperative vision ranged from 20/30 (PCF) to NLP. All patients with postoperative NLP had been NLP at baseline, except for one who was LP initially. Vision improved in 12 patients (20%), worsened in 17 patients (28%), and was unchanged in 32 patients (52%). Eleven of the unchanged patients were NLP before surgery. Eleven patients (18%) were NLP at initial presentation, and most were secondary to glaucoma. For these patients, enucleation may have been performed had conjunctival flap not been pursued.

Flaps

The underlying cornea1 diseases are listed in Table I. The largest group was herpetic disease (26%), both herpes zostel virus and herpes simplex virus, followed by cornea1 ulcers (20%), graft-related conditions (18%), and bullous keratopathy (15%). We looked at quality-of-life aspects in this study by evaluating topical medications regarding the number and frequency of drops, as well as the number of visits to the physician. Before surgery, patients were taking an average of 2.58 types of topical medications daily. Seventeen percent of the patients required medicated drops at least every 2 hours. At 1 month after surgery, patients were taking 0.80 kind of drops daily, none more frequent than four times daily. Additionally, 48% of the patients no longer required any drops after flap surgery. Office visits averaged 2.59 per month during the 3 months before conjunctival flap surgery. After surgery, this figure fell to 1.04 average visits per month for the 3-month postconjunctival flap placement. Two patients underwent simultaneous penetrating keratoplasties at the time of their conjunctival flap. Both patients started with counting fingers, and both ended with 20/400 visual acuity. There were seven complications (11.4%) in our study, all flap retractions. Five patients (8%) experienced retraction of their conjunctival flap, which was significant enough to warrant surgical repair (2 in the TCF group and 3 in the PCF group). Two TCFs required resuturing of their retracted flaps, one at 10 days and the other at 21 days after surgery. Additionally, there were two other patients who experienced retraction of their TCFs within 2 weeks and were left untreated without any subsequent complications. Seven days after flap placement, another patient required conversion from a partial to a full conjunctival flap with lamellar keratoplasty. This patient’s TCF retracted at 4 weeks and led to slowly progressive cornea1 thinning. The cornea perforated at 4 months, requiring penetratmg keratoplasty with tarsorrhaphy. One patient with PCF underwent cornea1 gluing 4 weeks after retraction because of progressive cornea1 thinning. The third PCF (retraction) was the only vertical flap that retracted; it was resutured and did well.

Discussion Treating recalcitrant cornea1 surface disease can present a formidable challenge. Described in the German literature in the 187Os,* Gundersen’,” reintroduced, popularized, and described his technique for thin conjunctival

Table 1. Underlying

Cornea1 Disease Preconlunctlval

Underlying Cornea1 Disease

No. (%)

HZV HSV

7 9 I2 10 9

UlCfX

Graft-asaoc1‘3tcd Rullou~ keratquthy Neurotrophlc Neuroparalytlc Drscemetocelc Rand keratopathy HZV

= hcrpcs zostcl wru;

of Patients (11) (15) (20) (16) (15)

6 (10) 3 (5) 3 (5) 1 (2) HSV

= herpea simplex

wru.

Flap

Ophthalmology Table

Disease Herpes (HSV/HZV) Bullous keratopathy Ulcer Graft-related Trauma Neuroparalytlc Neurotrophlc

2. Conjunctival Gundersen’ (1958) (%)

Volume Flap

Studies:

Gundersen” (1969) (%I

Underlying

6, June 1998 Cornea1

Diseases

Paton’ (1970) (%)

InsleP (1988) (%)

Brown’ (1992) (%)

Alino (current) (%)

36 3 22 3 3

64

26 15 20 18 7 5 3

68 13

35 25

30 27 19

3

6 5

6 5

flaps in 1958. Today, the role of the conjunctival flap is well established as an immediate, effective, and definitive treatment for persistent ocular surface disease.7-‘4 In restoring the cornea1 surface, patients are relieved of discomfort, unsightliness, and the arduous task of frequent administration of topical medication, as well as more invasive surgery. The underlying cornea1 diseases in our patient population are similar to those treated originally by Gundersen’.” and those of subsequent authors.9”o+‘2 Table 2 illustrates the various cornea1 diseases among authors performing conjunctival flap surgery. Nine (16%) of our patients underwent conjunctival flap surgery for bullous keratopathy. Gundersen’ asserted that optimal success for these patients occurred when his conjunctival flaps were combined with a lamellar keratectomy. It has been our experience that the conjunctival flap alone has been sufficient. Paton and Milsaukas’ reported similar success using a conjunctival flap alone in patients with bullous keratopathy. Twenty percent of our patients experienced improvement of their visual acuity after surgery. The patients whose visually acuity had improved had a variety of underlying cornea1 diseases, including graft failure, bullous keratopathy, herpes simplex virus, and ulcer, with no one disease dominating. The improvement in vision was caused by thinning of the conjunctival flap and resolution of cornea1 edema or, in some instances, cornea1 infection. No patient with a full conjunctival flap had visual improvement beyond 20/200 without subsequent penetrating keratoplasty. For some, the cornea was sufficiently stabilized by the conjunctival flap to permit subsequent penetrating keratoplasty. This was especially true in the herpetit group, in which conjunctival flap surgery often was the only means of quieting the inflammatory process. We did not observe any perforations under a conjunctival flap, a complication described by Lesher et al” related to herpetic stromal keratitis. In 1987, Lugo12 reported that visual acuity improved in three (43%) of seven patients after conjunctival flap surgery. In 1987, Insler and Pechous” described a 26% improvement in visual acuity after conjunctival flap surgery. All of the patients who Insler and Pechous cited as improving had undergone penetrating keratoplasties subsequent to their conjunctival flaps. This leaves the interpretation of the final visual results of their study somewhat clouded as to the benefit of a conjunctival flap alone.

1122

10.5, Number

Eighteen percent of our patients presented with NLP vision. The role of the conjunctival flap in blind painful eyes should be emphasized. The flap often is placed in lieu of an enucleation, a prospect that can be associated with significant psychological trauma. By placing a conjunctival flap, patients can keep their eyes while usually obtaining good cosmesis. Additionally, the flap can provide a surface for the placement of a cosmetic scleral shell or painted contact lens. Some may argue that opaque blind eyes are difficult to examine adequately and should be enucleated prophylactically because they may harbor malignancy. Indeed, past authors have quoted as high as 40% of opaque blind eyes contained unsuspected malignant melanoma.14 More recent literature finds this number to be much lower: approximately 2% and possibly lower.14 The decreased percentage over the years is attributed to improved diagnostic evaluation, especially ultrasonography. In our NLP group, no patient required enucleation, and all noted significant lessening of their ocular pain. The amount of time and effort that patients need to devote to their eye condition was enormous. Many of the patients were elderly and required frequent administration of multiple eye medications, as well as closely spaced visits to the ophthalmologist, often a cornea1 specialist. After surgery, many of our patients did not require any medications and had far fewer office visits. Other authors have emphasized the significant improvement in patients’ lifestyles after flap placement.5s9 Complications encountered after conjunctival flap surgery are relatively uncommon. The most frequently reported complications include flap retraction, conjunctival buttonholes and erosions, epithelial inclusion cysts, and cornea1 perforations. Many of these can be avoided at the time of surgery by avoiding excess tension on the conjunctiva. Of our patients, 11.4% experienced complications, and all were flap retractions. Our flap retractions all occurred within 1 month of surgery. Similarly, Paton and Milsaukas’ had 11 (85%) of 13 flaps retract within the first postoperative month. Our complication rate of 11.4% is similar to the rates reported by other authors.‘-” This rate is influenced by patient population and severity of disease in addition to technical skill. Previous cataract or glaucoma surgery exponentially increases the technical difficulty of performing conjunctival flap surgery. In the full conjunctival flap group, there were 4 (8.3%) of 48 flap retractions versus 3 (23.1%) of 13 in the partial

Alino et al -

Conjunctival

flap group (Fisher’s exact test, P = 0.11). This suggests a trend toward increased flap retraction in the partial flap group. Intuitively, there should be no difference in the degree of conjunctival retraction between the two groups. Perhaps in the partial group, the horizontal flaps were more likely to be affected by gravity and blinking action rather than the vertical partial flaps. In addition, with the partial flap, the edge of the flap starts out closer to the edge of the defect, and with postoperative retraction, the defect becomes exposed more often. Ptosis has been reported as a complication of conjunctival flap surgery.‘,9 Gundersen thought the ptosis was caused by the incorporation of Tenon’s capsule within the flap or insufficient dissection of the flap from its lateral and medial attachments. Paton and Milsaukas” similarly had attributed postconjunctival flap ptosis caused by undue traction. We did encounter ptosis in one patient after conjunctival flap placement. This low number may be because of strict attention to complete conjunctival dissection before bringing down the flap. The incidence of ptosis after conjunctival flap surgery is clear from the literature. Several reports on conjunctival flap surgery do not mention ptosis when discussing their complications.5,‘0X’5 In those articles that report ptosis as a complication, it does not appear to have been of significance. The authors refer to the ptosis as “transient” and “of a small amount.” ‘,9 Ptosis is difficult to evaluate because many patients may have varying degrees present before surgery. To fully answer this question, a prospective study must be performed. The placement of a conjunctival flap over the cornea does have some drawbacks. One potential problem is the persistence of infection under the flap. Although rare, cornea1 perforation under a conjunctival flap has been reported in two patients with stromal herpes keratitis.15’16 The conjunctival flap also hinders examination of the anterior and posterior segments. Penetration of topical medications (e.g., glaucoma medications) is compromised by the conjunctival flap. These undesirable effects must be weighed against the myriad of benefits of conjunctival flap placement. It is our conclusion that the conjunctival flap is a simple, useful, and effective treatment for ocular surface disease resistant to medical therapy. Prompt resolution of pain and stabilization of the ocular surface provide comfort and obviate the need for intensive management of ocular disease. Although there may be a dilemma for the physician and patient in “throwing in the towel” with regard to battling chronic ocular surface disease, in the long term, there often is a psychological, as well as a cosmetic, benefit to performing conjunctival flap surgery. The patient should be aware that this is a reversible procedure. Thinning of the conjunctival flap over the cornea allows the limbus to be discernible, and overall, redness of the eye usually decreases significantly. The eyes be-

Flaps

come very quiet and relatively white, with a difference in color being noted through the thinned conjunctival flap. Although the conjunctival flap does not permit good visual acuity nor good visualization of the anterior chamber details, it is good enough to provide a cosmetic benefit to the patient, who usually presents with very significantly inflamed eyes that often are unsightly. The potential for limited visual improvement has been shown. With its low complication rate and its proven immediate and sustained benefit, we believe that conjunctival flap surgery should be considered whenever ophthalmologists are faced with resistant ocular surface disease.

References 1. Gundersen T. Conjunctival flaps in the treatment of cornea1 disease with reference to a new technique of application. Arch Ophthalmol 1958;60:880-8. 2. Scholer. Jahresberichte uber die Wirksamkeit der AugenKlinik, in den Jahren 1874-80, Berlin, H. Peters, 18751881. 3. Donzis PB, Mondino BJ. Management of noninfectious cornea1 ulcers. Surv Ophthalmol 1987; 32:94- 110. 4. Portnoy SL, Insler MS, Kaufman HE. Surgical management of cornea1 ulceration and perforation. Surv Ophthalmol 1989;34:47-58. 5. Brown DD, McCulley JP, Bowman RW, Halsted MH. The use of conjunctival flaps in the treatment of herpes keratouveitis. Cornea 1992; 11:44-6. 6. Buxton JN, Fox ML. Conjunctival flaps in the treatment of refractory Pseudomonas cornea1 abscess. Ann Ophthalmol 1986; 18:315-g. 7. Thoft RA. Conjunctival and limbal surgery for cornea1 diseases. In: Smolin G, Thoft RA, eds. The Cornea: Scientific foundations and clinical practice, 3rd ed. Boston: Little, Brown, 1994; 7 1l-4. 8. Conjunctival flaps. Ophthalmic Surg 1987; 18:455-g. 9. Paton D, Milauskas AT. Indications, surgical technique, and results of thin conjunctival flaps on the cornea: a review of 122 Consecutive Cases. Int Ophthalmol Clin 1970; 10:329-45. 10. Insler MS, Pechous B. Conjunctival flaps revisited. Ophthalmic Surg 1987; 18:455-g. 11. Gundersen T, Pearlson HR. Conjunctival flaps for cornea1 disease: their usefulness and complications. Trans Am Ophthalmol Sot 1969;67:78-95. 12. Lugo M, Arentsen JJ. Treatment ofneurotrophic ulcers with conjunctival flaps. Am J Ophthalmol 1987; 103:71 l-2. 13. Buxton JN. Therapeutic conjunctival flaps and tarsorrhaphy. Contact Intraocul Lens Med J 1981;7:150-2. 14. Davidorf FH, Letson AD, Weiss ET, Levine E. Incidence of misdiagnosed and unsuspected choroidal melanomas. A 50-year experience. Arch Ophthalmol 1983; 101:410-2. 15. Lesher MP, Lohman LE, Yeakley W, Lass J. Recurrence of herpetic stromal keratitis after a conjunctival flap surgical procedure [letter]. Am J Ophthalmol 1992; 114:231-3. 16. Rosenfeld SI, Alfonso EC, Gollamudi S. Recurrent herpes simplex infection in a conjunctival flap [letter]. Am J Ophthalmol 1993; 116:242-4.

1123