- Email: [email protected]
Hard Palate and Free Tarsal Grafts as Posterior Lamella Substitutes in Upper Lid Surgery
Hard Palate and Free Tarsal Grafts as Posterior Lamella Substitutes in Upper Lid Surgery Igal Leibovitch, MD,1 Raman Malhotra, FRCOphth,2 Dinesh Selva, FRANZCO1,3 Objectives: To present the surgical outcomes and postoperative complications in a series of patients who underwent upper lid surgery using autogenous hard palate grafts (HPGs) or free tarsal grafts (FTGs) as posterior lamella replacement material. Design: Retrospective, comparative, interventional case series. Patients: Thirty-one consecutive patients who were operated in 2 oculoplastics centers between July 2000 and January 2005. Methods: All patients’ clinical records were reviewed. Main Outcome Measures: Postoperative upper eyelid contour and viability, ocular discomfort, keratopathy, and corneal edema, as well as assessment for donor site complications and final graft dimensions. Results: There were 31 patients who underwent upper lid surgery (15 HPGs, 16 FTGs). The complications in the HPG group included corneal edema or transient keratopathy (13.3%), partial graft dehiscence (13.3%), upper lid retraction (13.3%), and necrosis of the overlying skin flap (6.7%). There were no significant postoperative complications in the FTG group during a mean follow-up period of 13.5⫾5 months. Donor site complications included 2 cases of mild upper lid retraction and central peaking. There were an average of 17% decrease in FTG vertical height and a 24% decrease in HPG vertical height during the follow-up period. Conclusion: Hard palate grafts may be associated with a higher rate of complications in upper lid surgery relative to FTGs, although most complications are temporary. Graft contraction could be reduced by oversizing. Ophthalmology 2006;113:489 – 496 © 2006 by the American Academy of Ophthalmology.
Hard palate grafts (HPGs) and free tarsal grafts (FTGs) are widely used as posterior lamella substitutes in lid surgery.1–9 Most previous series report the outcome with these autogenous tissues in lower lid surgery, but less often in upper lid surgery. Although a number of series have reported a low complication and donor site morbidity rate with HPGs, these were based mostly on the outcome in lower lid surgery.1–7 Therefore, some clinicians remain concerned regarding the possible ocular side effects associated with using this material in upper lid surgery.
Originally received: August 22, 2005. Accepted: November 30, 2005. Manuscript no. 2005-792. 1 Oculoplastic & Orbital Division, Department of Ophthalmology & Visual Sciences, Royal Adelaide Hospital, University of Adelaide, Adelaide, Australia. 2 Corneoplastic Unit, Queen Victoria Hospital, East Grinstead, United Kingdom. 3 Departments of Surgery and Medicine, University of Adelaide, Adelaide, Australia. The authors state that no conflicting relationships exist. The work received no financial support. Correspondence to Dr Igal Leibovitch, Oculoplastic & Orbital Division, Department of Ophthalmology & Visual Sciences, Royal Adelaide Hospital, North Terrace, Adelaide, 5000, South Australia, Australia. E-mail: [email protected]. © 2006 by the American Academy of Ophthalmology Published by Elsevier Inc.
We report our experience with HPGs in upper lid surgery and compare the outcome and complications with those of a series of patients operated using FTGs.
Materials and Methods This is a retrospective, comparative, interventional case series of 31 consecutive patients who underwent upper eyelid surgery using autogenous FTGs or HPGs as posterior lamella replacement material, between July 2000 and January 2005. All patients underwent a complete ophthalmic evaluation before the operation, including upper eyelid contour, tarsal height, and anterior segment assessment. The potential donor sites (contralateral upper lid or hard palate) were also evaluated for any signs of pathology. The authors’ general preference is to use FTGs as the grafting material in upper lid reconstruction. The reasons for using HPGs and not FTGs are presented in “Results.”
Hard Palate Graft Harvesting and Grafting A standard harvesting technique was used.3,7 Local anesthetic (2% lidocaine and adrenaline 1:100 000) was injected into the hard palate mucosa and mucoperiosteum, including the area around the greater palatine and incisive foramina. After the required graft size was marked, 2 parallel incisions were made between the median raphe and the gingival mucosa using a no. 15 blade or a corneal pocket knife. An edge of the graft was lifted and a dissection ISSN 0161-6420/06/$–see front matter doi:10.1016/j.ophtha.2005.11.017
489
Ophthalmology Volume 113, Number 3, March 2006 carried out in the submucosal plane using the blade or the knife. Hemostasis was achieved using pressure, minimal cautery, or an absorbable gelatin sponge soaked in thrombin. A surgical stent was used in some cases. The harvested HPG was carefully thinned by removing fatty submucosa with scissors. The graft was then wrapped in moistened gauze. Suturing of the HPG to the upper lid defect was done with interrupted buried 6-0 Vicryl sutures (Ethicon, Somerville, NJ), which were passed half-thickness through the nonmucosal part of the graft. Special emphasis was given to avoid any exposed suture material that could cause ocular irritation. During suturing, the graft was trimmed to the required size, and at the end of the procedure, the final dimensions of the graft were recorded.
Free Tarsal Graft Harvesting and Grafting Local anesthetic (2% lidocaine and adrenaline 1:200 000) was injected beneath the pretarsal upper eyelid skin before eversion of the eyelid using a Desmarres retractor for subconjunctival injection of additional anesthetic above the superior tarsal margin. The tarsus was incised 4 to 5 mm above the lid margin, parallel to the eyelid margin, with the length of the horizontal incision up to 16 mm, depending on the available upper eyelid tarsus. This was followed by 2 vertical incisions at each end of the horizontal incision, towards the upper border of the tarsus. The graft was then dissected from the loosely attached levator aponeurosis, Müller’s muscle, and conjunctiva. The donor site was not sutured. The graft was then wrapped in moistened gauze. Suturing of the FTG to the
Figure 1. A, Left upper lid cicatricial entropion. B, Upper lid marginal rotation is performed, and the hard palate graft is prepared (arrow). C, The graft is sutured to replace the upper lid posterior lamella.
490
Leibovitch et al 䡠 Hard Palate and Free Tarsal Grafts in Upper Lid Surgery Table 1. Patients’ Characteristics, Postoperative Complications, and Follow-up in Cases Operated Using Hard Palate Grafts (HPGs) in Upper Lid Surgery Patient*
Gender/Age (yrs)
Upper Lid Pathology
Upper Lid Surgery Combined with HPG
1 2
M/70 M/63
Cicatricial entropion SCC
3
M/42
Cicatricial entropion
4
F/81
BCC
5
M/68
BCC
6
F/83
BCC
7 8 9
F/53 M/40 M/62
Cicatricial entropion Cicatricial entropion Cicatricial entropion
10
M/63
MCC
Surgical excision ⫹ rotation flap
11
F/90
BCC
12
F/84
BCC
Surgical excision ⫹ advancement flap and FTSG Surgical excision ⫹ advancement flap
13
M/67
Cicatricial entropion
14 15
F/53 F/72
Cicatricial entropion Cicatricial entropion
Marginal lid rotation Surgical excision ⫹ advancement flap and FTSG Marginal lid rotation Surgical excision ⫹ advancement flap and FTSG Surgical excision ⫹ periosteal flaps ⫹ advancement flap Surgical excision, Cutler–Beard flap and FTSG Marginal lid rotation Marginal lid rotation Marginal lid rotation
Marginal lid rotation ⫹ buccal mucosa graft Marginal lid rotation Marginal lid rotation ⫹ buccal mucosa graft
Postoperative Complications
Final Outcome of Complications
Follow-up Period (mos)
Ocular discomfort Ocular discomfort
Complete resolution Complete resolution
11 39
Transient epithelial keratopathy Upper lid retraction (1 mm)
Complete resolution
51
1-mm retraction
21
None
NA
Donor site bleeding
Bleeding controlled with packing
None None Partial graft dehiscence
NA NA Good final contour after surgical repair Good final contour after surgical repair of dehiscence Lateral eyelid peaking
Partial graft dehiscence, mild ocular discomfort, mild epithelial keratopathy Upper lid retraction (1.5 mm) Necrosis of skin flap at 3 mos, required removal of HPG and use of pericranial forehead flap None None None
Exenteration due to tumor recurrence
10 9 22 9 7 8 12 14
NA
10
NA NA
12 10
BCC ⫽ basal cell carcinoma; F ⫽ female; FTSG ⫽ full-thickness skin graft; M ⫽ male; MCC ⫽ Merkel cell carcinoma; NA ⫽ not applicable; SCC ⫽ squamous cell carcinoma. *Cases 1–11 were operated at the Royal Adelaide Hospital; cases 12–15 were operated at the Queen Victoria Hospital.
upper lid defect was done with interrupted buried 6-0 Vicryl sutures, which were passed half-thickness through the nonmucosal part of the graft, to avoid any exposed suture material that could cause ocular irritation. During suturing, the graft was trimmed to the required size, and at the end of the procedure, the final dimensions of the graft were recorded. All patients were treated postoperatively with topical antibiotics for 2 weeks. The postoperative follow-up schedule was generally 1 week and 1, 3, 6, and 6 months thereafter. Patients who could not attend at least the first 6 months’ follow-up period were excluded from the study. Postoperative evaluation included upper eyelid contour and position, reconstructed upper eyelid viability, evaluation of ocular discomfort, and keratopathy. The graft dimensions were measured after 6 months in all cases where the operated upper lid could be everted in the outpatient clinic. All cases treated with FTGs were assessed for donor site complications, including ocular discomfort, keratopathy, upper eyelid entropion, contour abnormalities, retraction, and ptosis. The hard palate was assessed regularly in all cases where HPGs were used.
Results There were 31 patients (mean age, 66⫾12 years [range, 40 –90]) who underwent upper lid surgery. Fifteen patients (7 female, 8 male; mean age, 66⫾15) were treated using an autogenous HPG, and 16 patients (9 female, 7 male; mean age, 65⫾9) were treated using an autogenous FTG. The indications for surgery in the HPG group included upper lid defect reconstruction after tumor excision using Mohs’ surgery (7 patients [47%]) and upper lid cicatricial entropion (8 patients [53%]) (Fig 1, Table 1). The indications for surgery in the FTG group included upper lid defect reconstruction after tumor excision using Mohs’ surgery (10 patients [62.5%]), upper lid cicatricial entropion (4 patients [25%]), and upper lid retraction from thyroid eye disease (2 patients [12.5%]) (Table 2). The reasons for choosing HPG as the preferred grafting material included insufficient height of the contralateral tarsal plate to enable harvesting of adequate FTG with preservation of 4 mm of residual tarsus (cases 1, 3, 4, 6 –9, 12–15), avoiding morbidity to the contralateral-only eye (case 2), and patient’s preference (cases 5, 10, 11).
491
Ophthalmology Volume 113, Number 3, March 2006 Table 2. Patients’ Characteristics, Postoperative Complications, and Follow-up in Cases Operated Using Free Tarsal Grafts (FTGs) in Upper Lid Surgery Patient*
Gender/Age (yrs)
Upper Lid Pathology
1 2
M/65 F/70
Retraction (TED) Entropion
Scar release Marginal lid rotation
3 4
F/51 M/69
Entropion SCC
5
F/79
6
F/64
7 8
F/64 M/72
9
F/72
BCC
Marginal lid rotation Surgical excision ⫹ pericranial flap and FTSG Surgical excision ⫹ advancement flap Surgical excision ⫹ advancement flap and FTSG Marginal lid rotation Surgical excision ⫹ Cutler–Beard flap Surgical excision ⫹ advancement flap
10
F/53
SCC
11 12
F/55 M/63
Retraction (TED) BCC
13
M/66
SCC
14
M/70
BCC
15
F/73
BCC
16
M/75
Sebaceous carcinoma BCC Entropion BCC
Entropion
Upper Lid Surgery Combined with FTG
Surgical excision ⫹ Cutler–Beard flap Canthal sling Surgical excision ⫹ Cutler–Beard flap Surgical excision ⫹ Cutler–Beard flap Surgical excision ⫹ anterior lamellar advancement Surgical excision ⫹ transposition flap Marginal lid rotation
Postoperative Complications
Final Outcome of Complications
Follow-up Period (mos)
None Mild upper lid retraction and central peak in donor eyelid None None
NA Good contour in donor eyelid after division of band NA NA
12 6 11 8
None
NA
15
None
NA
14
None None
NA NA
10 11
Mild upper lid retraction and central peak in donor eyelid None
Good contour in donor eyelid after division of band NA
12
Mild ocular discomfort None
Complete resolution NA
21 16
None
NA
18
None
NA
18
None
NA
22
None
NA
8
14
BCC ⫽ basal cell carcinoma; F ⫽ female; FTSG ⫽ full-thickness skin graft; M ⫽ male; NA ⫽ not applicable; SCC ⫽ squamous cell carcinoma; TED ⫽ thyroid eye disease. *Cases 1– 4 were operated at the Queen Victoria Hospital; cases 5–16 were operated at the Royal Adelaide Hospital.
Table 3. Initial and Final Graft Dimensions in Cases Operated Using Hard Palate Grafts (HPGs) in Upper Lid Surgery
Patient
Initial Graft Dimensions Used (Height*Width/Area) (mm*mm/mm2)
Graft Dimensions on Last Follow-up (Height*Width/Area) (mm*mm/mm2)
% Decrease in Graft Dimensions (Height/Width/Area) (mm/mm/mm2)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
7ⴱ23/91 6ⴱ20/120 6ⴱ18/108 7ⴱ16/112 6ⴱ20/120 7ⴱ21/147 8ⴱ19/152 NA 6ⴱ15/90 NA NA 6ⴱ14/84 7ⴱ17/119 7ⴱ14/98 7ⴱ15/105
4ⴱ18/72 5ⴱ18/90 6ⴱ16/96 5ⴱ14/70 NA 5ⴱ15/75 6ⴱ16/96 NA 4ⴱ13/52 NA NA 5ⴱ10/50 NA 7ⴱ14/98 NA
43/22/21 17/10/25 0/11/11 29/12/37 NA 29/29/49 25/16/37 NA 33/13/42 NA NA 17/29/40 NA 0/0/0 NA
NA ⫽ data not available.
492
Leibovitch et al 䡠 Hard Palate and Free Tarsal Grafts in Upper Lid Surgery Table 4. Initial and Final Graft Dimensions in Cases Operated Using Free Tarsal Grafts (FTGs) in Upper Lid Surgery
Patient
Initial Graft Dimensions Used (Height*Width/Area) (mm*mm/mm2)
Graft Dimensions on Last Follow-up (Height*Width/Area) (mm*mm/mm2)
% Decrease in Graft Dimensions (Height/Width/Area) (mm/mm/mm2)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
5ⴱ14/70 4ⴱ15/60 4ⴱ14/56 4ⴱ15/60 5ⴱ15/75 4ⴱ11/44 4ⴱ14/56 4ⴱ14/56 4ⴱ10/40 4ⴱ15/60 4ⴱ13/52 4ⴱ14/56 4ⴱ16/64 4ⴱ10/40 4ⴱ9/36 4ⴱ14/56
4ⴱ11/44 3ⴱ12/36 4ⴱ12/48 NA 4ⴱ13/52 3ⴱ9/36 NA NA 4ⴱ9/36 NA 4ⴱ12/48 2ⴱ9/18 NA 4ⴱ10/40 3ⴱ9/27 3ⴱ10/30
20/21/37 25/20/40 0/14/14 NA 20/13/31 25/18/18 NA NA 0/10/10 NA 0/8/8 50/36/68 NA 0/0/0 25/0/25 25/29/46
NA ⫽ data not available.
Hard palate graft dimensions were recorded in 12 cases. The average HPG dimensions were 6.7⫾0.6 mm in height and 17.7⫾3.0 mm in width. The average graft area was 118⫾24.3 mm2 (Table 3). All FTGs were harvested from the contralateral upper lid. The average graft dimensions were 4.1⫾0.3 mm in height and 13.2⫾2.0 mm in width. The average FTG area was 54.6⫾10.5 mm2 (Table 4). The complications recorded in patients treated with HPG during a mean follow-up period of 16.0⫾16 months (range, 7–51) included ocular irritation or discomfort in 3 patients (20%), corneal edema or transient keratopathy in 2 (13.4%), partial graft dehiscence in 2 (13.4%) (Fig 2), upper lid retraction in 2 (13.4%)
(Figs 3, 4), and necrosis of the overlying skin flap in 1 (6.7%) (Table 1). There were no significant complications in 8 cases (53.3%). No cases of HPG hypersecretion were recorded. Donor site complications included only one case (6.2%) of excessive bleeding from the hard palate site in the recovery room, which required packing. There were no significant complications that were recorded in the patients treated with FTGs during a mean follow-up period of 13.5⫾5 months (range, 6 –22) (Table 2). Only one patient (6.2%) had moderate ocular discomfort, which resolved after 4 weeks. The donor (contralateral) upper lid complications were 2 cases (12.4%) of mild upper lid retraction and central peaking from a
Figure 2. Dehiscence of the hard palate graft from the lower edge of the tarsal plate.
493
Ophthalmology Volume 113, Number 3, March 2006
Figure 3. A, Left upper lid defect after basal cell carcinoma excision. B, Postoperative result at 3 months shows lateral upper lid retraction.
fibrous band. Division of this band resulted in significant contour improvement. The final dimensions of the HPG were recorded in 9 patients (Table 3). The average graft height was 5.2⫾1.0 mm, and the width was 15.2⫾0.7 mm. The average graft area was 78⫾19 mm2, which equals a 36% decrease in average HPG area compared with initial grafting. The vertical graft height decreased on average by 24% (1.5 mm) compared with the initial graft placed. The final dimensions of the FTG were recorded in 11 patients (Table 4). The average graft height was 3.5⫾0.7 mm, and the width was 10.5⫾1.5 mm. The average graft area was 37⫾11 mm2, which equals a 27% decrease in average FTG area compared with initial grafting. The vertical graft height decreased on average by 17% (0.6 mm) compared with the initial graft placed.
494
Discussion The successful use of HPGs for eyelid reconstruction after tumor excision was first described by Siegel in 1985.3 Since then, HPGs have been used for different eyelid pathologies such as cicatricial entropion and trichiasis,4,7,8 eyelid retraction, and facial palsy.5–7,9 In one of the earliest and largest series,7 Cohen and Shorr reported their experience in using HPGs for eyelid reconstruction in 25 cases. Only 6 (24%) were upper lid cases. In a more recent series by Wearne et al,9 the authors presented 102 eyelids operated using autogenous HPGs, but all were performed for the treatment
Leibovitch et al 䡠 Hard Palate and Free Tarsal Grafts in Upper Lid Surgery
Figure 4. A, Left upper lid defect after basal cell carcinoma excision. B, Postoperative result at 3 months shows a mild lateral upper lid peak.
of lower eyelid displacement. To the best of our knowledge, our series of outcomes of upper lid surgery is the largest to date. The decision whether to use FTGs or HPGs is made individually for each case and is based on various factors. An autogenous FTG is an excellent material that provides flexibility, rigidity and contour.1 However, there are 2 important considerations when using FTGs: (1) the amount of available material is limited by the size of the donor tarsal plate and the requirement to leave at least 4 mm of the tarsal plate inferiorly to minimize complications,1,2 and (2) there are possible donor site complications. Contour peaking, which was noted in 2 donor upper lids in our series, was the only significant complication after FTG harvesting. In both
cases, it was attributed to a horizontal fibrotic band at the superior border of the remaining tarsus. Excision of the band resulted in significant contour improvement. From our experience, harvesting of FTGs does not lead to significant donor site morbidity; however, subsequent procedures to alter the position of the donor eyelid may be less predictable.2 The presence of ptosis or retraction in the proposed donor eyelid should lead the surgeon to consider correction before graft harvesting, or an alternative graft material. An HPG provides stiffness, thickness, and flexibility to the upper lid, as well as a mucosal lining.3,4,7 It is readily available and is not associated with any morbidity to the contralateral eyelid. In all previous series in which HPGs
495
Ophthalmology Volume 113, Number 3, March 2006 were used in lower lid surgery, the authors reported only minimal ocular and eyelid complications. Siegel3 reported no complications in his first series of 11 patients, and so did Kersten et al6 (25 lower eyelids) and Cohen and Shorr7 (19 lower lids and 6 upper lids). In Wearne et al’s more recent series9 (102 lower lids), 85% of cases achieved satisfactory lid position, and the rest had inadequate lengthening or significant recurrence of lower lid displacement. The complications recorded in our patients included ocular irritation or discomfort (20%), corneal edema or transient keratopathy (13.3%), partial graft dehiscence (13.3%), upper lid retraction (13.3%), and necrosis of the overlying skin flap (6.7%) (Table 1). There were no significant complications in 8 cases (53.3%). This rate of complications is higher than what was reported in previous series and could be related to placing the graft in the upper lid. The ocular irritation and the keratopathy could result from direct contact between the ocular surface and the keratinized epithelium of the HPG.6,7 In fact, the presence of keratinized epithelium has discouraged some clinicians from using HPGs in the upper lid due to the possible effects on the cornea. However, this side effect was temporary in all our cases and resolved after several weeks, possibly in correlation with the gradual metaplasia of the epithelium to a nonkeratinzed type.6,7 Hard palate graft dehiscence was not reported in previous series dealing with lower lid grafting. In our series, dehiscence occurred early and, possibly, due to infection or altered healing. One of the 2 patients with partial graft dehiscence was diagnosed with Stevens–Johnson syndrome and was treated with systemic steroids, which could have an effect on healing. Local postoperative infection is an important factor that can result in graft dehiscence, and although there were no obvious signs of infection in any of our cases, it could have been mild and not easily detected. Donor site complications after HPG harvesting have been reported in several studies. The most common complication is hemorrhage (10% of cases), which occurs during the immediate postoperative period or even later (7–12 days).7,9 In most cases, the hemorrhage is controlled with local pressure and systemic antibiotics. In our series, only one case of hemorrhage occurred (6.7%), which responded to local packing. Other less common complications previously reported are development of oronasal fistulas, and candidal infection.10 These complications were not encountered in our series. Another uncommon complication that was reported in HPG cases is mucus production from minor salivary glands located in the graft.11 These cases were treated successfully with cryotherapy or graft removal. Although we have had several similar cases of hypersecretion from HPGs that required graft removal, all occurred in lower lid grafts. On histology, these grafts showed mucus-secreting glands. This complication did not occur in any of our upper lid grafts. It is generally reported that the incidence of postoperative HPG contraction is low. In Siegel’s first report on HPGs, he stated that they contract minimally, and advised oversizing the graft by 1 to 2 mm.3 Cohen and Shorr7 measured 17 grafts and found that only 3 demonstrated contraction, and that no graft contracted ⬎0.5 mm. They did not specify their exact follow-up period. Other authors
496
reported contraction of ⱕ15%.5,6,12 In a recent study, Sullivan and Dailey compared graft contraction of acellular dermis with that of hard palate mucosa in lower eyelid surgery.13 They found that HPG vertical height contraction ranged from 0% to 44% (mean, 16%). In our study, there was significant contraction in HPG vertical height, with an average of 24% (range, 0%– 43%). This degree of contraction is higher than in previous reports, but is similar to Sullivan and Dailey’s findings. Interestingly, we also found a significant contraction in FTG size (27% in graft area and 17% in vertical height), but the exact reason for these findings is not clear. The small study groups, retrospective design, and lack of randomization preclude definite conclusions; however, we believe that oversizing the HPG may have a role in minimizing the effects of graft contraction on the final outcome. In conclusion, although HPGs are considered excellent material for posterior lamella replacement in lower lid surgery, they may be associated with a higher rate of complications in upper lid surgery and when compared with FTGs. Ocular irritation and keratopathy are temporary and usually resolve after several weeks.
References 1. Stephenson CM, Brown BZ. The use of tarsus as a free autogenous graft in eyelid surgery. Ophthal Plast Reconstr Surg 1985;1:43–50. 2. Leibovitch I, Selva D, Davis G, Ghabrial R. Donor site morbidity in free tarsal grafts. Am J Ophthalmol 2004;138:430 –3. 3. Siegel RJ. Palatal grafts for eyelid reconstruction. Plast Reconstr Surg 1985;76:411– 4. 4. Silver B. The use of mucous membrane from the hard palate in the treatment of trichiasis and cicatricial entropion. Ophthal Plast Reconstr Surg 1986;2:129 –31. 5. Bartley GB, Kay PP. Posterior lamellar eyelid reconstruction with a hard palate mucosal graft. Am J Ophthalmol 1989;107: 609 –12. 6. Kersten RC, Kulwin DR, Levartovsky S, et al. Management of lower-lid retraction with hard-palate mucosa grafting. Arch Ophthalmol 1990;108:1339 – 43. 7. Cohen MS, Shorr N. Eyelid reconstruction with hard palate mucosa grafts. Ophthal Plast Reconstr Surg 1992;8:183–95. 8. Mannor GE, Mathers WD, Wolfley DE, Martinez JA. Hardpalate mucosa graft in Stevens-Johnson syndrome. Am J Ophthalmol 1994;118:786 –91. 9. Wearne MJ, Sandy C, Rose GE, et al. Autogenous hard palate mucosa: the ideal lower eyelid spacer? Br J Ophthalmol 2001; 85:1183–7. 10. Kim JW, Kikkawa DO, Lemke BN. Donor site complications of hard palate mucosal grafting. Ophthal Plast Reconstr Surg 1997;13:36 –9. 11. Pelletier CR, Jordan DR, Brownstein S, Li S. An unusual complication associated with hard palate mucosal grafts: presumed minor salivary gland secretion. Ophthal Plast Reconstr Surg 1998;14:256 – 60. 12. Beatty RL, Harris G, Bauman GR, Mills MP. Intraoral palatal mucosal graft harvest. Ophthal Plast Reconstr Surg 1993;9: 120 – 4. 13. Sullivan SA, Dailey RA. Graft contraction: a comparison of acellular dermis versus hard palate mucosa in lower eyelid surgery. Ophthal Plast Reconstr Surg 2003;19:14 –24.
Hard palate grafts (HPGs) and free tarsal grafts (FTGs) are widely used as posterior lamella substitutes in lid surgery.1–9 Most previous series report the outcome with these autogenous tissues in lower lid surgery, but less often in upper lid surgery. Although a number of series have reported a low complication and donor site morbidity rate with HPGs, these were based mostly on the outcome in lower lid surgery.1–7 Therefore, some clinicians remain concerned regarding the possible ocular side effects associated with using this material in upper lid surgery.
Originally received: August 22, 2005. Accepted: November 30, 2005. Manuscript no. 2005-792. 1 Oculoplastic & Orbital Division, Department of Ophthalmology & Visual Sciences, Royal Adelaide Hospital, University of Adelaide, Adelaide, Australia. 2 Corneoplastic Unit, Queen Victoria Hospital, East Grinstead, United Kingdom. 3 Departments of Surgery and Medicine, University of Adelaide, Adelaide, Australia. The authors state that no conflicting relationships exist. The work received no financial support. Correspondence to Dr Igal Leibovitch, Oculoplastic & Orbital Division, Department of Ophthalmology & Visual Sciences, Royal Adelaide Hospital, North Terrace, Adelaide, 5000, South Australia, Australia. E-mail: [email protected]. © 2006 by the American Academy of Ophthalmology Published by Elsevier Inc.
We report our experience with HPGs in upper lid surgery and compare the outcome and complications with those of a series of patients operated using FTGs.
Materials and Methods This is a retrospective, comparative, interventional case series of 31 consecutive patients who underwent upper eyelid surgery using autogenous FTGs or HPGs as posterior lamella replacement material, between July 2000 and January 2005. All patients underwent a complete ophthalmic evaluation before the operation, including upper eyelid contour, tarsal height, and anterior segment assessment. The potential donor sites (contralateral upper lid or hard palate) were also evaluated for any signs of pathology. The authors’ general preference is to use FTGs as the grafting material in upper lid reconstruction. The reasons for using HPGs and not FTGs are presented in “Results.”
Hard Palate Graft Harvesting and Grafting A standard harvesting technique was used.3,7 Local anesthetic (2% lidocaine and adrenaline 1:100 000) was injected into the hard palate mucosa and mucoperiosteum, including the area around the greater palatine and incisive foramina. After the required graft size was marked, 2 parallel incisions were made between the median raphe and the gingival mucosa using a no. 15 blade or a corneal pocket knife. An edge of the graft was lifted and a dissection ISSN 0161-6420/06/$–see front matter doi:10.1016/j.ophtha.2005.11.017
489
Ophthalmology Volume 113, Number 3, March 2006 carried out in the submucosal plane using the blade or the knife. Hemostasis was achieved using pressure, minimal cautery, or an absorbable gelatin sponge soaked in thrombin. A surgical stent was used in some cases. The harvested HPG was carefully thinned by removing fatty submucosa with scissors. The graft was then wrapped in moistened gauze. Suturing of the HPG to the upper lid defect was done with interrupted buried 6-0 Vicryl sutures (Ethicon, Somerville, NJ), which were passed half-thickness through the nonmucosal part of the graft. Special emphasis was given to avoid any exposed suture material that could cause ocular irritation. During suturing, the graft was trimmed to the required size, and at the end of the procedure, the final dimensions of the graft were recorded.
Free Tarsal Graft Harvesting and Grafting Local anesthetic (2% lidocaine and adrenaline 1:200 000) was injected beneath the pretarsal upper eyelid skin before eversion of the eyelid using a Desmarres retractor for subconjunctival injection of additional anesthetic above the superior tarsal margin. The tarsus was incised 4 to 5 mm above the lid margin, parallel to the eyelid margin, with the length of the horizontal incision up to 16 mm, depending on the available upper eyelid tarsus. This was followed by 2 vertical incisions at each end of the horizontal incision, towards the upper border of the tarsus. The graft was then dissected from the loosely attached levator aponeurosis, Müller’s muscle, and conjunctiva. The donor site was not sutured. The graft was then wrapped in moistened gauze. Suturing of the FTG to the
Figure 1. A, Left upper lid cicatricial entropion. B, Upper lid marginal rotation is performed, and the hard palate graft is prepared (arrow). C, The graft is sutured to replace the upper lid posterior lamella.
490
Leibovitch et al 䡠 Hard Palate and Free Tarsal Grafts in Upper Lid Surgery Table 1. Patients’ Characteristics, Postoperative Complications, and Follow-up in Cases Operated Using Hard Palate Grafts (HPGs) in Upper Lid Surgery Patient*
Gender/Age (yrs)
Upper Lid Pathology
Upper Lid Surgery Combined with HPG
1 2
M/70 M/63
Cicatricial entropion SCC
3
M/42
Cicatricial entropion
4
F/81
BCC
5
M/68
BCC
6
F/83
BCC
7 8 9
F/53 M/40 M/62
Cicatricial entropion Cicatricial entropion Cicatricial entropion
10
M/63
MCC
Surgical excision ⫹ rotation flap
11
F/90
BCC
12
F/84
BCC
Surgical excision ⫹ advancement flap and FTSG Surgical excision ⫹ advancement flap
13
M/67
Cicatricial entropion
14 15
F/53 F/72
Cicatricial entropion Cicatricial entropion
Marginal lid rotation Surgical excision ⫹ advancement flap and FTSG Marginal lid rotation Surgical excision ⫹ advancement flap and FTSG Surgical excision ⫹ periosteal flaps ⫹ advancement flap Surgical excision, Cutler–Beard flap and FTSG Marginal lid rotation Marginal lid rotation Marginal lid rotation
Marginal lid rotation ⫹ buccal mucosa graft Marginal lid rotation Marginal lid rotation ⫹ buccal mucosa graft
Postoperative Complications
Final Outcome of Complications
Follow-up Period (mos)
Ocular discomfort Ocular discomfort
Complete resolution Complete resolution
11 39
Transient epithelial keratopathy Upper lid retraction (1 mm)
Complete resolution
51
1-mm retraction
21
None
NA
Donor site bleeding
Bleeding controlled with packing
None None Partial graft dehiscence
NA NA Good final contour after surgical repair Good final contour after surgical repair of dehiscence Lateral eyelid peaking
Partial graft dehiscence, mild ocular discomfort, mild epithelial keratopathy Upper lid retraction (1.5 mm) Necrosis of skin flap at 3 mos, required removal of HPG and use of pericranial forehead flap None None None
Exenteration due to tumor recurrence
10 9 22 9 7 8 12 14
NA
10
NA NA
12 10
BCC ⫽ basal cell carcinoma; F ⫽ female; FTSG ⫽ full-thickness skin graft; M ⫽ male; MCC ⫽ Merkel cell carcinoma; NA ⫽ not applicable; SCC ⫽ squamous cell carcinoma. *Cases 1–11 were operated at the Royal Adelaide Hospital; cases 12–15 were operated at the Queen Victoria Hospital.
upper lid defect was done with interrupted buried 6-0 Vicryl sutures, which were passed half-thickness through the nonmucosal part of the graft, to avoid any exposed suture material that could cause ocular irritation. During suturing, the graft was trimmed to the required size, and at the end of the procedure, the final dimensions of the graft were recorded. All patients were treated postoperatively with topical antibiotics for 2 weeks. The postoperative follow-up schedule was generally 1 week and 1, 3, 6, and 6 months thereafter. Patients who could not attend at least the first 6 months’ follow-up period were excluded from the study. Postoperative evaluation included upper eyelid contour and position, reconstructed upper eyelid viability, evaluation of ocular discomfort, and keratopathy. The graft dimensions were measured after 6 months in all cases where the operated upper lid could be everted in the outpatient clinic. All cases treated with FTGs were assessed for donor site complications, including ocular discomfort, keratopathy, upper eyelid entropion, contour abnormalities, retraction, and ptosis. The hard palate was assessed regularly in all cases where HPGs were used.
Results There were 31 patients (mean age, 66⫾12 years [range, 40 –90]) who underwent upper lid surgery. Fifteen patients (7 female, 8 male; mean age, 66⫾15) were treated using an autogenous HPG, and 16 patients (9 female, 7 male; mean age, 65⫾9) were treated using an autogenous FTG. The indications for surgery in the HPG group included upper lid defect reconstruction after tumor excision using Mohs’ surgery (7 patients [47%]) and upper lid cicatricial entropion (8 patients [53%]) (Fig 1, Table 1). The indications for surgery in the FTG group included upper lid defect reconstruction after tumor excision using Mohs’ surgery (10 patients [62.5%]), upper lid cicatricial entropion (4 patients [25%]), and upper lid retraction from thyroid eye disease (2 patients [12.5%]) (Table 2). The reasons for choosing HPG as the preferred grafting material included insufficient height of the contralateral tarsal plate to enable harvesting of adequate FTG with preservation of 4 mm of residual tarsus (cases 1, 3, 4, 6 –9, 12–15), avoiding morbidity to the contralateral-only eye (case 2), and patient’s preference (cases 5, 10, 11).
491
Ophthalmology Volume 113, Number 3, March 2006 Table 2. Patients’ Characteristics, Postoperative Complications, and Follow-up in Cases Operated Using Free Tarsal Grafts (FTGs) in Upper Lid Surgery Patient*
Gender/Age (yrs)
Upper Lid Pathology
1 2
M/65 F/70
Retraction (TED) Entropion
Scar release Marginal lid rotation
3 4
F/51 M/69
Entropion SCC
5
F/79
6
F/64
7 8
F/64 M/72
9
F/72
BCC
Marginal lid rotation Surgical excision ⫹ pericranial flap and FTSG Surgical excision ⫹ advancement flap Surgical excision ⫹ advancement flap and FTSG Marginal lid rotation Surgical excision ⫹ Cutler–Beard flap Surgical excision ⫹ advancement flap
10
F/53
SCC
11 12
F/55 M/63
Retraction (TED) BCC
13
M/66
SCC
14
M/70
BCC
15
F/73
BCC
16
M/75
Sebaceous carcinoma BCC Entropion BCC
Entropion
Upper Lid Surgery Combined with FTG
Surgical excision ⫹ Cutler–Beard flap Canthal sling Surgical excision ⫹ Cutler–Beard flap Surgical excision ⫹ Cutler–Beard flap Surgical excision ⫹ anterior lamellar advancement Surgical excision ⫹ transposition flap Marginal lid rotation
Postoperative Complications
Final Outcome of Complications
Follow-up Period (mos)
None Mild upper lid retraction and central peak in donor eyelid None None
NA Good contour in donor eyelid after division of band NA NA
12 6 11 8
None
NA
15
None
NA
14
None None
NA NA
10 11
Mild upper lid retraction and central peak in donor eyelid None
Good contour in donor eyelid after division of band NA
12
Mild ocular discomfort None
Complete resolution NA
21 16
None
NA
18
None
NA
18
None
NA
22
None
NA
8
14
BCC ⫽ basal cell carcinoma; F ⫽ female; FTSG ⫽ full-thickness skin graft; M ⫽ male; NA ⫽ not applicable; SCC ⫽ squamous cell carcinoma; TED ⫽ thyroid eye disease. *Cases 1– 4 were operated at the Queen Victoria Hospital; cases 5–16 were operated at the Royal Adelaide Hospital.
Table 3. Initial and Final Graft Dimensions in Cases Operated Using Hard Palate Grafts (HPGs) in Upper Lid Surgery
Patient
Initial Graft Dimensions Used (Height*Width/Area) (mm*mm/mm2)
Graft Dimensions on Last Follow-up (Height*Width/Area) (mm*mm/mm2)
% Decrease in Graft Dimensions (Height/Width/Area) (mm/mm/mm2)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
7ⴱ23/91 6ⴱ20/120 6ⴱ18/108 7ⴱ16/112 6ⴱ20/120 7ⴱ21/147 8ⴱ19/152 NA 6ⴱ15/90 NA NA 6ⴱ14/84 7ⴱ17/119 7ⴱ14/98 7ⴱ15/105
4ⴱ18/72 5ⴱ18/90 6ⴱ16/96 5ⴱ14/70 NA 5ⴱ15/75 6ⴱ16/96 NA 4ⴱ13/52 NA NA 5ⴱ10/50 NA 7ⴱ14/98 NA
43/22/21 17/10/25 0/11/11 29/12/37 NA 29/29/49 25/16/37 NA 33/13/42 NA NA 17/29/40 NA 0/0/0 NA
NA ⫽ data not available.
492
Leibovitch et al 䡠 Hard Palate and Free Tarsal Grafts in Upper Lid Surgery Table 4. Initial and Final Graft Dimensions in Cases Operated Using Free Tarsal Grafts (FTGs) in Upper Lid Surgery
Patient
Initial Graft Dimensions Used (Height*Width/Area) (mm*mm/mm2)
Graft Dimensions on Last Follow-up (Height*Width/Area) (mm*mm/mm2)
% Decrease in Graft Dimensions (Height/Width/Area) (mm/mm/mm2)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
5ⴱ14/70 4ⴱ15/60 4ⴱ14/56 4ⴱ15/60 5ⴱ15/75 4ⴱ11/44 4ⴱ14/56 4ⴱ14/56 4ⴱ10/40 4ⴱ15/60 4ⴱ13/52 4ⴱ14/56 4ⴱ16/64 4ⴱ10/40 4ⴱ9/36 4ⴱ14/56
4ⴱ11/44 3ⴱ12/36 4ⴱ12/48 NA 4ⴱ13/52 3ⴱ9/36 NA NA 4ⴱ9/36 NA 4ⴱ12/48 2ⴱ9/18 NA 4ⴱ10/40 3ⴱ9/27 3ⴱ10/30
20/21/37 25/20/40 0/14/14 NA 20/13/31 25/18/18 NA NA 0/10/10 NA 0/8/8 50/36/68 NA 0/0/0 25/0/25 25/29/46
NA ⫽ data not available.
Hard palate graft dimensions were recorded in 12 cases. The average HPG dimensions were 6.7⫾0.6 mm in height and 17.7⫾3.0 mm in width. The average graft area was 118⫾24.3 mm2 (Table 3). All FTGs were harvested from the contralateral upper lid. The average graft dimensions were 4.1⫾0.3 mm in height and 13.2⫾2.0 mm in width. The average FTG area was 54.6⫾10.5 mm2 (Table 4). The complications recorded in patients treated with HPG during a mean follow-up period of 16.0⫾16 months (range, 7–51) included ocular irritation or discomfort in 3 patients (20%), corneal edema or transient keratopathy in 2 (13.4%), partial graft dehiscence in 2 (13.4%) (Fig 2), upper lid retraction in 2 (13.4%)
(Figs 3, 4), and necrosis of the overlying skin flap in 1 (6.7%) (Table 1). There were no significant complications in 8 cases (53.3%). No cases of HPG hypersecretion were recorded. Donor site complications included only one case (6.2%) of excessive bleeding from the hard palate site in the recovery room, which required packing. There were no significant complications that were recorded in the patients treated with FTGs during a mean follow-up period of 13.5⫾5 months (range, 6 –22) (Table 2). Only one patient (6.2%) had moderate ocular discomfort, which resolved after 4 weeks. The donor (contralateral) upper lid complications were 2 cases (12.4%) of mild upper lid retraction and central peaking from a
Figure 2. Dehiscence of the hard palate graft from the lower edge of the tarsal plate.
493
Ophthalmology Volume 113, Number 3, March 2006
Figure 3. A, Left upper lid defect after basal cell carcinoma excision. B, Postoperative result at 3 months shows lateral upper lid retraction.
fibrous band. Division of this band resulted in significant contour improvement. The final dimensions of the HPG were recorded in 9 patients (Table 3). The average graft height was 5.2⫾1.0 mm, and the width was 15.2⫾0.7 mm. The average graft area was 78⫾19 mm2, which equals a 36% decrease in average HPG area compared with initial grafting. The vertical graft height decreased on average by 24% (1.5 mm) compared with the initial graft placed. The final dimensions of the FTG were recorded in 11 patients (Table 4). The average graft height was 3.5⫾0.7 mm, and the width was 10.5⫾1.5 mm. The average graft area was 37⫾11 mm2, which equals a 27% decrease in average FTG area compared with initial grafting. The vertical graft height decreased on average by 17% (0.6 mm) compared with the initial graft placed.
494
Discussion The successful use of HPGs for eyelid reconstruction after tumor excision was first described by Siegel in 1985.3 Since then, HPGs have been used for different eyelid pathologies such as cicatricial entropion and trichiasis,4,7,8 eyelid retraction, and facial palsy.5–7,9 In one of the earliest and largest series,7 Cohen and Shorr reported their experience in using HPGs for eyelid reconstruction in 25 cases. Only 6 (24%) were upper lid cases. In a more recent series by Wearne et al,9 the authors presented 102 eyelids operated using autogenous HPGs, but all were performed for the treatment
Leibovitch et al 䡠 Hard Palate and Free Tarsal Grafts in Upper Lid Surgery
Figure 4. A, Left upper lid defect after basal cell carcinoma excision. B, Postoperative result at 3 months shows a mild lateral upper lid peak.
of lower eyelid displacement. To the best of our knowledge, our series of outcomes of upper lid surgery is the largest to date. The decision whether to use FTGs or HPGs is made individually for each case and is based on various factors. An autogenous FTG is an excellent material that provides flexibility, rigidity and contour.1 However, there are 2 important considerations when using FTGs: (1) the amount of available material is limited by the size of the donor tarsal plate and the requirement to leave at least 4 mm of the tarsal plate inferiorly to minimize complications,1,2 and (2) there are possible donor site complications. Contour peaking, which was noted in 2 donor upper lids in our series, was the only significant complication after FTG harvesting. In both
cases, it was attributed to a horizontal fibrotic band at the superior border of the remaining tarsus. Excision of the band resulted in significant contour improvement. From our experience, harvesting of FTGs does not lead to significant donor site morbidity; however, subsequent procedures to alter the position of the donor eyelid may be less predictable.2 The presence of ptosis or retraction in the proposed donor eyelid should lead the surgeon to consider correction before graft harvesting, or an alternative graft material. An HPG provides stiffness, thickness, and flexibility to the upper lid, as well as a mucosal lining.3,4,7 It is readily available and is not associated with any morbidity to the contralateral eyelid. In all previous series in which HPGs
495
Ophthalmology Volume 113, Number 3, March 2006 were used in lower lid surgery, the authors reported only minimal ocular and eyelid complications. Siegel3 reported no complications in his first series of 11 patients, and so did Kersten et al6 (25 lower eyelids) and Cohen and Shorr7 (19 lower lids and 6 upper lids). In Wearne et al’s more recent series9 (102 lower lids), 85% of cases achieved satisfactory lid position, and the rest had inadequate lengthening or significant recurrence of lower lid displacement. The complications recorded in our patients included ocular irritation or discomfort (20%), corneal edema or transient keratopathy (13.3%), partial graft dehiscence (13.3%), upper lid retraction (13.3%), and necrosis of the overlying skin flap (6.7%) (Table 1). There were no significant complications in 8 cases (53.3%). This rate of complications is higher than what was reported in previous series and could be related to placing the graft in the upper lid. The ocular irritation and the keratopathy could result from direct contact between the ocular surface and the keratinized epithelium of the HPG.6,7 In fact, the presence of keratinized epithelium has discouraged some clinicians from using HPGs in the upper lid due to the possible effects on the cornea. However, this side effect was temporary in all our cases and resolved after several weeks, possibly in correlation with the gradual metaplasia of the epithelium to a nonkeratinzed type.6,7 Hard palate graft dehiscence was not reported in previous series dealing with lower lid grafting. In our series, dehiscence occurred early and, possibly, due to infection or altered healing. One of the 2 patients with partial graft dehiscence was diagnosed with Stevens–Johnson syndrome and was treated with systemic steroids, which could have an effect on healing. Local postoperative infection is an important factor that can result in graft dehiscence, and although there were no obvious signs of infection in any of our cases, it could have been mild and not easily detected. Donor site complications after HPG harvesting have been reported in several studies. The most common complication is hemorrhage (10% of cases), which occurs during the immediate postoperative period or even later (7–12 days).7,9 In most cases, the hemorrhage is controlled with local pressure and systemic antibiotics. In our series, only one case of hemorrhage occurred (6.7%), which responded to local packing. Other less common complications previously reported are development of oronasal fistulas, and candidal infection.10 These complications were not encountered in our series. Another uncommon complication that was reported in HPG cases is mucus production from minor salivary glands located in the graft.11 These cases were treated successfully with cryotherapy or graft removal. Although we have had several similar cases of hypersecretion from HPGs that required graft removal, all occurred in lower lid grafts. On histology, these grafts showed mucus-secreting glands. This complication did not occur in any of our upper lid grafts. It is generally reported that the incidence of postoperative HPG contraction is low. In Siegel’s first report on HPGs, he stated that they contract minimally, and advised oversizing the graft by 1 to 2 mm.3 Cohen and Shorr7 measured 17 grafts and found that only 3 demonstrated contraction, and that no graft contracted ⬎0.5 mm. They did not specify their exact follow-up period. Other authors
496
reported contraction of ⱕ15%.5,6,12 In a recent study, Sullivan and Dailey compared graft contraction of acellular dermis with that of hard palate mucosa in lower eyelid surgery.13 They found that HPG vertical height contraction ranged from 0% to 44% (mean, 16%). In our study, there was significant contraction in HPG vertical height, with an average of 24% (range, 0%– 43%). This degree of contraction is higher than in previous reports, but is similar to Sullivan and Dailey’s findings. Interestingly, we also found a significant contraction in FTG size (27% in graft area and 17% in vertical height), but the exact reason for these findings is not clear. The small study groups, retrospective design, and lack of randomization preclude definite conclusions; however, we believe that oversizing the HPG may have a role in minimizing the effects of graft contraction on the final outcome. In conclusion, although HPGs are considered excellent material for posterior lamella replacement in lower lid surgery, they may be associated with a higher rate of complications in upper lid surgery and when compared with FTGs. Ocular irritation and keratopathy are temporary and usually resolve after several weeks.
References 1. Stephenson CM, Brown BZ. The use of tarsus as a free autogenous graft in eyelid surgery. Ophthal Plast Reconstr Surg 1985;1:43–50. 2. Leibovitch I, Selva D, Davis G, Ghabrial R. Donor site morbidity in free tarsal grafts. Am J Ophthalmol 2004;138:430 –3. 3. Siegel RJ. Palatal grafts for eyelid reconstruction. Plast Reconstr Surg 1985;76:411– 4. 4. Silver B. The use of mucous membrane from the hard palate in the treatment of trichiasis and cicatricial entropion. Ophthal Plast Reconstr Surg 1986;2:129 –31. 5. Bartley GB, Kay PP. Posterior lamellar eyelid reconstruction with a hard palate mucosal graft. Am J Ophthalmol 1989;107: 609 –12. 6. Kersten RC, Kulwin DR, Levartovsky S, et al. Management of lower-lid retraction with hard-palate mucosa grafting. Arch Ophthalmol 1990;108:1339 – 43. 7. Cohen MS, Shorr N. Eyelid reconstruction with hard palate mucosa grafts. Ophthal Plast Reconstr Surg 1992;8:183–95. 8. Mannor GE, Mathers WD, Wolfley DE, Martinez JA. Hardpalate mucosa graft in Stevens-Johnson syndrome. Am J Ophthalmol 1994;118:786 –91. 9. Wearne MJ, Sandy C, Rose GE, et al. Autogenous hard palate mucosa: the ideal lower eyelid spacer? Br J Ophthalmol 2001; 85:1183–7. 10. Kim JW, Kikkawa DO, Lemke BN. Donor site complications of hard palate mucosal grafting. Ophthal Plast Reconstr Surg 1997;13:36 –9. 11. Pelletier CR, Jordan DR, Brownstein S, Li S. An unusual complication associated with hard palate mucosal grafts: presumed minor salivary gland secretion. Ophthal Plast Reconstr Surg 1998;14:256 – 60. 12. Beatty RL, Harris G, Bauman GR, Mills MP. Intraoral palatal mucosal graft harvest. Ophthal Plast Reconstr Surg 1993;9: 120 – 4. 13. Sullivan SA, Dailey RA. Graft contraction: a comparison of acellular dermis versus hard palate mucosa in lower eyelid surgery. Ophthal Plast Reconstr Surg 2003;19:14 –24.