Subperiosteal Midface Lift with or without
a Hard Palate Mucosal Graft for Correction
of Lower Eyelid Retraction
Guy J. Ben Simon, MD, Seongmu Lee, BS, Robert M. Schwarcz, MD,
John D. McCann, MD, PhD, Robert A. Goldberg, MD
placement of a hard palate mucosal graft (HPMG) in patients with lower eyelid retraction.
Retrospective, comparative, interventional case series.
Thirty-four patients with lower eyelid retractions who underwent surgery at the Jules Stein Eye
Institute in a 5-year period.
Medical record review of all patients who underwent surgery for lower eyelid retraction by a
subperiosteal midface lift with or without an HPMG. Preoperative and postoperative digital photographs were
taken in all patients.
Main Outcome Measures:
Change in margin reflex distance 2 (MRD2), measured from the pupillary margin
to the upper margin of the lower eyelid; patient discomfort; and surgical complications.
Thirty-four patients (20 female; mean age, 64 years) participated in the study; 11 underwent
bilateral surgery, with overall 43 surgeries performed. Eighteen patients (42%) had lower eyelid retraction
secondary to previous transcutaneous lower eyelid blepharoplasty. Postoperatively, patients attained a
better lower eyelid position, with improvement of lower eyelid height of 1.4 mm (P?0.001, 1-sample t test).
Patients operated using an HPMG (12 surgeries) achieved a greater reduction in MRD2postoperatively as
compared with patients operated by subperiosteal midface lift alone (31 surgeries; 2.2 mm vs. 1.1 mm,
respectively; P ? 0.02, Wilcoxon Mann–Whitney). One patient needed reoperation secondary to symptom-
atic lower eyelid retraction postoperatively.
The subperiosteal midface lift is effective in correction of lower eyelid retraction of various
causes. The use of an HPMG spacer may enhance surgical outcomes and results in a better lower eyelid position.
Ophthalmology 2006;113:1869–1873 © 2006 by the American Academy of Ophthalmology.
To compare functional and surgical outcomes of a subperiosteal midface lift with and without the
Lower eyelid retraction is a relatively uncommon condition
that may occur in association with various orbital or sys-
temic diseases and eyelid surgery.1Thyroid-related or-
bitopathy can manifest as upper and lower eyelid retrac-
tions, which give the typical stare appearance along with
widening of the vertical palpebral fissure. It is believed that
overactivity of the sympathetically innervated Müller’s
muscle equivalent may be the actual mechanism for lower
eyelid retraction. Postoperative transcutaneous lower eyelid
blepharoplasty with excess removal of skin and orbicularis
could result in a vertical shortage of anterior lamella or
middle and posterior lamella tethering.2,3It also may com-
plicate chronic facial nerve palsy or occur with no under-
lying pathology.4,5Clinically, ocular discomfort, lagoph-
thalmos, and exposure may ensue.
Surgeons differ in their approach to the surgical repair of
lower eyelid retraction with or without midface descent. Many
surgical techniques have been described. They can involve
relatively simple maneuvers, such as a full-thickness skin graft
or myocutaneous switch flaps, or more complicated surgeries,
such as middle and posterior lamella lengthening or midface
lifting, all with or without spacer material.6–13For the latter,
different autogenous graft materials have been used, including
tarsoconjunctiva,14,15hard palate,16–18buccal membrane,6,16
ear or conchal cartilage,19,20autogenous dermis skin,21or
biosynthetic materials such as acellular human dermis (Al-
loDerm, LifeCell Corp., The Woodlands, TX), polytetra-
fluoroethylene,22and porous polyethylene.23,24To date,
controversy exists about the optimal surgical correction and
long-term outcomes of each procedure.
In our institution, the subperiosteal midface lift generally
is performed with securing of the subperiosteally dissected
midfacial tissue to the inferior orbital arcus marginalis with
or without a hard palate mucosal graft (HPMG). The pur-
pose of the current study is to compare in a retrospective
fashion efficacies of this procedure performed alone versus
with an HPMG.
Originally received: October 11, 2005.
Accepted: May 12, 2006.
From the Jules Stein Eye Institute and Department of Ophthalmology,
David Geffen School of Medicine at UCLA, Los Angeles, California.
Correspondence to Guy J. Ben Simon, MD, Goldschleger Eye Institute,
Manuscript no. 2005-976.
© 2006 by the American Academy of Ophthalmology
Published by Elsevier Inc.
ISSN 0161-6420/06/$–see front matter
Materials and Methods
An electronic medical record review of all patients with lower
eyelid retraction referred to the orbitofacial unit of the Jules Stein
Eye Institute between January 1999 and December 2004 was
performed. Patients were included only if lower eyelid retraction
was secondary to thyroid eye disease or occurred after blepharo-
plasty. Patients were excluded if a traumatic eyelid and orbital
injury were evident, because a more complex mechanism of eyelid
retraction exists in these cases. The study was approved by the
local institutional review board.
All patients underwent comprehensive eye examinations, in-
cluding determination of visual acuity (VA) and intraocular pres-
sure (IOP) and slit-lamp examinations. Preoperative and postop-
erative digital photographs were obtained in primary gaze. Margin
reflex distance 2 (MRD2) was defined as the distance of the
pupillary light reflex from the superior edge of the inferior eyelid
and was measured in millimeters preoperatively and 6 to 12
months postoperatively. All patients had a minimal follow-up time
of 6 months. Patients were evaluated specifically for the presence
of ocular discomfort, dry eyes, use of topical lubricants, and
subjective cosmetic appearance. Additional measurements were
performed by an independent masked observer based on digital
images using a computer program.
All surgeries were performed by 2 of the authors (JDM, RAG),
and the decision to use an HPMG was made based on individual
Subperiosteal Midface Lift. The lower eyelid was infiltrated with
a mixture of lidocaine and marcaine with 1:100 000 adrenaline.
Injections were performed in the lower eyelid fornix toward the
inferior arcus marginalis as well as the midface full thickness.
The lower eyelid then was retracted using a Desmarres retrac-
tor, and an inferior fornix incision was made using monopolar
cautery using a Colorado needle. In all cases, lateral canthotomy
and inferior cantholysis were performed and a lateral tarsal strip
As dissection was carried inferiorly, an incision was made in
the periosteum, leaving a cuff anteriorly on the orbital rim. A no.
15 blade was used to make a periosteotomy, continuing a subperi-
osteal dissection with a blunt periosteal elevator. When needed,
middle lamella scar lysis was done. Care was taken not to sever the
inferior orbital nerve while dissection was performed isolating it.
The levator labii superioris ala nasi muscle was detached from its
origin inferior to the inferomedial orbital rim.
Three to 5 sutures (Prolene [Ethicon, Inc., Somerville, NJ] or
PDS [Ethicon]) were placed in a mattress-type fashion from the
inferior orbital rim to the periosteum and deep fibrofatty tissue of
the midface after a periosteotomy was performed at the level of the
nasal alae to allow for a release of the midfacial tissues to be lifted.
The sutures were tied down to the inferior periosteal cuff; the
lateral tarsal strip was then attached to the lateral orbital rim using
a 5/0 Vicryl (Ethicon) or PDS suture on a half-circle needle, and
the lateral tarsal angle was reformed. The conjunctival incision
was left unsutured.
The lower eyelid was placed on 3 Frost sutures that were taped
to the forehead or sutured down to the eyebrow. The eye was
patched for 5 days when the Frost sutures were removed.
Subperiosteal Midface Lift Using a Hard Palate Mucosal
Graft. Surgery was performed in a similar fashion, with the
HPMG harvested from the hard palate lateral to the midline raphe
before securing the lateral tarsal strip. The length of the HPMG
harvested was measured according to the height of the lower
The HPMG was thinned, and fat tissue was removed using sharp
dissection. It was sutured to both the inferior conjunctiva–retractors
complex and the inferior tarsal edge.
At the last stage, the lateral tarsal strip was sutured to the lateral
orbital rim and the lateral canthal angle was reformed. The eyelid
was placed on Frost sutures, and the eye was patched for 5 days.
When patients required bilateral surgery, it was performed as a
staged procedure to avoid patching both eyes for 5 days, and eyelid
pull-up sutures were required to ensure wound healing at the
The paired-samples t test was used to evaluate preoperative and
postoperative data such as MRD2, VA, and IOP. Conversion of
VA to logarithm of the minimum angle of resolution was per-
formed. The 1-sample t test was used to compare ? values of these
parameters to zero value. An independent-samples t test was
performed to compare these numerical variables between two
groups of patients—subperiosteal midface lift with HPMG and
subperiosteal midface lift without HPMG. One-way analysis of
variance (ANOVA) was used to calculate the difference in ?
MRD2between different diagnoses. The nonparametric chi-square
analysis and Fisher exact test with cross-tabulations were used to
calculate proportions of patients achieving improvement with sur-
gery in both groups; improvement in dry eyes, using ocular lubri-
cants; and presence of punctate epithelial keratopathy as evidence
of ocular exposure and lagophthalmos. Kaplan–Meier survival
analysis was used to calculate longevity of surgery in both groups.
Statistical analysis was performed using Excel 2003 (Microsoft
Corp., Redmond, WA) and SPSS (version 13.0, SPSS, Inc., Chi-
Thirty-four patients (20 female; mean age, 64 years) were treated
with a midface lift for lower eyelid retraction between January
1999 and December 2004. Demographics of the study population
are summarized in Table 1. Eleven patients underwent bilateral
surgery; overall, 43 surgeries were performed.
Most cases (18 [42%]) were diagnosed with lower eyelid
retraction secondary to lower blepharoplasty, followed by midface
descent with lagophthalmos (11 cases [26%]), thyroid-related or-
bitopathy, facial nerve palsy, and anterior cicatricial ectropion.
Thirty-one cases (72%) underwent a transconjunctival midface
lift, and 12 (28%) underwent a transconjunctival midface lift with
Table 1. Demographics of 34 Patients (43 Surgeries) Who
Were Operated for Lower Eyelid Retraction in a 5-Year Period
Age (yrs) (range)
Facial nerve palsy
Midface lift with hard palate mucosal graft
Volume 113, Number 10, October 2006
Postoperatively, patients attained a better lower eyelid position,
with improvement of lower eyelid height of 1.4 mm (P?0.001,
1-sample t test). Most patients were pleased with the surgical
outcome (both functional and cosmetic results) (Table 2, Fig 1).
Only 3 cases (7%) had mild superficial punctate keratopathy
postoperatively. Mean follow-up time was 13 months.
Visual acuity and IOP remained unchanged after surgery; in-
terestingly, lagophthalmos decreased by only 0.2 mm, and this was
not statistically significant relative to the baseline measurement.
Comparative analysis between patients who underwent a mid-
face lift and patients who underwent a midface lift with an HPMG
showed that although patients were similar in all baseline charac-
teristics such as age and diagnosis and extent of lower eyelid
retraction, patients operated using an HPMG achieved a greater
reduction in MRD2postoperatively (2.2 mm vs. 1.1 mm; P ? 0.02,
Wilcoxon Mann–Whitney) (Fig 2). Subgroup analysis showed
similar reductions in MRD2in patients with different preoperative
diagnoses (P ? 0.79, 1-way ANOVA); similarly, multiple com-
parisons within each preoperative diagnosis using a nonparametric
Wilcoxon Mann–Whitney test showed no difference in ? MRD2.
Postoperatively, 9 cases (20.9%) had mild residual lower eyelid
retraction with lagophthalmos; 1 of these patients was reoperated.
Corneal abrasion was noted in 2 cases; both were operated using
an HPMG, and corneal abrasion resolved with topical treatment.
One patient achieved a higher than normal lower eyelid position
(overcorrection). Similar complication rates were observed in the
A subperiosteal midface lift with or without an HPMG is
effective in correction of lower eyelid retraction. Surgery is
successful in achieving better lower eyelid position and
improving eyelid asymmetry and lagophthalmos. In the
current study, better results were found using an HPMG, but
Table 2. Preoperative and Postoperative Data for 34 Patients
(43 Surgeries) Who Were Operated for Lower Eyelid Retraction
in a 5-Year Period
No. of patients
IOP ? intraocular pressure; MRD2? marginal reflex distance 2, measured
from the pupillary light reflex to the upper border of the lower eyelid in
primary position; NS ? not significant.
*Calculated using paired-samples t test.
†Fisher exact test.
Figure 1. A 43-year-old male with facial nerve palsy on the left side (A) before and (B) after a midface lift using a hard palate mucosal graft. Note marked
improvement in left lower eyelid position, with good symmetry and no residual retraction.
Ben Simon et al ? Subperiosteal Midface Lift with or without a Hard Palate Graft
this procedure may be associated with a longer operation
and transient patient discomfort.
Pathophysiology of lower eyelid retraction may involve
middle and posterior lamella tethering, midface descent, and
lateral canthal tendon laxity. It is imperative to identify and
address these conditions. Different authors achieve similar
improvements in lower eyelid position with resolution of
scleral show with a midface lift and spacer graft, with
numbers ranging from 1.6 to 2.5 mm23,25,26; failure to
improve lower eyelid position is seen in up to 25% of
patients. Twenty-one percent of our patients had mild re-
sidual asymptomatic eyelid retraction, with only 1 patient
A midface lift with or without a spacer graft may be a
relatively robust surgery for an allegedly minor problem;
however, many authors believe that a larger surgery is
associated with better long-term results. It has been shown
that addressing more than one element in the pathophysiol-
ogy of lower eyelid retraction may result in a better surgical
Shorr and Fallor were the first to describe our technique
of subperiosteal midface lifting, which was specified as the
Madame Butterfly procedure.30Recently, Li et al23pub-
lished their results using this technique comparing an
HPMG and an acellular human dermis graft (AlloDerm).
They compared 35 patients undergoing AlloDerm grafting
with 25 patients undergoing an HPMG and found similar
improvement in eyelid height in both groups. In general it is
accepted that hard palate mucosa is a better graft material
because it tends to retract less than other autogenous or
autologous materials. However, HPMG harvesting may be
associated with bleeding, sensory lesions, and patient dis-
comfort. These complications can be reduced by meticulous
surgical technique, paramedian harvesting, and postopera-
tive care, such as compression using a mouth guard.16,31We
also had 2 cases of transient corneal abrasions, both in
patients with an HPMG spacer.
Lower eyelid retraction with scleral show can manifest
upper eyelid blepharoptosis.32It is proposed that disinser-
tion of the levator aponeurosis from the tarsus enhances
contraction of the superior rectus muscle through the inter-
muscular fascia, resulting in upward rotation of the globe.
As a result, additional contraction of the inferior rectus
muscle is induced to maintain a horizontal visual axis with
the head in primary gaze position, leading to pulling on the
inferior suspensory ligament of Lockwood and the capsu-
lopalpebral fascia. Both result in a dynamically lower
scleral show. Surgical advancement of the levator aponeu-
rosis can correct this problem.32We recommend assessment
of all patients preoperatively for upper eyelid ptosis to rule
out this condition.
We achieved an average of 2.2 mm of improvement of
the lower eyelid position using an HPMG; this is slightly
better than a free tarsoconjunctival graft without a midface
lift, for which an improvement of 1.6 to 2.0 mm was
reported.14,15Porous polyethylene was found to improve
eyelid height by 1 to 1.5 mm.24Different studies report
similar extents of improvement using other graft materials
such as hard palate mucosa.28Acellular dermis contracts
significantly more than hard palate mucosa when used as a
lower eyelid spacer graft, although both materials were
found to be successful in treating lower eyelid retraction
with a subperiosteal midface lift.29
An interesting study compared the use of a donor scleral
graft with the use of partial tenotomy of the anterior part of
lower eyelid retractors with adjunctive antimetabolites in thy-
roid eye disease.28The authors report better results with the
donor scleral graft, with 25% of patients in the tenotomy–
Figure 2. Box plot showing postoperative change in margin reflex distance 2 (MRD2) in 34 patients (43 surgeries) with lower eyelid retractions.
Comparative analysis between patients operated using a midface lift (31 cases) and patients operated using a midface lift and hard palate mucosal graft
(HPG) (12 cases). *P ? 0.02 (Wilcoxon Mann–Whitney).
Volume 113, Number 10, October 2006
antimetabolites group requiring additional surgery using a
In conclusion, a subperiosteal midface lift is an effective
procedure in lower eyelid elevation, and the use of a spacer
material such as hard palate mucosa may enhance surgical
1. Cohen M, Lessell S. Retraction of the lower eyelid. Neurology
2. Morax S. Complications of blepharoplasty. J Fr Ophtalmol
3. Patipa M. Transblepharoplasty lower eyelid and midface
rejuvenation: part I. Avoiding complications by utilizing les-
sons learned from the treatment of complications. Plast Re-
constr Surg 2004;15:1459–68.
4. Elner V, Mauffray RO, Fante RG, et al. Comprehensive mid-
facial elevation for ocular complications of facial nerve palsy.
Arch Facial Plast Surg 2003;5:427–33.
5. Olver J. Raising the suborbicularis oculi fat (SOOF): its role in
chronic facial palsy. Br J Ophthalmol 2000;84:1401–6.
6. Sullivan S, Dailey RA. Endoscopic subperiosteal midface lift:
surgical technique with indications and outcomes. Ophthal
Plast Reconstr Surg 2002;18:319–30.
7. Frueh B, Su CS. Medial tarsal suspension: a method of ele-
vating the medial lower eyelid. Ophthal Plast Reconstr Surg
8. Kim J, Ellis DS, Stewart WB. Correction of lower eyelid
retraction by transconjunctival retractor excision and lateral
eyelid suspension. Ophthal Plast Reconstr Surg 1999;15:
9. Patel BC, Patipa M, Anderson RL, et al. Management of
postblepharoplasty lower eyelid retraction with hard palate
grafts and lateral tarsal strip. Plast Reconstr Surg 1997;99:
10. Feldman K, Putterman AM, Farber MD. Surgical treatment of
thyroid-related lower eyelid retraction: a modified approach.
Ophthal Plast Reconstr Surg 1992;8:278–86.
11. Holds J, Anderson RL, Thiese SM. Lower eyelid retraction: a
minimal incision surgical approach to retractor lysis. Ophthal-
mic Surg 1990;21:767–71.
12. Hurwitz J, Archer KF, Gruss JS. Treatment of severe lower
eyelid retraction with scleral and free skin grafts and bipedicle
orbicularis flap. Ophthalmic Surg 1990;21:167–72.
13. Hamako C, Baylis HI. Lower eyelid retraction after blepha-
roplasty. Am J Ophthalmol 1980;89:517–21.
14. Ferri M, Oestreicher JH. Treatment of post-blepharoplasty
lower lid retraction by free tarsoconjunctival grafting. Orbit
15. Gardner T, Kennerdell JS, Buerger GF. Treatment of dysthy-
roid lower lid retraction with autogenous tarsus transplants.
Ophthal Plast Reconstr Surg 1992;8:26–31.
16. Wearne M, Sandy C, Rose GE, et al. Autogenous hard palate
mucosa: the ideal lower eyelid spacer? Br J Ophthalmol 2001;
17. Cohen M, Shorr N. Eyelid reconstruction with hard palate
mucosa grafts. Ophthal Plast Reconstr Surg 1992;8:183–95.
18. Kersten R, Kulwin DR, Levartovsky S, et al. Management of
lower-lid retraction with hard-palate mucosa grafting. Arch
19. Marks M, Argenta LC, Friedman RJ, et al. Conchal cartilage
and composite grafts for correction of lower lid retraction.
Plast Reconstr Surg 1989;83:629–35.
20. Baylis H, Perman KI, Fett DR, et al. Autogenous auricular
cartilage grafting for lower eyelid retraction. Ophthal Plast
Reconstr Surg 1985;1:23–7.
21. Brock W, Bearden W, Tann T 3rd, et al. Autogenous dermis
skin grafts in lower eyelid reconstruction. Ophthal Plast Re-
constr Surg 2003;19:394–7.
22. Karesh J, Fabrega MA, Rodrigues MM, et al. Polytetrafluoroeth-
ylene as an interpositional graft material for the correction of
lower eyelid retraction. Ophthalmology 1989;96:419–23.
23. Li T, Shorr N, Goldberg RA. Comparison of the efficacy of
hard palate grafts with acellular human dermis grafts in lower
eyelid surgery. Plast Reconstr Surg 2005;116:873–8.
24. Tan J, Olver J, Wright M, et al. The use of porous polyeth-
ylene (Medpor) lower eyelid spacers in lid heightening and
stabilisation. Br J Ophthalmol 2004;88:1197–200.
25. Patel M, Shapiro MD, Spinelli HM. Combined hard palate
spacer graft, midface suspension, and lateral canthoplasty for
lower eyelid retraction: a tripartite approach. Plast Reconstr
26. Taban M, Douglas R, Li T, et al. Efficacy of “thick” acellular
human dermis (AlloDerm) for lower eyelid reconstruction:
comparison with hard palate and thin AlloDerm grafts. Arch
Facial Plast Surg 2005;7:38–44.
27. Baylis H, Nelson ER, Goldberg RA. Lower eyelid retraction
following blepharoplasty. Ophthal Plast Reconstr Surg 1992;
28. Olver J, Rose GE, Khaw PT, et al. Correction of lower eyelid
retraction in thyroid eye disease: a randomised controlled trial
of retractor tenotomy with adjuvant antimetabolite versus
scleral graft. Br J Ophthalmol 1998;82:174–80.
29. Sullivan S, 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.
30. Shorr N, Fallor MK. “Madame Butterfly” procedure: com-
bined cheek and lateral canthal suspension procedure for post-
blepharoplasty, “round eye,” and lower eyelid retraction. Oph-
thal Plast Reconstr Surg 1985;1:229–35.
31. Guyot L, Layoun W, Benso-Layoun C, et al. Hard palate
mucosal graft for posterior lamella repair. J Fr Ophtalmol
32. Matsuo K, Kondoh S, Kitazawa T, et al. Pathogenesis and
surgical correction of dynamic lower scleral show as a sign of
disinsertion of the levator aponeurosis from the tarsus. Br J
Plast Surg 2005;58:668–75.
Ben Simon et al ? Subperiosteal Midface Lift with or without a Hard Palate Graft