Acta Ophthalmologica 2014

Levator recession with adjustable sutures for

correction of upper eyelid retraction in thyroid eye
disease
Hans Olav Ueland,1* Aleksander Uchermann2* and Eyvind Rødahl1,2
1
Department of Ophthalmology, Haukeland University Hospital, Bergen, Norway
2
Department of Clinical Medicine, University of Bergen, Bergen, Norway

ABSTRACT. normalized. A tight rectus inferior

Purpose: To examine the outcome of levator muscle recession with adjustable muscle may cause secondary retraction,
sutures for correcting upper eyelid retraction in thyroid eye disease. in which there is an overshoot on
Methods: All patients treated at Haukeland University Hospital between 2001 upgaze of the affected eyelid. Increased
and 2011 were invited to participate in a follow-up examination. Preoperative proptosis may contribute to a widening
measurements were obtained from medical records. The following parameters of the palpebral fissure, affecting
were recorded: eyelid aperture, the distance from limbus to eyelid margin, mainly the position of the lower eyelid.
vertical motility of the eyelid in up- and downgaze, lagophthalmos, proptosis, Retraction of the upper eyelid causes
vertical motility of the eye, and strabismus. Symmetry and eyelid contour were exposure of the conjunctiva and cornea
with ocular discomfort, foreign body
evaluated from photographs. Results were categorized as good, acceptable or
sensation, dryness, photophobia and
unacceptable.
tearing. The symptoms become more
Results: Sixty-four out of a total of 80 patients attended the follow-up severe if lagophthalmos is present.
examination. Six were excluded from further analysis due to relapse of their Aesthetically, the patients may have a
thyroid eye disease with increased inflammation. Among the 58 remaining staring, angry or frightened appear-
patients, a good result was observed in 34 patients, an acceptable result in 19 and ance. Correction of upper eyelid retrac-
an unacceptable result in 5. Lagophthalmos was seen in 30% of the lids tion is important to improve the
preoperatively and in 7% postoperatively. A second procedure was performed in quality of life of TED patients.
13/80 patients, in 6 because of under-correction and in 7 because of ptosis. The normal position of the upper
Wound infection occurred in 2 patients. eyelid margin is approximately 1 mm
Conclusion: Levator recession with adjustable sutures can be used to correct any below the limbus. Various definitions
degree of retraction. Good results can be obtained, but the procedure is time- have been used to indicate when upper
consuming, and patients must be informed about the risk for reoperation. eyelid retraction is present, including
‘at the limbus or above’ (Bartley et al.
Key words: contour – eyelid surgery – keratopathy – lagophthalmos – symmetry
1996) and ‘more than 5 mm above the
midpupil’ (Shortt et al. 2011). Upper
eyelid retraction can be accompanied
Acta Ophthalmol. 2014: 92: 793–797 by lid lag, which is a static situation in
ª 2014 Acta Ophthalmologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
which the eyelid is higher than normal
doi: 10.1111/aos.12404 with the globe in downgaze, von Grae-
fe’s sign, which is a dynamic sign
*Equal contribution
describing the retarded descent of the
1996). Unilateral presentation is seen eyelid during movement of the globe
Introduction in approximately one of five patients from primary position to downgaze,
(Bartley et al. 1996). Usually, the and reduced levator excursion (Harvey
Thyroid eye disease (TED) is the most lateral part of the upper eyelid is more & Anderson 1981; Gaddipati & Meyer
common cause of upper eyelid retrac- retracted than the medial part (‘lateral 2008; Lelli et al. 2010).
tion (Bartley 1996), and upper eyelid flare’). During the acute phase of Histological studies of the upper
retraction is observed during the dis- hyperthyroidism, transient retraction eyelid in TED has demonstrated the
ease course in more than 90% of of the upper eyelid may occur, which presence of fat infiltration, inflamma-
affected individuals (Bartley et al. disappears when thyroid function is tory cells, fibrosis and reduced number

793
Acta Ophthalmologica 2014

of muscle cells both in the levator considered a quality improvement Results were categorized as ‘good’,
palpebrae and M€ uller0 s muscle (Grove study by the Regional Medical Ethics ‘acceptable’ and ‘unacceptable’ accord-
1981; Cockerham et al. 2002; Shih Committee of Western Norway, and ing to Mourits and Sasim (Mourits &
et al. 2006). It is thought that inflam- was approved as such by the Hauke- Sasim 1999). A ‘good’ result is defined
mation may cause changes in eyelid land University Hospital review board as the upper eyelid covering 0.5–
structures that progress towards fibro- (project no. 2010/474). The study 1.5 mm of the cornea in the 12 o’clock
sis resulting in a tight, retracted eyelid. adhered to the tenets of the Declara- position, the difference in lid aperture
There are several options for tion of Helsinki. between the two sides is <1 mm, the
treatment of upper eyelid retrac- The patients underwent a clinical eyelid contour is smooth, the lid crease
tion. Injections with botulinum toxin, examination that included photogra- is within 7–10 mm of the lid margin,
triamcinolone (Lee et al. 2013) and phy, and their hospital records were the skin folds of the right and left lid
hyaluronic acid (Mancini et al. 2011) reviewed. Before surgery, all patients are symmetrical, and the patient is
may produce temporary relief. A had been euthyroid for at least satisfied. An ‘acceptable’ result is pres-
number of techniques have been 12 months, and none had any signs of ent when the upper eyelid margin is
described for surgical treatment. Most active inflammation according to the within 0.5 mm of the limbus or covers
are variations on weakening or length- Clinical Activity Score (CAS) (Mourits no more than 2 mm of the cornea in
ening of the upper eyelid retractors et al. 1997). They were all operated by the 12 o0 clock position, the difference
(Schaefer 2007). The effect has in some the same surgeon (ER). between the right and the left side is
cases only been evaluated by a single The position of the eyelid was mea- <2 mm, and otherwise as in ‘good’
group of investigators, and frequently sured relative to the limbus with the result.
only in small series. Comparing data normal position being 1–1.5 mm below Parameters were recorded in an
from individual studies can be difficult the limbus. Retraction was considered Excel spreadsheet and analysed using
as patient features and criteria for to be present when the eyelid was the statistical package for the social
success may vary from one study to <0.5 mm below the limbus. sciences (SPSS Version 21.0; IBM Cor-
another. The surgical procedure has been poration, Armonk, NY, USA). Paired
Surgically, the main challenges are described previously (Collin & O’Don- t-test was performed to evaluate the
to avoid contour defects, under-correc- nell 1994). It was carried out under change in pre- and postoperative mea-
tion and secondary ptosis. Levator local anaesthesia using a 50/50 solution surements. p-values <0.05 were consid-
recession with adjustable sutures of marcaine 0.5% and lidocaine 1% ered significant.
addresses many of the problems that with adrenaline 0.005%. In cases with
may result in poor outcome following bilateral involvement, both upper lids
surgery for upper eyelid retraction. were operated in the same session.
Results
This includes cutting the lateral horn Briefly, following a lid crease incision A total of 80 patients, 74 women and
of the levator aponeurosis to eliminate down to the tarsal plate, the levator 6 men underwent levator recession
lateral flare, recessing both the levator aponeurosis was separated from with adjustable sutures as the primary
and M€ uller muscle and releasing the M€ uller0 s muscle. The septum was procedure for correcting upper eyelid
septum to ensure sufficient lowering of opened across the eyelid, and the retraction between 2001 and 2011. At
the lid. The sutures allow adjustments lateral horn of the levator aponeurosis follow-up, 4 were dead and 12 were
of the eyelid contour and symmetry was cut. Except in the most severe unable to attend. Among the 64
relative to the opposite eyelid. The cases, the medial horn was left intact patients that attended, 6 had suffered
adjustments can be done without the (Ceisler et al. 1995). M€ uller0 s muscle from recurrent, active inflammation
influence of local anaesthesia. The was then dissected from the underlying with worsening of their thyroid eye
positioning of the sutures restores the conjunctiva, and a vertical incision was disease after they had undergone leva-
eyelid crease. So far, however, the made in the lateral part of the muscle tor recession and were therefore
results of treatment with adjustable to detach it from its lateral extension. excluded from further analysis. The
sutures have only been reported in Three double-armed polyglactin 910 12 patients unable to attend were
small series (Levine & Chu 1991; Small (Vicryl) 6-0 sutures were passed from contacted by telephone. They were
1995; Tucker & Collin 1995; Woog the levator aponeurosis to the upper all satisfied with the outcome and
et al. 1996), and there are no long-term part of the tarsus. The sutures were none requested any further surgery.
studies. In this study, we have re- then passed through the skin edges, one All of them had only been operated
examined 58 patients (86 eyelids) trea- arm above and the other below the once.
ted with this procedure. incision and tied temporarily. Polyester Among the 58 patients that partici-
(Ethicon, Kirkton, Scotland) 6-0 pated in the follow-up examination,
sutures were used for additional skin median age at surgery was 49 years
Materials and Methods closure. Eyelid position was adjusted (range 32–72 years), and median time
All patients that had primary surgery the day after surgery. No anaesthesia from surgery to the follow-up exam
due to upper eyelid retraction with was used during the adjustment. The was 56 months (range 6–123 months).
levator recession using adjustable patients were then seen after 1 week. In Correction was performed bilaterally in
sutures in the period 2001–2011 were under-corrected patients, both the po- 28 patients, and unilaterally in 30
invited to participate in the study. A lyglactin and the polyester sutures were patients. During the active stage of
written informed consent was obtained removed, otherwise only the polyester the disease, 23 patients had been
from the participants. The study was sutures were removed. treated with corticosteroids. Of these,

794
 Acta Ophthalmologica 2014

6 had additional treatment with retro-

bulbar irradiation, while 4 received
retrobulbar irradiation only. Orbital
decompression had been performed in
35 patients (21 lateral, 3 medial endo-
scopic, 5 combined lateral and medial (A) (B)
endoscopic and 6 three-wall decom-
pressions). Strabismus surgery for ver-
tical deviation had been performed in 6
patients and for esotropia in 6. Treat-
ment of the thyroid disease included no
treatment in 5, carbamizole or propyl- (C) (D)
thiouracil in 7, radioactive iodine in 23,
thyroidectomy in 17 and radioactive Fig. 1. Photographs of patients before (A,B) and after (C,D) correction of bilateral (A,C) and
iodine and thyroidectomy in 6. At the unilateral (B,D) upper eyelid retraction with levator recession and adjustable sutures. Both
time of surgery, all patients had been patients were judged as having a good result.
euthyroid for at least 12 months.
Retraction was defined as mild (eye-
lid margin <1 mm above limbus) in 8 Median (range) central height of the Discussion
lids, moderate (1–2 mm above limbus) palpebral fissure was 13 mm (10–
in 62 lids, and severe (>2 mm above 19 mm) preoperatively, and 10 mm Adjustable sutures have been used
limbus) in 16 lids. Before operation, (7–12 mm) at follow-up. Median val- previously both in pre-Whitnall levator
lagophthalmos and keratopathy was ues of eyelid retraction measured as the recession (Small 1995; McNab et al.
recorded in 30% and 21% of the eyes, distance from the limbus to the eyelid 2004) and in recession of the levator
respectively. Proptosis of more than margin was 1 mm preoperatively aponeurosis (Levine & Chu 1991;
21 mm was seen in 12 patients. (range 0.5 to 3 mm), and 1.5 mm Small 1995; Tucker & Collin 1995;
Information on smoking habits was at follow-up (range 3 to 1 mm). Woog et al. 1996). Small reported bet-
available for 29 patients, 22 were Mean reduction of eyelid retraction ter results with pre-Whitnall levator
smokers. was 2.5 mm (p < 0.001). recession than with levator aponeurosis
Among the 80 patients that under- After surgery, vertical motility of the recession and was concerned about the
went levator recession with adjustable upper eyelid improved. Lagophthal- high reoperation rate in recession of
sutures, reoperations had been per- mos, which was present in 30% of lids the levator aponeurosis (32%). Collin
formed in 13 patients. All of these preoperatively, was seen only in 7% at and O’Donnell (Collin & O’Donnell
attended the follow-up examination. In follow-up. 1994) modified the technique for reces-
6 eyelids, reoperation was due to Eyelid symmetry and contour were sion of the levator aponeurosis, and
under-correction, and in 7 because of evaluated from digital photographs. Tucker and Collin reported good
secondary ptosis. All cases with under- A perfect symmetrical appearance results in 67% (6/9 eyelids in 7 patients)
correction were caused by tight sutures, was seen in 34 patients (59%), and of patients with thyroid eye disease
and were recognized already at 1 week 62 lids (72%) had perfect eyelid using criteria comparable to those of
after the operation. Despite removing contour. Mourits and Sasim (Mourits & Sasim
the adjustable sutures together with the After bilateral upper lid surgery, a 1999). Levine and Chu had successful
skin sutures at this time-point, under- good result (Fig. 1) was found in 20 outcome in 1 patient, while Woog and
correction still persisted. Two patients patients, an acceptable result in 6 and co-workers obtained good results in
had postoperative infection, and both an unacceptable result in 2. Among 83% (10/12 lids in 9 patients with
developed a secondary ptosis. Another those who underwent unilateral lid thyroid eye disease) with this tech-
patient developed ptosis gradually sev- surgery, a good result (Fig. 1) was nique.
eral years after surgery, while in the observed in 14, an acceptable result in Graded levator and M€ uller muscle
remaining 4, ptosis was seen already at 13 and an unacceptable result in 3. The recession can also be performed with-
1 week after surgery. Patients with proportion of patients achieving only out sutures using either a transcon-
under-correction underwent 1 or 2 (1 an acceptable result is significantly junctival (Putterman 1981; Ben Simon
case) additional procedures with higher among those with unilateral et al. 2005) or transcutaneous
adjustable sutures to achieve a good versus bilateral surgery (43% versus approach (Harvey & Anderson 1981).
or acceptable result. Two patients with 21%, p < 0.05). Good results were reported in 81–90%
ptosis had 4 additional procedures, 3 We did not find that proptosis with the transconjunctival and in 67%
with adjustable sutures and finally 1 >21 mm, or impaired vertical motility (14 patients, 24 lids) with the transcu-
with fixed sutures. The remaining 5 of the eye or the lid, including lagoph- taneous approach. Using a transcuta-
patients with ptosis needed only 1 thalmos, significantly influenced the neous approach with a fixed suture in
additional procedure for correcting surgical outcome. Patients with severe the lateral part of the eyelid, Mourits
the ptosis. The final outcome among retraction (>2 mm) were slightly over- and Sasim obtained good results in
those who had been reoperated was corrected with a mean distance from 56% and acceptable results in 31% (50
good in 5, acceptable in 6 and unac- the limbus to the lid margin of patients/78 lids) (Mourits & Sasim
ceptable in 2. 1.69 mm. 1999).

795
Acta Ophthalmologica 2014

Perioperatively, all patients were Using the same procedure in 41 All degrees of eyelid retraction can be
overcorrected as it is easier to raise patients and evaluating the outcome corrected, but the procedure is time-
the eyelid than to lower it postopera- with the criteria of Mourits and Sasim, consuming, and patients must be
tively. In patients with tight sutures, Hintschich and Haritoglou reported a informed about the risk for reopera-
both the skin sutures and the adjust- perfect result in 72% (43/60 lids) and tion. Compared with other procedures,
able sutures were removed 1 week acceptable in 23% (14/60 lids) after 1 it appears that equally good or better
postoperatively. However, the position or 2 procedures (Hintschich & Hari- results can be achieved with less
of the eyelid is fairly stable at this time- toglou 2005). In a recent study, Shortt demanding techniques. Still, there is a
point (Mourits & Sasim 1999; Shortt and co-workers used a graded need for randomized studies to deter-
et al. 2011), and unless there is an approach with levator recession, lateral mine the optimal procedure for cor-
immediate lowering of the lid following and in some cases medial M€ ullerecto- recting upper eyelid retraction.
suture removal, a reoperation should my, and, if necessary, division of the
be considered as it is quite easy to pull conjunctiva (Shortt et al. 2011). Suc-
the tissues apart and insert a new set of cessful outcome using similar criteria as
Acknowledgements
sutures. Mourits and Sasim was reported in The assistance of photographer B
ard
The procedure, as described by Col- 87% of the eyelids (45/52 eyelids). Kjersem is greatly appreciated.
lin and O‘Donnell, included cutting the The reoperation rates in the more
medial horn of the levator aponeurosis recent studies include 18% (7/38 lids)
(Collin & O’Donnell 1994). We find (McNab et al. 2004), 13.5% (7/52 lids) References
that in lids with mild or moderate (Shortt et al. 2011), 12% (7/60 lids)
retraction this may not be necessary to (Hintschich & Haritoglou 2005), 8% Bartley GB (1996): The differential diagnosis
and classification of eyelid retraction. Oph-
achieve overcorrection perioperatively. (1/12 lids) (Woog et al. 1996), 6.5% (7/
thalmology 103: 168–176.
While good results could be achieved 107 lids) (Ben Simon et al. 2005), 3% Bartley GB, Fatourechi V, Kadrmas EF,
for all degrees of retraction, patients (2/72 lids) (Ceisler et al. 1995), and 2% Jacobsen SJ, Ilstrup DM, Garrity JA &
with severe retraction were on aver- (1/50 lids) (Elner et al. 2004). The main Gorman CA (1996): Clinical features of
age slightly overcorrected. A similar reason for reoperation is either under- Graves’ ophthalmopathy in an incidence
experience was reported by Elner or over-correction with subsequent cohort. Am J Ophthalmol 121: 284–290.
and co-workers in their series on full- asymmetry between the two upper Ben Simon GJ, Mansury AM, Schwarcz RM,
thickness anterior blepharotomy. Here, eyelids. Contour abnormalities, wound Modjtahedi S, McCann JD & Goldberg RA
(2005): Transconjunctival Muller muscle
the final heights in severe retraction dehiscence and skin crease recession
recession with levator disinsertion for cor-
were more variable, and patients with less frequently lead to reoperation. rection of eyelid retraction associated with
mild lid retraction had the least chance When comparing the results from thyroid-related orbitopathy. Am J Ophthal-
of having postoperative asymmetry different studies, it is important to bear mol 140: 94–99.
(Elner et al. 2004). in mind that the studies vary with Ceisler EJ, Bilyk JR, Rubin PA, Burks WR &
The outcome after unilateral surgery respect to the number of patients stud- Shore JW (1995): Results of Mullerotomy
was less successful than after bilateral ied, the severity of lid retraction and and levator aponeurosis transposition for
surgery with a larger proportion of the criteria used for evaluating the the correction of upper eyelid retraction in
Graves disease. Ophthalmology 102: 483–
patients achieving only an acceptable outcome. Also, in some series, there is
492.
result. Others have not observed any a mix of patients with thyroid eye Cockerham KP, Hidayat AA, Brown HG,
difference in the outcome after unilat- disease and other conditions, and some Cockerham GC & Graner SR (2002): Clin-
eral and bilateral surgery (Mourits & include a re-examination of patients, icopathologic evaluation of the Mueller
Sasim 1999). However, some patients while others are based solely on hospi- muscle in thyroid-associated orbitopathy.
with unilateral retraction may actually tal records. It is important to re-exam- Ophthalmic Plast Reconstr Surg 18: 11–17.
have bilateral disease, and this in addi- ine the patients since unacceptable Collin JR & O’Donnell BA (1994): Adjustable
tion to the effect of Hering‘s law may results may not be noticed if the sutures in eyelid surgery for ptosis and lid
retraction. Br J Ophthalmol 78: 167–174.
influence the outcome. patients are only interviewed by phone
Elner VM, Hassan AS & Frueh BR (2004):
Other techniques for correcting or by answering a questionnaire. There Graded full-thickness anterior blepharoto-
upper eyelid retraction in thyroid eye are very few studies where different my for upper eyelid retraction. Arch Oph-
disease have been introduced more procedures are compared. In a retro- thalmol 122: 55–60.
recently. Levator aponeurosis transpo- spective study, Tucker and Collin Gaddipati RV & Meyer DR (2008): Eyelid
sition has been used in 2 studies. found that levator recession with retraction, lid lag, lagophthalmos, and von
Ceisler and coworkers reported excel- adjustable sutures gave better results Graefe’s sign quantifying the eyelid features
lent results in 81% (30/37 patients, 58/ than levator recession without sutures of Graves’ ophthalmopathy. Ophthalmol-
ogy 115: 1083–1088.
72 eyelids) and good results in 6/37 or with non-adjustable sutures (Tucker
Grove AS Jr (1981): Upper eyelid retraction
patients (13/72 eyelids) (Ceisler et al. & Collin 1995). and Graves’ disease. Ophthalmology 88:
1995), while Schaefer obtained excel- 499–506.
lent results in 91% (32 patients, 48
eyelids) (Schaefer 2007). Elner and
Conclusion Harvey JT & Anderson RL (1981): The
aponeurotic approach to eyelid retraction.
co-workers obtained good results with Good results can be obtained after Ophthalmology 88: 513–524.
anterior blepharotomy in 75% (32 levator recession with adjustable Hintschich C & Haritoglou C (2005): Full
patients 50 eyelids) (Elner et al. 2004). sutures for the treatment of upper thickness eyelid transsection (blepharotomy)
for upper eyelid lengthening in lid retraction
Only 1 patient had asymmetry >1 mm. eyelid retraction in thyroid eye disease.

796
 Acta Ophthalmologica 2014

associated with Graves’ disease. Br J Oph- Mourits MP & Sasim IV (1999): A single Small RG (1995): Surgery for upper eyelid
thalmol 89: 413–416. technique to correct various degrees of retraction, three techniques. Trans Am Oph-
Lee SJ, Rim TH, Jang SY, Kim CY, Shin DY, upper lid retraction in patients with Graves’ thalmol Soc 93: 353–365; discussion 365–
Lee EJ, Lee SY & Yoon JS (2013): Treat- orbitopathy. Br J Ophthalmol 83: 81–84. 359.
ment of upper eyelid retraction related to Mourits MP, Prummel MF, Wiersinga WM & Tucker SM & Collin R (1995): Repair of upper
thyroid-associated ophthalmopathy using Koornneef L (1997): Clinical activity score eyelid retraction: a comparison between
subconjunctival triamcinolone injections. as a guide in the management of patients adjustable and non-adjustable sutures. Br J
Graefes Arch Clin Exp Ophthalmol 251: with Graves’ ophthalmopathy. Clin Endo- Ophthalmol 79: 658–660.
261–270. crinol 47: 9–14. Woog JJ, Hartstein ME & Hoenig J (1996):
Lelli GJ Jr, Duong JK & Kazim M (2010): Putterman AM (1981): Surgical treatment of Adjustable suture technique for levator
Levator excursion as a predictor of both thyroid-related upper eyelid retraction. recession. Arch Ophthalmol 114: 620–624.
eyelid lag and lagophthalmos in thyroid eye Graded Muller’s muscle excision and levator
disease. Ophthalmic Plast Reconstr Surg 26: recession. Ophthalmology 88: 507–512.
7–10. Schaefer DP (2007): The graded levator hinge
Received on September 30th, 2013.
Levine MR & Chu A (1991): Surgical treat- procedure for the correction of upper eyelid
Accepted on February 23rd, 2014.
ment of thyroid-related lid retraction: a new retraction (an American Ophthalmological
variation. Ophthalmic Surg 22: 90–94. Society thesis). Trans Am Ophthalmol Soc
Correspondence:
Mancini R, Khadavi NM & Goldberg RA 105: 481–512.
Dr. Eyvind Rødahl
(2011): Nonsurgical management of upper Shih MJ, Liao SL, Kuo KT, Smith TJ &
Department of Ophthalmology
eyelid margin asymmetry using hyaluronic Chuang LM (2006): Molecular pathology of
Haukeland University Hospital
acid gel filler. Ophthalmic Plast Reconstr Muller’s muscle in Graves’ ophthalmopathy.
N-5021 Bergen
Surg 27: 1–3. J Clin Endocrinol Metab 91: 1159–1167.
Norway
McNab AA, Galbraith JE, Friebel J & Caesar Shortt AJ, Bhogal M, Rose GE & Shah-Desai
Tel: +47 55 97 41 10
R (2004): Pre-Whitnall levator recession S (2011): Stability of eyelid height after
Fax: +47 55 97 41 43
with hang-back sutures in Graves orbitopa- graded anterior-approach lid lowering for
Email: eyvind.rodahl@helse-bergen.no
thy. Ophthalmic Plast Reconstr Surg 20: dysthyroid upper lid retraction. Orbit 30:
301–307. 280–288.

797