Correction of lower eyelid retraction is usually challenging because it involves different pathologies and addressing all of them is necessary to achieve acceptable outcome. Cicatricial retraction is even more challenging compared to other causes such as Grave’s orbitopathy, because the normal anatomy has been violated and scars and adhesions further complicate the corrective measures.
Deep understanding of eyelid and periorbital anatomy and etiologies of lower lid retraction is necessary to make the proper choice of surgical methods that resolve the underlying pathologies. Involvement of any of the three layers (lamellae) of the eyelid can lead to lower lid retraction especially in conjunction with horizontal lid laxity, canthal dystopia, or malar ptosis caused by aging that reduces the intrinsic lower lid support.(4) However, involvement of middle lamella (the orbital septum) plays the most important role in severity of retraction and is almost always present in moderate and severe cases.
Different procedures have been proposed to correct lower eyelid retraction, including lateral canthal surgery (canthoplasty or canthopexy), transverse tarsotomy and tarsoconjunctival flap, autogenous spacers, retractor release, cheek suspension techniques and usage of fillers.(2, 4, 13–17, 5–12)
In mild cases, typically a single underlying etiology exists that is usually sufficiently resolved by a simple procedure. However, in moderate and severe cases and particularly in cicatricial retractions, numerous pathologies are involved and a combination of several methods should be used to achieve optimal outcome.
Our suggested combined procedure addresses four major factors involved in cicatricial lower eyelid retraction, namely scars and adhesions, midface descent, middle and posterior lamellar shortening and lateral canthal tendon laxity.
Midface elevation is commonly used to correct lower lid retraction and has even been reported to be employed as the only procedure required in some cases.(18) It is particularly helpful when the problem affects the anterior lamella, as a proper midface lift recruits more skin and eliminates the need for an extra skin graft. Our approach to the midface was through the transconjunctival incision between lower edge of tarsus and retractors which obviated the need for a separate incision.
Surgical correction of the lower lid retraction frequently requires a spacer graft to keep the retractors recessed and support the tarsus in an upward position. The role of spacer grafts is more accentuated in cases with lower eyelid middle and posterior lamella scar or shortage in providing additional augmentation by lengthening the lower lid retractors and supporting the lower eyelid following release of the cicatrix.(1)
Currently, a large number of grafts are used as spacers in middle and posterior lamellae, including homologous tissues (e.g., hard palate, ear cartilage, nasal septal cartilage, tarsus, periosteum, temporalis fascia, fascia lata, and sclera), and alloplastic implants (e.g., human cadaveric acellular dermis (AlloDerm; LifeCell Corp), decellularized porcine-derived membrane (tarSys; IOP Inc), high density polyethylene (Medpor), Mersilene mesh (Ethicon), and polytetrafluoroethylene (PTFE)).(2, 3, 17, 19–22, 4, 5, 7, 9, 11–14) However, the optimal choice of spacer material is still controversial. Galindo-Ferreiro et al. in 2019 performed a web-based survey of 196 oculoplastic surgeon members of Ojoplast (Spanish and Brazilian Oculoplastic Societies) on the management of lower lid retraction and concluded that ear cartilage (39.2%), hard palate (22.7%) and tarso-conjunctival graft (19.6%) are the most preferred autogenous spacers. They also acknowledged that synthetic materials were suggested only by a small number of participants with controversy about the optimal surgical correction and long-term outcomes of each material.(7)
Although buccal mucosal membrane graft (BMMG) has been reported to be used in other oculoplastic conditions such as anophthalmic socket reconstruction(22–26), to the best of our knowledge, there have been no reports in the literature which describe using BMMG in lower eyelid malpositions. Therefore, our research represents the first study of use of BMMG as spacer graft for correction of lower lid retraction.
The ideal spacer graft should be biocompatible and easily accessible with a low rate of contracture and some degree of stiffness to provide support. Additionally, it should promote tissue integration with minimal inflammation and allow mucosalization on the conjunctival side. Ideally, it should have thickness, rigidity and contour characteristics that approximate the lower lid tarsus, and a mucosal surface so as not to irritate the corneal surface and have low risk of rejection and absorption.(3, 27) We found BMMG to be a suitable and accessible material without the disadvantages and limitations of other spacers.
Ear cartilage is one of the most commonly used materials. It is too stiff to handle, so its thickness and contour usually don’t match the tarsus and it does not easily conform to the globe. It is also limited in size, and can be palpable for years after the procedure. Downgaze movement may also be affected because of a relatively immobile lower eyelid.(28, 29)
The hard palate graft is a composite autograft with a texture and thickness similar to tarsus but donor-site morbidity and patient discomfort may discourage its use. Donor site secondary hemorrhage, granuloma, oronasal fistula, chronic mucoid discharge, formation of torus palatinus and compromised healing are examples of donor-site morbidity. Ocular complications include graft keratinization causing ocular surface irritation and potential infection from oronasal flora. So, the hard palate should be placed in a low position in the eyelid to minimize contact with the cornea and should perhaps be avoided in patients with bleeding diathesis.(3, 8, 30–32) On the other hand, some autogenous materials such as ear cartilage, hard palate and especially free tarsal graft are limited by the amount of donor material available. Besides, harvesting tarsal graft from the upper eyelid may exacerbate preexisting upper eyelid retraction in Graves’ orbitopathy patients.(20, 33)
Donor sclera is reported to have a high resorption rate and recurrence is expectable. Therefore, the use of scleral grafts has largely fallen out of favor due to high rates of degradation and shrinkage.(34)
Human acellular dermal matrix materials like AlloDerm have been reported to be contracted more often than other materials and lack the stiffness of similar grafts. Sometimes growth of fine hairs is observed and there is also a theoretical risk of transmission of infectious agents and immunogenic rejection.(35, 36)
High-density porous polyethylene (Medpor), a biointegrable implant material, can be used as a spacer graft, but the high rate of complications such as implant extrusion/exposure, poor lid motility, and ectropion limit the popularity of this material. It is associated with high failure rate when used in the lower eyelid and is more suited for skeletal augmentation.(14, 37)
Similarly, implanting Polytetrafluoroethylene (PTFE) has the risk of graft extrusion because of its exposed area that cannot support the overgrowth of conjunctival or fibrous tissue. Also, it cannot be fully vascularized as a result of insufficient size of its pores that impede the rapid ingrowth of fibrovascular tissue into the implant.(27, 38)
Decellularized porcine derived membrane (tarSys; IOP Inc) is a relatively new bioengineered eyelid spacer graft that has been used in recent years. But some disadvantages have been reported such as melting, contraction or resorption over time. Prolonged lower lid swelling and puffy appearance were other frequent complications which might be due to allergic inflammatory reaction. Besides, this implant takes at least 20 minutes to hydrate during surgery, which may add to the operating time.(20)
BMMG, has several advantages over other spacer graft materials, including readily accessible mucosa, absence of keratinizing epidermis or dermal appendages, and an easily vascularized surface from the surrounding conjunctiva to promote graft survival. Being an autologous tissue, it imposes less cost to the patient compared to alloplastic materials and lacks the risk of transmitting infections., but reducing the operation cost is an important note especially for low socioeconomic patients and donor-site morbidity is not that much intolerable as our experience. Furthermore, BMMG does not have disadvantages of other alloplastic materials such as incomplete vascularization, extrusion/exposure, and poor lid motility. Nevertheless, some surgeons still prefer alloplastic materials because of lack of the need for harvesting procedures that prolong the operating time and bear the potential for donor-site morbidity.
In our study, donor site morbidity was minimal. All patients began oral feeding after 1 or 2 days after surgery mostly with fluids, returned to regular diet in less than 1 week and recovered without scarring or deformity within 2–3 weeks. Our results were more similar to Kim et al (23), who published on their experience with buccal mucosal graft for anophthalmic socket reconstruction of 44 patients. Their subjects had only minor and tolerable morbidity of the donor site and recovered without deformities or scars. Kumar et al.(39) reported that patients who underwent buccal mucosa harvesting had problems with oral feeding and were not able to resume their regular activities until 1 month postoperatively. On the other hand, Neuschl et al.(29) concluded that short- to medium-term donor site morbidity is tolerable and long-term donor site morbidity is rare. Another study demonstrated that donor site morbidity was only noticeable within the first 3 weeks after surgery, and was tolerable as measured by a validated quality-of-life tool.(40) Some studies have found that closure of the harvest donor site seems to worsen the pain and leaving the harvest site open mitigates the pain and facilitates earlier return to a full diet. (41, 42) However, our patients, whose donor sites were sutured and repaired, had no remarkable or intolerable pain and a short recovery time. Only in two patients there was slight pain in donor site area that lasted 4–5 days and completely resolved by analgesic pills. Compared to hard palate grafts, donor site morbidity is less with BMMG and some complications like oronasal fistula do not occur.
In addition, tissue perfusion and oxygenation of the graft harvest site completely recover to a level equal to the unoperated mucosa 24 weeks after the harvesting procedure.(40, 43) Kim et al.(23) reported complete vascularization within 1 month postoperatively in 94.4% and within 2 months in the other 5.6% of their cases.
Postoperative graft shrinkage and contracture are possible in BMMG, but the rate is lower than other materials. Mean graft contraction rates have been 57% for acellular dermis and 16% for hard palate mucosal graft in lower eyelid surgery.(35) Graft contracture has been reported in only 4.5% of socket reconstructions using BMMG, less than any other spacer grafts.(23) In our study no graft contracture was observed during follow up. As postoperative contraction of the mucosal surface usually develops within 6 months after implantation, we believe further late-onset graft contracture or shrinkage in our cases would be highly unlikely.
BMMG, unlike most other autologous materials, has no limitation in harvesting, as there is enough mucosa in either side of the mouth to correct even the most severe eye lid retractions. In addition, its stiffness is suitable for posterior lamellar grafting and it can be readily modified in desirable shape and size to provide an appropriate spacer graft.
Previous studies on surgical treatment of lower eyelid retraction have showed varying outcomes.
Oestreicher et al.(11) compared the outcome of three different posterior lamellar grafts (including hard palate mucosa, free tarsoconjunctival, and free scleral grafts) in a retrospective study of 659 eyelids of 400 patients with four retraction etiologies. They reported a mean reduction of 1.3 mm in scleral show. Korn et al.(21) achieved a mean improvement of 2.03 mm in MRD2 in 16 eyes of 11 patients who underwent midfacial lifting and dermis fat as a spacer. The patients had different types and etiologies of retraction. In 24 retracted eyelids of 17 patients, Patel et al.(4) described an average of 2.5 mm of scleral show improvement using a surgical technique involving hard palate graft placement, canthopexy, and midface suspension. In another study, Patel et al.(15) described 17 patients with post blepharoplasty lower eyelid retraction who had a mean improvement in scleral show of at least 1.8 mm after hard palate graft placement and lateral tightening.. Wearne et al.(8) reported 102 eyelids of 68 patients treated with hard palate graft who had a mean reduction of 2.3 mm of scleral show. Ferri and Oestreicher (44) described 20 post blepharoplasty retracted eyelid of 10 patients treated with free tarsoconjunctival graft placement with an average decrease in scleral show of 1.6 mm and improvement in lagophthalmos. Tan et al.(14) used Medpor sheet as a posterior lamellar spacer in 35 eyelids of 32 patients. They had an average reduction in scleral show of 1 mm. Holds et al.(45) reported at least 1 mm of scleral show improvement in 200 eyelids, with an in-glove retractor lysis technique. However, 10% of the cases had at least 1 mm of residual scleral show.
In comparison, the average 2.66 mm reduction in MRD2 of our patients, is equal and slightly higher than the previous studies discussed. Also, we had residual scleral show in only one patient (8.3%), a rate that is lower than some of the studies.
These results support the efficacy of our surgical technique and spacer graft. However, the small sample size of our patients limited to only cicatricial retraction cases, can affect our findings. There are other important limitations to our study like lack of a comparison group or blinding of the researchers and the patients.
In conclusion, the RETRACT surgery, which is combination of scar release, elevation of midface, transfer of buccal mucosal membrane graft to the posterior lamella as a spacer, and lateral canthal tightening, is an effective procedure to correct cicatricial lower eyelid retractions with acceptable outcomes and a low morbidity rate.