TITLE: Periocular Skin Malignancies
SOURCE: Grand Rounds Presentation, UTMB, Dept. of Otolaryngology
DATE: May 8, 2008
RESIDENT PHYSICIAN: Jeffrey Buyten, MD
FACULTY PHYSICIAN: Vicente Resto, MD, PhD
SERIES EDITORS: Francis B. Quinn, Jr., MD
"This
material was prepared by resident physicians in partial fulfillment of
educational requirements established for the Postgraduate Training Program =
of
the UTMB Department of Otolaryngology/Head and Neck Surgery and was not
intended for clinical use in its present form. It was prepared for the purp=
ose
of stimulating group discussion in a conference setting. No warranties, eit=
her
express or implied, are made with respect to its accuracy, completeness, or
timeliness. The material does not necessarily reflect the current or past
opinions of members of the UTMB faculty and should not be used for purposes=
of
diagnosis or treatment without consulting appropriate literature sources and
informed professional opinion."
Anatomy
The periocular region is a common site for head = and neck skin malignancy presentation. It is bounded by the brow superiorly, the infraorbital rim inferiorly, the nose medially and the lateral orbital rim. The main functions of the eyelids are to protect the eye by acting as a barrier to light and physical trauma, as well as prevent dessication by means of tear production. Although they are thin structures,= both the upper and lower lids contain distinct layers which can be split into anterior and posterior lamellas. The anterior lamella consists of skin and the orbicularis oculi musc= le. The posterior lamella consists of the tarsus and the conjunctive. The upper= lid is much thinner than the lower lid.
The tarsus provides rigidity to the upper and lo= wer lids. It is made up of dense, fibrous tissue and contains meibomian glands. The tarsus is about 25 mm in horizontal length, one mm thick and varies in height. The upper tarsal plat= e is about 10 mm in height. The lower tarsal plate is 4 mm in height.
The orbicularis oculi muscle is a sphincter that= facilitates eye closure. Since there is very little subcutaneous tissue in the periocul= ar region, it lies just below the skin. The periosteum and orbital septum lie = one layer deep to the orbicularis muscle. The orbital septum is a fascial barri= er that prevents the spread of superficial infection into the orbit.
Basal=
Cell
Carcinoma (BCCA)
Basal cell carcinoma is the most common skin malignancy. It is locally destructive and slow growing with an extremely low metastatic rate (0.0028 = to 0.55%). Most cases (80%) pres= ent in the head and neck because of its chronic sun exposure. Almost all (80-95%) of lid and med= ial canthal malignancies are BCCA. In the periocular region, the lower lid is the most common site (43%), followe= d by the medial canthus (26%), upper lid (12%) and the lateral canthus (8%).
The incidence of BCCA is increasing up to 10% per
year. In
Once a patient is diagnosed with BCCA, a diligent search for other suspicious lesions must be performed. Up to 60% of patients present with multiple lesions and between 20-40% of lesions are missed clinically. Some patients present with large a= nd locally destructive lesions that are referred to as rodent ulcers. Periocular lesions typically grow = and extend laterally because of the tarsal plate. If left untreated, the lesions can extend into the orbit via the periosteum.&= nbsp; They can eventually enter the cranial cavity and paranasal sinus involvement for large, neglected lesions is common.
Microscopically, BCCA is a proliferation of basa= loid cells from the epidermis that invade the dermis. There are four major types of BCCA: nodular, superficial, infiltrative and mixed. Nodular BCCA makes up 50% of lesio= ns and presents grossly as a pearly papule with a rolled border and telangiectasias. A linear var= iant of nodular BCCA tends to occur in the periocular region and may be more aggressive clinically. Superf= icial BCCA accounts for 15% of lesions and is more common on the trunk than the h= ead and neck. Grossly, it present= s as a scaly, erythematous plaque or patch that may mimic psoriasis, discoid eczem= a or carcinoma in situ. Infiltrati= ve BCCA makes up 10-20% of cases. It presents as a more indurated lesion with indistinct margins. Microscopically, the tumor cells a= re parts of irregular groups. The morpheic variant is marked by a dense, sclerotic stroma and it has a higher rate of perineural invasion. Perineural invasion (PNI) is an uncommon feature for BCCA, occurring= in one to three percent of cases histologically. When present, PNI is a risk factor= for recurrence.
High risk lesions occur in the setting of prior non-surgical therapy, recurrence and prior incomplete excision. Other risk factors include: size greater than 2 cm, H-zone location, medial canthal extension, poorly defined margins, infiltrative/morpheic/micronodular subty= pes, perineural invasion and Immunosuppression..
Molecular markers associated with BCCA include p= 53 and Bcl-2. p53 is associated with sclerosing subtypes and more aggressive behav= ior. In contrast, Bcl-2 is associated with low risk lesions and slower tumor gro= wth. One unique molecular difference between BCCA and SCCA is that nearly all BC= CA lesions express Gli1. This protein is regulated by sonic hedgehog (Shh) transcription factors. The deregulation of the Shh-Gli pathway has been fou= nd in both familial and sporadic cases of BCCA.
The Australian Mohs Database review of periocula= r BCCA shows that nodular lesions are the most common primary BCCA (45%) while infiltrating lesions were the most common recurrent lesions (39%). The overall recurrence rate for periocular BCCA is between 1-35% depending on the surgical treatment modality. Two thirds of recur= rences happen within three years of treatment and eighteen percent of recurrences happen between five and ten years post treatment. The best treatment is Mohs microgr= aphic surgery which has been shown to have a 1-2% recurrence rate for primary tum= ors and 5.6-7.8% recurrence rate for recurrent tumors. For non-Mohs excision, recurrences= range widely. En face frozen section techniques, which mimic Mohs excision has similarly low recurrence rates of 2.1% for primary tumors and 4.4% for recurrent tumors.
Squamous Cell Carcinoma (SCCA)
SCCA accounts for 5-10% of periocular malignanci=
es. It
typically presents as a keratinzed plaque that eventually erodes tissues and
ulcerates. It is characterized by more rapid growth rates and a much higher
metastatic rate (4-45%) when compared to BCCA. The overall metastatic rate =
is
less than 10%. Its incidence mirrors BCCA when comparing
In the periocular region most lesions are well differentiated (50%). Moderately differentiated lesions make up 37% and poo= rly differentiated lesions occur in 6% of cases. As with BCCA, perineural invas= ion is a poor prognostic indicator but it is much more common (2.5-14%). PNI is associated with large tumors (> 2cm), head and neck primary, prior recurrence, poor differentiation and overall aggressive tumor behavior.
High risk lesions occur in the setting of incomp= lete excision, recurrence and prior non-surgical therapy. They also occur on the lip, ear, lids. Other risk factors include: size greater than 2 cm, poor differentiation, deep lesions (greater than 4 mm), perineural invasion, scar carcinomas and immunocompromised patients.
Patients that fall into the high risk category s= hould undergo a more extensive evaluation of the neck and parotid gland. These ar= eas should be imaged appropriately and treatment adjusted accordingly. Patients= at high risk should be counseled regarding parotidectomy and neck dissection.<= span style=3D'mso-spacerun:yes'> Patients with advanced parotid dis= ease (nodes greater than 6 cm, facial nerve or skull base involvement) have a statistically significant lower survival rate.
The Australian Mohs database for periocular SCCA= shows that 5 year local recurrence rates for non-Mohs therapies range between 3-2= 3%. For patients treated by Mohs techniques, the study demonstrated a recurrence rate of 4%. Based on prior Mohs studies, excision margins of 4 mm for low r= isk SCCA and 6 mm for high risk SCCA should produce negative margins in 95% of patients. High risk lesions should be excised with subcutaneous fat because= up to 30% invade to this level.
Hybrid lesions have been named Basosquamous carc= inoma. The presence of both squamous cell characteristics and basal cells is a confounding sight. There are three reasons for this to happen. Differentiat= ion from BCCA to SCCA or vice versa is theoretically feasible and immunostaining has demonstrated true basal and squamous transition zones in lesions. Another possibility is the presence of individual BCCA and SCCA lesions occurring at the same time and lie adjacent to one another. The third possibility is a variant of nodular BCCA named keratinizing BCCA. It is important to make the distinction between these possible etiologies because treatment regimens wi= ll differ. Keratinizing BCCA can be treated as a low risk lesion if it is recognized. Likewise, adjacent lesions, or collision tumors, may be treated= as separate low risk lesions. True differentiation between tumor types is considered to be a highly aggressive lesion with reported recurrence rates = of 52% and metastatic rates of 9.7%.
Mohs micrographic surgery
Frederick Mohs introduced his excision technique= in 1941. His initial report on 1986 periorbital cases boasts a 5 year cure rat= e of 99% for BCCA and 98% for SCCA. These results have been reproduced by numero= us studies. The technique hinges on the fact that the entire surface of the excised lesion is examined microscopically. If any margins are positive, further excision is performed. The lesions are sliced in the plane of the lesion instead of vertically. Each cycle of excision and microscopic evalua= tion is termed a stage. The other key factor for Mohs surgery is that the surgeon and pathologist are the same person. The precision of Mohs surgery allows f= or maximal tissue preservation and function of surrounding structures. In the periocular region, BCCA Mohs defects are 4.2 to 4.6 times larger than the original tumor. Morpheic lesions are up to 6.1 times larger than the origin= al tumor and require the most stages. SCCA Mohs defects are up 2.6 times larger than the original tumor.
Non-Mohs excision
Mohs surgery is not widely available; so many su= rgeons have to excise cutaneous lesions without the precision that it provides. Traditional surgical excision with delayed margin evaluation has recurrence rates ranging from 23 to 35%. Other non-Mohs modified margin control techni= ques have recurrence rates that match standard Mohs surgery. Standard excision w= ith immediate repair and postoperative histology for BCCA is not uncommon. Site directed frozen sections in this technique are not always accurate and re-excision in a repaired site may not be straight forward if a rotational = flap was utilized. Excision with delayed repair, leaving the wound open while traditional paraffin sections are processed overnight yields good results b= ut is not as expedient as Mohs. Excision with en face frozen section margin control mirrors Mohs results but requires the combined effort and coordinat= ion of both the surgeon and pathologist in the operating room.
For periocular BCCA, Hamada et al showed that in= itial tumor excision with 4 mm margins completely removed the tumor in 84% of lesions. Re-excision of positive margins only revealed malignant cells in 5= 3% of cases. The 5 year recurrence rate was 4.35%. This study utilized standard paraffin sections for histologic analysis and performed delayed closure for poorly demarcated tumors or large defects requiring complex repair. Khandwa= la et al had a recurrence rate of 1.23% for primary BCCA lesions and 12.5% for recurrent BCCA lesions using an overnight paraffin section analysis of the margins. Wong et al use en-face frozen section analysis that is similar to = Mohs because of the quick feedback and communication between the pathologist and surgeon.
For SCCA, frozen section control of margins is m= uch more effective. Standard surgical excision for periocular SCCA has a recurr= ence rate of 2.8% at 6 years. Goysal et al reviewed 76 patients over ten years w= ith periocular SCCA. The lesions were excised with 4-5 mm margins and were clos= ed primarily if possible or delayed if the defects were large. Post op radiati= on therapy was given to 20% of patients. Patients with lesions less than one centimeter (n=3D21) had no recurrences. One of 22 patients with lesions gre= ater than one centimeter recurred and almost 50% of patients with orbital invasi= on recurred (n=3D33).
Radiation Therapy (XRT)
Primary radiation therapy is reserved for perioc= ular skin malignancies is a useful treatment option. Patients that are poor surg= ical candidates, have unresectable tumors or do not desire surgery should be adv= ised that XRT is an option. Relative contraindications to XRT are patients young= er than sixty, readily excisable tumors and lesions lying over the lacrimal gl= and. Side effects of XRT in the periorbital region include: erythema, skin atrop= hy, subcutaneous fibrosis, ulcers, epiphora, dry eyes, cataracts, neovascular glaucoma, retinopathy and optic neuropathy. Zagrodnik et al reviewed 148 patients with BCCA. These lesions were treated with primary XRT and were located on all body sites. The overall recurrence rate for all sites was 15= .8%. Nodular lesions had a 8.2% rate, superficial 16% and sclerosing 27.2% recurrence rate. Rodriguez-Sains et al reviewed 631 patients with periocular BCCA, of which 55 were treated with primary XRT. The average time to recurr= ence following therapy was 5.3 years. There were 7 recurrences, which were all located in the medial canthus. Almost half of patients that recurred requir= ed orbital exenteration. Lesions greater than one centimeter were much more li= kely to recur also (9.5% vs 2%).
Orbital invasion
Because of the orbital septum, orbital invasion = by periocular skin malignancies is rare. SCCA is more likely to invade due to = its more aggressive nature, with reported frequencies ranging from 0.2 to 8.2%. BCCA can invade the orbit, but at a lower rate (0.8 to 3.6%). The time from first detection of the lesion and orbital invasion ranges from 2-25 years. = SCCA lesions averages only one year from detection to invasion, while BCCA lesio= ns average 9.8 years. The most common signs and symptoms of orbital invasion a= re: painless mass, tumor fixation to bone, limited ocular motility, extreme gaze diplopia (early sign, globe displacement, ptosis and proptosis (uncommon). Tumors spread along the periosteum into the orbit but rarely invade the glo= be itself. They can spread intracranially via the superior orbital fissure and intranasally. Perineural invasion is found in 19% of patients with orbital invasion. Patients with suspe= cted orbital invasion should undergo both CT and MRI in order to evaluate bony erosion, involvement of orbital contents and intracranial spread. Orbital exenteration is the standard therapy with or without radiation therapy depending on the tumor characteristics and margin status. Leibovitch et al&= #8217;s review of BCCA patients found that 56% of patients underwent exenteration without any radiotherapy. 19% of patients underwent exenteration with post = op radiotherapy, 6% underwent excision, 9% underwent excision with post op radiotherapy and 6% underwent primary radiotherapy.
Orbital exenteration
About half of patients referred to ophthalmologi= sts for exenteration have periocular cutaneous malignancies. Orbital exenterati= on removes the globe and all orbital contents within the bony socket. The exci= sion may extend to include the eyelids, bony orbital walls, nasal sinuses and sk= ull base. Absolute indications for orbital exenteration include: orbital apex involvement, extraocular muscle involvement, bulbar conjunctiva involvement, sclera involvement, lid involvement beyond a reasonable hope for reconstruc= tion and nonresectable full thickness invasion through the periorbita into the retrobulbar fat. The overall mortality rate for exenteration patients is 93% after one year but decreases to 57% at five years and 37% at ten years. In = 54% of cases, mortality is linked to the orbital tumor but 38% of patients die = of unrelated medical conditions and 8% are due to metastatic disease. Interestingly, there is no significant difference in the 5 year survival between BCCA lesions and all other tumors requiring exenteration.
Lid reconstruction
Reconstruction of the eyelids requires a detailed understanding of their anatomy and function. The basic principles of closure are the same for both upper and lower lid defects. Small lesions which are = up to 33% of the lid may be closed primarily. Primary closure can still be performed for defects measuring 40% of the lid surface if the patient has l= id laxity. Rotational flaps may be used for medium sized defects between 33-50= % of the lid. For defects greater than 50% of the lid, bridging procedures are utilized to borrow tissue from the opposite lid.
Primary closure reapproximates the tarsus, orbicularis, conjunctiva and skin, using the grey line as a guide. To decre= ase tension, a canthotomy and cantholysis may be performed.
The primary rotational flap used for medium sized defects is the Tenzel flap. This flap is a semicircle based at the lateral canthus. A myocutaneous flap is raised and cantholysis and canthotomy are performed. The flap is then advanced and reapproximated in layers.
The Hughes procedure is a bridging procedure for= the large lower lid defects. It uses the upper lid posterior lamella to create a new posterior lamella for the lower lid. A tarsoconjunctival flap is raised from the upper lid. This flap is sutured to the lower lid remnant to recrea= te the conjunctival layer. The lower lid anterior lamella is recreated using l= ocal myocutaneous flaps or skin grafts. The flap is left in place for a few weeks and then is taken down. Another option for large lower lid lesions is the single stage Mustarde cheek flap.
Upper lid lesions are repaired using the same ba= sic scheme. The Tenzel flap is inferiorly oriented for upper lid repairs. For large upper lid regions, the C= utler Beard bridge flap may be utilized. This flap takes a full thickness pedicled graft below the ciliary line on the lower lid. The flap is split into the conjunctiva and the myocutaneous portion. Once split, the tarsus can be reconstructed using septal or conchal cartilage. The new anterior lamella is then sutured in to place. The flap is taken down after a few weeks and clos= ed primarily.
Orbital exenteration reconstruction
The reconstructive ladder for the exenterated or= bit starts with local options including spontaneous granulation, and skin graft= ing. Secondary intention healing takes several months and can produce skin contractures, brow ptosis and is at risk for infection. Skin grafting the s= ite allows the cavity to heal much more quickly with less contracture. Regional options in clued temporalis muscle flaps, cervicofacial flaps, temperoparie= tal flaps, forehead flaps and frontal flaps. Distal options are typically microvascular free flaps. In patients with large cavities, rectus flaps are commonly used. With the advances in anterolateral thigh flap harvest, this = flap may be used more frequently because of the more minimal donor site problems= and the flap’s ability to fill in large defects. Free flap reconstruction= has opened the door to surgical treatment of previously unresectable lesions. Craniofacial resections with dural exposure are now more easily repaired wi= th free flaps. Also, the well vascularized tissue is able to withstand post op radiation more easily than patients that heal by secondary intention or are akin grafted. Another reasonable alternative to reconstruction is prosthesis placement. Prosthetics are used in conjunction with local and regional reconstructions because the prosthesis is able to fit into the repaired cav= ity.
Bibliography