The surgical management of childhood orbito-temporal neurofibromatosis

A retrospective interventional case series analysis was performed in all patients who first presented under the age of 16 years with an established diagnosis of NF1, treated by one of the authors (JROC) from 1978 to 1998. As the study aim was to document disease progression, patients who had postoperative follow up for less than 5 years were excluded (Table 1) .

There were 13 NF1 patients under the age of 16 in our series. Four patients with inadequate follow up were excluded. The remaining patients included five males and four females with ages ranging from 3 to 15 years at the time of their first clinic visit. The follow-up ranged from 5 to 22 years. Four patients had undergone previous surgery for their orbito-temporal deformities. All had unilateral involvement (5 right and 4 left), and the visual acuity in the affected eye ranged from 6/6 to NPL (no perception of light). Seven of the nine patients had worse visual acuity on their affected side, with one patient who already had his blind eye enucleated prior to the first clinic visit. Two more patients underwent enucleation of their blind eyes during the follow up period. All had upper lid and two also had lower lid involvement; all had orbital neurofibromatosis with varying degrees of ocular dystopia. All underwent blepharoptosis surgery, and eight patients had debulking and lateral canthal reattachment surgery. Two patients also underwent extra-ocular muscle surgery. The number of operations undergone by the patients during the study period ranged from 1 to 7. Our classification of peri-orbital deformities: (1) upper lid infiltration with ptosis (2) lower lid infiltration (3) lateral canthal disinsertion (4) conjunctival and lacrimal gland infiltration.

Neurofibromas, along with other major facial hamartomas such as haemangiomas and lymphangiomas, form during embryologic development and are seamlessly integrated into the facial structures preventing total excision. Most studies^{,  ,}  concurred that these lesions were usually obvious by the first year of age and there have been several publications over the years concerning orbito-temporal neurofibromatosis and its surgical treatment.^{,  ,}  However there have been no longitudinal studies of a cohort of paediatric patients with predominantly periorbital disease—documenting the progression of the deformities and the management over several years. One long term study examining the surgical management of head and neck neurofibromatosis in children recommended early removal of lesions in order to produce better cosmesis with removal of large painful conspicuous or strategically placed tumours.

The severity of the deformities of orbito-temporal neurofibromatosis were classified by Jackson into three groups—Group 1 (Orbital soft tissue involvement with a seeing eye), Group 2 (Orbital soft tissue and significant bony involvement with a seeing eye) and Group 3 (Orbital soft tissue and significant bony involvement with a blind or absent eye. The skeletal changes that occur in 7% of patients with NF1 consist of partial or complete absence of the greater wing of sphenoid leading to an enlargement of the superior orbital fissure, through which the temporal lobe could herniate into the orbit causing (pulsating) exophthalmos and ocular dystopia (as in Patients 3, 4 and 6). There have been many published studies in the cranio-facial literature describing transcranial approaches to repair the bony defect and debulk the orbit.^,  The main palpebral lesion, which forms the focus of this paper, is the plexiform neurofibroma which causes full-thickness infiltration of the lid leading to mechanical blepharoptosis. Localised hypertrophy of the tissues with fragmentation of the elastin results in lax lids and canthal tendons.

In keeping with previous descriptions all our patients have unilateral disease, and in all but one (Patient 9) of the patients the visual acuity was worse on the affected side. The main surgical procedures included (1) debulking surgery, (2) ptosis surgery (anterior levator resection or frontalis suspension) and (3) lateral canthal re-attachment. The lateral part of the upper lid is typically more infiltrated (creating the classical S-shaped deformity) and a full thickness pentagonal wedge can be excised in this area. Care must be taken to spare the lacrimal gland ductules where there is extensive conjunctival involvement. Pentagonal excisions can also be used to excise nodular neurofibromas causing mechanical ptosis in the upper lid or ectropion in the lower lid. Although we do not recommend routine enucleation of all blind eyes, especially where a cosmetic shell could be worn for improved cosmesis, elective enucleation of a blind eye may be part of the debulking process, and usually no orbital implant insertion is required, as there is often increased orbital volume due to temporal lobe prolapse or neurofibroma infiltration. Blepharoptosis surgery is indicated if the visual axis is compromised and there is a chance of limiting amblyopia (many patients have poor vision from congenital glaucoma or optic nerve glioma). Ptosis surgery in NF1 patients is challenging, and it is often difficult to judge the correct amount of levator resection due to the heavy infiltration. In spite of this, all of our patients, even those with complete ptosis preoperatively, achieved a satisfactory elevation of the upper lid above the pupillary axis. This is in contrast to previous reports of poor outcome of ptosis surgery. The use of scleral shells may be appropriate to hide unsightly blind eyes following blepharoptosis surgery and may function as partial blepharoptosis props, so that once removed, the lids would close satisfactorily. This decreases the risk of corneal exposure secondary to maximal levator resection.

For lateral canthal reattachments, we prefer to use nonabsorbable sutures from the tarsus into the orbital periosteum and avoid the use of periosteal flaps which may inhibit skeletal growth or lower lid fascial slings in view of further facial growth. Most of our patients had several surgical procedures staged over many years to correct the deformities. All but two patients had undergone more than two operations during the study period.

The question of the timing and extent of surgery in children with orbito-temporal neurofibromatosis is a controversial topic. Intrauterine development of neurofibromas typically involves the orbital regions and usually increase in size and number throughout childhood. However, plexiform neurofibromas may remain quiescent until the onset of puberty, a time which usually heralds a major increase in the number and size of neurofibromas. Middle and late adulthood is characterised by a progressive increase in the size and number of neurofibromas although with age this becomes more quiescent. Advocates of conservative surgery recommend selective conservative, nerve sparing, resection of functionally impairing or cosmetically deforming masses, with multiple excisions over a period of years. Proponents of more aggressive surgery^,  felt that ‘the reluctant surgeon has no place in the care of these patients’ and that repeated debulking procedures to remove the main bulk of the tumours offered the best chance of avoiding further deformities. These included Morax who advocated total one-piece lid removal and Jackson who treated the more localised form of disease with radical resection and free tissue transfer, and advocated enucleation in severe orbital disease. Likewise, Poole found that transcranial repair of the sphenoidal bony defect and orbital debulking surgery produced disappointing results unless the eye was removed. Although the main focus of this paper is on the periorbital aspects of neurofibromatosis, all orbito-temporal NF patients should optimally be managed in a multidisciplinary unit with a cranio-facial surgeon, neurosurgeon, plastic surgeon and ophthalmologist. As oculoplastic surgeons, we tailored our surgery to the achieve reasonable cosmesis and to alleviate any threat to the vision and comfort of a seeing eye, but ultimately the surgical result are always ‘compromised by the very nature of the tumour, its diffuse position, its widespread involvement of all the constituents of a region or organ, and its tendency to recur’. Definitive cosmetic surgery is best delayed until past puberty, or when the disease progression has stabilised, unless there are compelling social reasons for earlier surgery. Both the surgeon and patient need to understand that childhood surgical interventions may signal the beginning of a life long series of operations. We recommend that the oculoplastic surgeons should be involved in the care of paediatric orbito-temporal NF1 patients to allow timely application of appropriate surgical procedures to preserve vision and to improve ocular cosmesis and comfort.