Neurofibromatosis type 1: Surgical Perspectives
Nthumba PM, MMed(Surg),FCS(ECSA), Plastic Surgery Fellowship, Juma PI, MMed(Surg)
Affi liation: AIC Kijabe Hospital, Kijabe, Kenya, Africa Correspondence: Peter M. Nthumba, Plastic, Reconstructive and Hand Surgery
Unit, AIC Kijabe Hospital, Kijabe 00220, Kenya, Africa. Fax: +254-020-3204-6355. E-mail:
Introduction: Neurofibromatosis type 1 (NF1) affects about 1 in 3000 people. The indications for surgical intervention in patients with NF1 are not always clear-cut. In low-income economies, where scarcity of resources and skilled manpower often dictate levels of healthcare, a broad knowledge base of NF1 is required in order to adequately manage NF1 complications.
Materials and Methods: The authors performed PubMed/internet searches for articles on surgical aspects of NF1, as well as a review of the pathology department database for reports on all specimens submit-ted from patients with NF1 over a 16 year period. A retrospective chart review was performed on all patients with NF1 referred to the authors’ institution for surgical intervention between January 2004 and January 2011.
Results: Forty five articles describing aspects of surgical care of patients with NF1 qualified for inclusion in the review. Pathological specimens were submitted from a total of 333 patients with NF1 between 1992 and 2008. These represented 0.4% of all submitted specimens during this period. The male to female ratio was 1.05:1; 9.3% of these specimens were reported as malignant peripheral nerve sheath tumors. Fifteen of sixteen NF1 patients referred to the institution for surgical interven-tion over a seven year period underwent an average 1.7 interventions. Most presented late, some with malignancies (27%), making it difficult to obtain good cosmetic and functional results. Three representative case presentations are reported from these patients to show that (i) NF1 complications affects any system, cannot be predicted, making prognosis uncertain, (ii) cervical spine instability, excessive intra-operative bleeding and post-operative edema are important peri-operative considerations, essential role of access to safe blood for transfusion and an intensive care service.
Conclusions: Surgical symptom control (cosmesis, function, and pain) rather than surgical cure is the primary, achievable goal of most interven-tions in patients with NF1. Lifelong surveillance is necessary for all NF1 patients to avert or minimize complications, and thus improve surgical outcomes.
The term ‘neurofibromatosis’ includes neurofibromato-sis type 1 (NF1), neurofibromatosis type 2 (NF2) and others (1,2). NF1 affects about 1 in 3000 people and has an autosomal dominant inheritance pattern with com-plete penetrance, variable expression, and a high rate of new mutation. Fifty percent of NF1 patients represent new mutations. An affected individual has a 50% chance of transmitting NF1 to each offspring (3).
A positive diagnosis of NF1 is based on two or more of the following: neurofibromas, café-au-lait macules, skin-fold freckling, skeletal dysplasia, Lisch nodules, optic gliomas and a first degree relative with NF1(2). Local-ized, diffuse or plexiform neurofibromas can involve any organ in the body, resulting in varying degrees of deformities and/or dysfunction (3,4).
The psychological effects of these tumors are immense. Many affected patients are deprived of social integration, development and self actualization (5,6). Ostracized from society, children and adults with NF1 often live in poverty, hidden away only to present with advanced le-sions (Figure 1).
A multi-specialty approach is required in the care of NF1 patients. Although not always clear-cut, broad indica-tions for surgical intervention include functional im-provement (pseudo-arthrosis, penile lesions Figure 2), cosmesis (orbito-temporal lesions Figure 3 and 4), and malignancy (malignant peripheral nerve sheath tumor - MPNST Figure 5) (7). Palliative debulking is an option when this may result in an improved quality of life (Fig-ure 1). Goldberg et al. proposed a classification of NF1 patients that describes the threat to life, and can be used in long term patient follow ups in determining progres-sion (Table 1).
A broad base of knowledge of NF1 is therefore needed for the surgical treatment of NF1 complications. This paper outlines the surgical pathologies in NF1 patients, their management, anticipated difficulties and possible solutions.
Materials and methods
An internet/PubMed search was performed on the terms ‘neurofibromatosis type 1’, ‘neurofibromatosis type 1 + Africa’, and ‘surgery + neurofibromatosis type 1 + Africa’. Data on all patients with a diagnosis of NF1 from whom specimen had been obtained and submitted for histo-pathological examination were extracted. The authors also reviewed clinical data on all patients with NF1 re-ferred to the hospital for surgical intervention, between January 2004 and January 2011. From these, three repre-sentative case presentations are made.
There were 140 relevant publications retrieved from the internet/PubMed search: of these, only 45 were of surgi-cal relevance.
From a database of over 75,000 specimens submitted between 1992 and 2008, there were 333 specimens from NF1 patients evaluated, representing 0.4% of total sub-mitted specimen over this period. The male to female ratio of NF1 patients was 1.05 to 1. Histopathological examination revealed that 9.3% of the specimens were malignant peripheral nerve sheath tumors (MPNST).
A total of sixteen patients with NF1 were referred to the authors’ institution for surgical intervention between January 2004 and January 2011; one patient was deemed not to require surgery. Their average age was 19 years, with a male to female ratio of 1.2 to 1. While 44% of the patients had a good surgical outcome, two patients con-sidered their cosmesis ‘fair’ or acceptable (Figure 4). Five patients got lost to follow-up; some of these had been referred to an oncologist, but did not return for review. Twenty seven percent of the sixteen patients were found to have MPNST. Their average age was 24 years (Table 2). Thirteen percent of patients who had surgery had grade 1 lesions, 47% had grade 2 lesions, while 40% had grade 4 lesions. Their average ages were 17, 17.7 and 17.8 years, respectively.
A 27 year-old male with a longstanding history of diffi-culty in walking. He had not been in contact with anyone outside his immediate family for two years. Six months prior to presentation, he had developed dyspnea at rest, with orthopnea. A clinical examination revealed large plexiform neurofibromas on all four extremities, multiple neurofibromas, café au lait spots, a macrocephaly and severe kyphoscoliosis (Figure 1). He was planned for an initial debulking of extremities to improve his mobility, and later thoracic surgery to debulk intrathoracic neu-rofibromas.
The limbs were debulked under tourniquet control. Four units of blood were transfused peri-operatively. Post-operatively he devel-oped severe whole-body edema which improved towards the end of the first post-operative week. Attempts at extubating him were unsuccessful as he would desaturate, and a tracheostomy was per-formed to permit his weaning from the ventilator. After two weeks, he developed progressive pulmonary failure and died.
An 11 year-old male with NF1 and a large left orbito-temporal plexi-form neurofibroma and a pulsatile blind eye. He had never been to school, and could not play with other children because of this de-formity (Figure 3). He underwent two debulking procedures before he was happy to mix with other children and go to school (Figure 4). A mesh was used to replace the subcutaneous fascia; intra-oper-atively he required a unit of blood. Mesh was also used to close the left orbital cavity after enucleation and replacement of dura into the cranium through the superior orbital fissure. After the initial opera-tion he was taken to the intensive care unit (ICU) for the immediate post-operative recovery. Prolonged post-operative edema prevented extubation, necessitating a tracheostomy.
A 35 year-old male with NF1 presented with abdominal pain, early satiety and a large abdominal mass. On examination, his entire body was covered with small neurofibromas. He had a tense abdo-men with a palpable firm mass (Figure 5). A computerized tomo-gram scan revealed a large retroperitoneal tumor with para-aortic lymphadenopathy. There were no metastases in the liver, lungs or bones. He developed bowel obstruction just before his scheduled surgery. Tumor debulking with excision of tumor obstructing bowel was performed. He required intra-operative transfusion with five units of blood. He did not require ICU care post-operatively. His-tological examination of the specimen revealed an MPNST. The pa-tient recovered uneventfully and was referred to an oncologist for further management.
Ramanjam et al. presented the first of few reports on NF1 pheno-type in Africa, documenting findings consistent with Western litera-ture (8,9). There are even fewer detailing surgical interventions from
this region (10-13). Neurofibromas constituted 12.5% of 225 primary neural tumors of the head and neck re-gion, over a 10 year period in a Nairobi hospital, Kenya (14). Akadiri and Jackson aptly outlined the difficulties of surgical management in resource-poor environments (12).
The complications of NF1 may affect any system and are difficult to predict, even within families (1,15). Conse-quently the prognosis is uncertain for virtually all patients with NF1-related complications (1,16). NF1 is a progressive disorder; age is the most important determi-nant of NF1 severity, with a 25% estimated likelihood of developing grade 4 by age 30 years (Table 1) (3). The six patients (40%) with grade 4 lesions had an average age of 17.8 years. Patients with grade 1 and 2 lesions also had an average age of 17 years. The numbers in this study are small, and therefore difficult to draw any con-clusions from, but the data does suggest that sub-Saharan African NF1 patients may present with advanced le-sions at an earlier age than in Western populations. Fortunately, most NF1 patients never develop major complications (Table 2). The mainstay of their care is therefore anticipatory guidance and surveillance for treatable complications (1).
The three case presentations illustrate the difficulties as-sociated with the surgical care of NF1 patients in sub-Saharan Africa. The availability of an intensive care unit was central to the care of two of the three patients pre-sented. All three patients required blood transfusion for intra-operative hypotension from blood loss. One pa-tient died peri-operatively from pulmonary failure. Al-though this possible outcome had been discussed, the potential anticipated gains from the surgery, especially mobility, independence and social interaction, out-weighed the risks of the debulking operation.
Vascular and Cardiothoracic complications
Intrathoracic tumors can produce dyspnea, hemothorax and other serious complications (17). Vascular anoma-lies include aneurysms. Hypertension in NF1 patients may be secondary to a number of surgically correctable conditions, including: increased incidence of coarcta-tion of the aorta, phaeochromocytoma and renal artery stenosis (which may be from aneurysm formation or ex-trinsic compression by an adjacent tumor) (3).
Abnormal vessels account for massive bleeding encoun-tered during surgical debulking of neurofibromas, for which different techniques, including hypotensive an-esthesia have been used (17). Peri-operative transfusion was necessary in eight (53%) of the 15 patients who underwent surgery. The diffuse but significant bleeding encountered in these patients is difficult to control even with cautery. Application of pressure dressings and un-dersawing with multiple deep sutures aid in the control of bleeding.
Skeletal involvement may include macrocephaly, dys-plasia of the greater wing of sphenoid, bony defects of the skull, vertebral dysplasia, scoliosis, pseudoarthro-sis (usually distal tibia and fibula) and a short stature amongst others (4,5). Non-ossifying cyst-like fibromas of the long bones, genu varum and genu valgum, cortical bone defects and focal bone gigantism can occur with variable frequencies. Pressure on bone from adjacent plexiform neurofibromas may lead to erosion or bone demineralization (18-21).
Central nervous system (CNS) and the spinal cord
A large head in patients with NF1 is usually due to mac-rocephaly. A child with NF1 and macrocephaly does not need to be investigated unless there are other symptoms or signs suggestive of intracranial pathology (1). About 50% of patients with NF1 have learning disabilities, some behavioral problems and a heightened risk of epi-lepsy and headaches. These seizures may be due to the existence of a recognized complication of NF1, such as tumors, hydrocephalus or cerebrovascular disease, but may be idiopathic or a consequence of non-NF1-related pathology (1).
Gliomas are well-recognized complications of NF1 (22). They most commonly involve the optic pathways, espe-cially in children. Bilateral optic gliomas are considered by some as pathognomonic of NF1. Medulloblastomas and ependymomas have also been reported. The ma-jority of NF1-associated CNS lesions may not require specific treatment (1). However, approximately 50% of patients may show clinical or radiological progression at some stage, with some experiencing symptoms or neuro-logical deficits. Shehu and Hassan reported a case of an arachnoid cyst in a NF1 patient who had presented with quadriparesis, but recovered after surgery (23).
Other CNS complications include: migraine headaches, aqueductal stenosis, cerebrovascular disease, and spinal meningoceles, amongst others. An increased incidence of multiple sclerosis has also been reported (1).
Intraspinal and intramedullary tumors and spinal root neurofibromas usually present with impairment of pain and temperature sensation, reflex loss at the level of the tumor, sphincter disturbances or progressive paraparesis and generally have a poor outcome. Extraspinal plexi-form neurofibromas may extend into the spinal cord leading to spinal cord compression and/or radiculopa-thies (15).
Neurofibromatosis neuropathy is a progressive senso-rimotor peripheral neuropathy due to accumulation of multiple peripheral neurofibromata on the nerve roots. Hypotonia and poor muscle coordination affect some NF1 patients.
Plexiform neurofibromas are the second most common NF1 complication, after learning difficulties, affecting 30% of the NF1 patient population. They are clinically and pathologically a distinct entity from other types of neurofibromas. Plexiform neurofibromas are congenital lesions that present as large subcutaneous swellings, soft in consistency, with ill-defined margins. These lesions affect long portions of peripheral nerves and surround-ing tissue, and often lead to significant disfigurement and mechanical pressure. Nowhere else do neurofibro-mas cause as much deformity however, as in the orbito-temporo-facial region. Here, the smallest neurofibroma produces a prominent abnormality, while larger ones can produce monstrous and even hideous deformities in a person (3,5,6,24-26) (Figure 3). Lesions in this re-gion should therefore be resected/debulked in order to improve cosmesis and self/social acceptance, as in the second case presentation. Seven (44%) of the series of 16 patients presented with lesions of the head region, in-cluding four orbito-temporal plexiform neurofibromas. Plexiform neurofibromas are difficult to resect because of their diffuse nature, and the currently recommended protocol is conservative unless sudden enlargement of the mass, functional deficits, pain or bleeding from the mass develop (1).
The commonest GI lesions are stromal tumors, located mostly in the stomach and jejunum. These may present with bleeding, obstruction, or perforation. Carcinoid tumors (usually duodenal or periampullary) are fairly
common GI lesions in NF1 patients (3,27,28).
Authors have reported a higher incidence of preterm labor, Caesarian section rate, intra-uterine growth retar-dation and stillbirths amongst NF1 patients than in the general population. Most of these complications are re-lated to hypertensive complications (29,30).
Malignant tumors occur four times more frequent in the NF1 patient group than in the general population matched for age, gender, and time of follow-up (1). NF1 has a well established relationship with malignan-cies such as MPNST, CNS tumors, rhabdomyosarcomas, leukemia, and pheochromocytomas (7,22). Breast can-cer, Non-Hodgkin’s lymphomas and Wilm’s tumors amongst others have also been reported in patients with NF1 (22). MPNSTs may also arise as second tumors in the field of irradiation for another neoplasm. Triton tu-mors occur at a younger age and have a poor prognosis (31).
MPNSTs, the commonest malignancies in NF1 patients, can occur at any site in the peripheral nervous system, most commonly from plexiform neurofibromas. They represent 10% of all soft tissue sarcomas; 50% of all MPNSTs occur in NF1 patients. Twenty five percent of the patients in the current study had MPNSTs; two of the 4 patients had radiological evidence of metastases.
Total excision of localized neurofibromatosis with skele-tal and soft tissue reconstruction has had favorable long term results, but is rarely achievable. Excision requires multiple stages in the most extensive lesions (31-35). The extent of resection has proven prognostic signifi-cance in children; a greater extent of surgical resection predicts both lower risk of tumor progression and longer interval to progression. Near-total or sub-total excision of the tumor ensures recurrence rates less than 20% and 40% respectively, while removal of less than 90% of the tumor mass leads to a recurrence rate of 60% or greater (37-40).
Total resection and reconstruction is, however often tempered by the inability to perform total excisions of lesions in the head and neck, because of the need to pre-serve vital structures, such as the seeing eye, or the functional facial nerve (3,40). Functioning upper and lower extremity nerve plexuses are deterrents to total resection in the limbs. Plexiform neurofibromas involving the trunk and limbs may lead to soft tissue overgrowth, and even gigantism. The use of prosthetic mesh in the face has been shown to be effective in the long-term preven-tion of recurrence and drooping of tissue in the facial region; it acts as an inelastic subcutaneous fascia layer (39). Prosthetic mesh was used in the four patients with orbito-temporal neurofibromas in the current study.
Near-total excision was only possible in two patients: one who had a chin lesion and a second with a penile lesion. The other patients underwent what were largely debulking procedures due to the extensive nature of their lesions.
Timing of operation and extent of resection
Early resection may help minimize cosmetic and func-tional complications. However, recurrences are common in patients aged ten or less, after incomplete resection and in tumors located in the head and neck region.
The multiple procedures required in order to obtain ac-ceptable cosmesis and reduce recurrences, are a signifi-cant strain on resources. In the experience of the authors, the primary determinants of the extent of resection in sub-Saharan Africa therefore include: finances, skilled manpower, availability of safe blood for transfusion, and access to an intensive care unit (for some patients). Therefore, only the single-stage procedure likely to ob-tain the best functional and cosmetic result, with the least risk to the life of the patient, should be performed. Emergency surgical intervention may be required in cases such as acute intestinal obstruction, acute hemor-rhage into a lesion, acute neurological deterioration due to an enlarging lesion, or a life endangering mediastinal tumor (3, 41).
Prior to intubation and administration of anesthesia, cervical spine stability should be ensured. Kyphosco-liosis and fibrosing alveolitis, found in two percent and 20% of NF1 patients respectively, may increase the perioperative respiratory complications (42). Severe in-traoperative bleeding is a feature reported by many, the result of abnormal vasculature including vascular steno-ses, aneurysms, and arteriovenous fistulae (33,41-44). Blood for transfusion must be available before proceeding with the surgery. The use of tourniquets on extremi-ties significantly reduces blood loss during surgical exci-sion (45). Other techniques used include pre-operative angiography and embolization, undersawing of the neurofibroma prior to excision as well as intra-operative wound packing with gauze.
Post-operative edema, that may occasionally make the initial surgical result appear worse than the pre-opera-tive pathology is a troublesome complication that re-solves over the course of a few weeks (44). Indeed in the oro-facial region, post-resection edema may necessitate a tracheostomy, to obviate the need for prolonged intu-bation, as in two of the three case reports in this study.
While malignancy (MPNSTs and other soft tissue sarco-mas) is the leading cause of death in adults, intracra-nial tumors are the main cause of mortality in children. Acute hemorrhage (gastro-intestinal and intracranial), hypertension and its related complications are other im-portant causes in adult deaths, with MPNSTs and leuke-mia being other causes of pediatric mortality (17).
Surgical symptom control (cosmesis, function, and pain) rather that surgical cure is the primary, achievable goal of most interventions in patients with NF1. Lifelong surveillance is necessary for all NF1 patients to avert or minimize complications.
Ruggieri M. The different forms of neurofibromatosis. Childs Nerv Syst 1999;15:295–308.
National Institutes of Health Consensus Development Con-ference Statement Neurofibromatosis. Arch Neurol 1988; 45:575-578.
Goldberg Y, Dibbern K, Klein J, et al. Neurofibromatosis type 1 – an update and review for the primary pediatrician. Clin Pediatr 1996;35:545–561.
Korf BR. Neurofibromas and malignant tumors of the pe-ripheral nervous system. In: Friedman JM, Gutmann DH, MacCollin M, Riccardi VM, eds. Neurofibromatosis: Pheno-type, Natural History, and Pathogenesis. 3rd ed. Baltimore: Johns Hopkins University Press. 1999:142-161.
Ablon J. 'The Elephant Man' as 'self' and 'other': the psycho-social costs of a misdiagnosis. Soc Sci Med 1995;40:1481-1489.
Ablon J. Parents' responses to their child's diagnosis of neurofibromatosis 1. Am J Med Genet 2000;93:136-142.
Nthumba PM, Juma PI. Malignant peripheral nerve sheath tumors in Africa – a clinic-pathological study. ISRN Surgery 2011;2011. doi:10.5402/2011/526454.
Wallis CE, Slater CP. Neurofibromatosis in the South African Indian community--further evidence for heterogeneity? S Afr Med J 1987;72:478-480.
Ramanjam V, Adnams C, Ndondo A, et al. Clinical pheno-type of South African children with neurofibromatosis 1. J Child Neurol 2006;21:63-70.
Craig JB, Govender S. Neurofibromatosis of the cervical spine. A report of eight cases. J Bone Joint Surg Br 1992;74:575-578.
Floyd A, Percy-Lancaster R. The Elephant Woman; Neurofi-bromatosis associated with pseudoarthrosis of the Humerus. J Bone Joint Surg 1987; 69:121-123.
Akadiri OA, Jackson IT. Craniofacial neurofibromatosis Type 1: Clinical features, challenges of management, and a report of 2 Nigerian patients. The Internet Journal of Head and Neck Surgery. 2009;3:1.
Sheikh N, McLigeyo SO. Neurofibromatosis type 1: report of two contrasting cases. East Afr Med J 2002;79:614-617.
Chindia ML, Dimba E. Neural tumours of the head and neck. East Afr Med J 2000;77:531-533.
Riccardi VM. Type 1 neurofibromatosis and the pediatric pa-tient. Curr Probl Pediatr 1992;22:66-106.
Pollack IF, Colak A, Fitz C, et al. Surgical management of spi-nal cord compression from plexiform neurofibromas in pa-tients with neurofibromatosis 1. Neurosurgery 1998;43:248-255.
Seymour -Dempsey K, Andrassy RJ. Neurofibromato-sis: implications for the general surgeon. J Am Coll Surg 2002;195:553-563
Abdel-Wanis ME, Kawahara N. Bone development in neuro-fibromatosis 1. Med Hypotheses 2003;60:459-462.
Joseph KN, Bowen JR, MacEwen GD. Unusual orthopedic manifestations of neurofibromatosis. Clin Orthop Relat Res 1992;278:17-28.
Tonsgard JH. Clinical manifestations and management of neurofibromatosis type 1. Semin Pediatr Neurol 2006;13:2-7.
Elefteriou F, Kolanczyk M, Schindeler A, et al. Skeletal abnor-malities in neurofibromatosis type 1: approaches to thera-peutic options. Am J Med Genet A 2009;149A:2327-2338.
Yohay K. Neurofibromatosis type 1 and associated malig-nancies. Curr Neurol Neurosci Rep 2009;9:247-253.
Shehu BB, Hassan I. Cervicothoracic arachnoid cyst in a patient with neurofibromatosis: case report. East Afr Med J 2006;83:515-517.
Reynolds RM, Browning GGP, Nawroz I, et al. Von Reckling-hausen’s neurofibromatosis: neurofibromatosis type 1. Lan-cet 2003;361:1552–1554.
Cohen M. Jr. Understanding Proteus Syndrome, unmasking the Elephant Man, stemming Elephant fever. Neurofibroma-tosis 1988;1:260.
Krastinova-Lolov D, Hamza F. The surgical manage-ment of cranio-orbital neurofibromatosis. Ann Plast Surg 1996;36:263-269.
Cavallaro G, Basile U, Polistena A, et al. Surgical manage-ment of abdominal manifestations of type 1 neurofibroma-tosis: experience of a single center. Am Surg 2010;76:389-396.
Wheeler MH, Curley IR, Williams ED. The association of neurofibromatosis, pheochromocytoma, and somatostatin rich duodenal carcinoid tumor. Surgery 1986;100:1163– 1169.
Segal D, Holcberg G, Sapir O, Sheiner E, et al. Neurofibro-matosis in pregnancy. Maternal and perinatal outcome. Eur J Obstet Gynecol Reprod Biol 1999;84:59-61.
Sharma JB, Gulati N, Malik S. Maternal and perinatal com-plications in neurofibromatosis during pregnancy. Int J Gyn-aecol Obstet 1991;34:221-227.
Wanebo JE, Malik JM, Vandenberg SR, et al. Malignant pe-ripheral nerve sheath tumors: a clinicopathologic study of 28 cases. Cancer 1993;71:1247–1253.
Jackson IT, Carbonnel A, Potparic Z, et al. Orbitotemporal neurofibromatosis: classification and treatment. Plast Recon-str Surg 1993; 92:1–11.
Poole MD. Experiences in the surgical treatment of cra-nio-orbital neurofibromatosis. Plast Reconstr Surg 1989; 42:155–162.
Friedrich RE, Heiland M, Kehler U, et al. Reconstruction of sphenoid wing dysplasia with pulsating exophthalmos in a case of neurofibromatosis type 1 supported by intra-opera-tive navigation using a new skull reference system. Skull Base 2003;13:211-217.
Mukherji MM. Giant neurofibroma of the head and neck. Plast Reconstr Surg 1974; 53:185–189.
Nagata S. A systematic multiple stage surgical approach for attainment of satisfactory and favourable surgical results in an extremely severe von Recklinghausen's disease, el-ephantiasis neurofibromatosa. J Plast Reconstr Aesthet Surg 2006;59:662-674.
Needle MN, Cnaan A, Dattilo J. et al. Prognostic signs in the surgical management of plexiform neurofibroma: the Chil-dren’s Hospital of Philadelphia experience. J Pediatr Surg 1997;131:678-682.
Maceri R, Saxon KG. Neurofibromatosis of the head and neck. Head Neck Surg 1984; 6:842-850.
Park BY, Hong JP, Lee W. Netting operation to control neu-rofibroma of the face. Plast Reconstr Surg 2002;109:1228-1236.
Lee V, Ragge NK, Collin JR. The surgical management of childhood orbito-temporal neurofibromatosis. Br J Plast Surg 2003;56:380-387.
White N, Gwanmesia I, Akhtar N, et al. Severe haemorrhage in neurofibromatoma: a lesson. Br J Plast Surg. 2004;57:456-457.
Delgado JM, de la Matta Martín M. Anaesthetic implications of von Recklinghausen's neurofibromatosis. Paediatr An-aesth 2002;12:374.
Van der Meulen J. Orbital neurofibromatosis. Clin Plast Surg 1987;14:123-135.
Earley MJ, Moriarty P, Yap LH. Isolated bilateral orbital neurofibromatosis in a twelve-year-old. Br J Plast Surg 2001;54:162-164.
Power KT, Giannas J, Babar Z, et al. Management of mas-sive lower limb plexiform neurofibromatosis--when to inter-vene? Ann R Coll Surg Engl 2007;89:W3-5.