Neurofibromatosis, Type I

A number sign (#) is used with this entry because neurofibromatosis type I (NF1) is caused by heterozygous mutation in the neurofibromin gene (NF1; 613113) on chromosome 17q11.

Neurofibromatosis type I is an autosomal dominant disorder characterized by cafe-au-lait spots, Lisch nodules in the eye, and fibromatous tumors of the skin. Individuals with the disorder have increased susceptibility to the development of benign and malignant tumors. NF1 is sometimes referred to as 'peripheral neurofibromatosis.' The worldwide incidence of NF1 is 1 in 2,500 to 1 in 3,000 individuals (reviews by Shen et al., 1996 and Williams et al., 2009).

Type II neurofibromatosis (NF2; 101000) is a genetically distinct disorder caused by mutation in the gene encoding merlin (NF2; 607379) on chromosome 22q12. NF2, sometimes known as 'central neurofibromatosis,' is characterized by bilateral acoustic neuroma and meningioma, but few skin lesions or neurofibromas (Rouleau et al., 1993).

Some patients with homozygous or compound heterozygous mutations in mismatch repair genes (see, e.g., MLH1; 120436 and MSH2; 609309) have a phenotype characterized by early onset malignancies and mild features of NF1, especially cafe-au-lait spots; this is known as the mismatch repair cancer syndrome (276300), sometimes referred to as brain tumor-polyposis syndrome-1 or Turcot syndrome. These patients typically do not have germline mutations in the NF1 gene, although a study by Wang et al. (2003) suggested that biallelic mutations in mismatch repair genes may cause somatic mutations in the NF1 gene, perhaps resulting in isolated features resembling NF1.

See also Legius syndrome (611431), a genetically distinct disorder with a similar phenotype to NF1.

Sorensen et al. (1986) conducted a valuable follow-up study of the natural history of NF1 in a nationwide cohort of 212 patients with the disorder identified in Denmark by Borberg (1951). Malignant neoplasms or benign CNS tumors occurred in 45% of the probands, giving a relative risk of 4.0 compared with expected numbers. All 76 probands had been ascertained through hospitals and were more severely affected than their incidentally identified relatives, although relatives had poorer survival rates than persons in the general population. The worst prognosis was shown by female probands.

Friedman et al. (1993) described a central database designed to collect information on NF1 from 16 centers around the world. The aspects of the disorder for which information was being collected included renal artery stenosis and cerebral artery stenosis.

Dugoff and Sujansky (1996) reported outcome data of 247 pregnancies in 105 women with NF1. The 247 pregnancies resulted in 44 first trimester spontaneous abortions. The cesarean section rate (36%) was greater than in the general population (9.1 to 23.5%). In 7 of the patients, cesarean section was required because of maternal complications of NF1 including pelvic neurofibromas, pelvic bony abnormality with or without kyphoscoliosis, pheochromocytoma, and spinal cord neurofibroma. Dugoff and Sujansky (1996) reported that 80% of the women in their study experienced either the appearance of new neurofibromas, growth of existing neurofibromas, or both. Thirty-three percent of these women noted a decrease in the size of their neurofibromas in the postpartum period. Eighteen percent of the women reported no changes in neurofibromas and no appearance of new neurofibromas during pregnancy.

Friedman and Birch (1997) summarized clinical information about NF1 patients based on the International Database maintained by the National NF Foundation (NNFF), which contained information on 1,479 probands and 249 of their affected relatives with NF1 at the time of analysis. The age at diagnosis of NF1 was 8 years younger in the probands than in the affected relatives, and many of the manifestations of NF1 were more frequent in the probands than in their affected relatives. The age-specific prevalence of most manifestations of NF1 increased with age. Despite biases inherent in a convenience sample from specialist clinics, the frequency of manifestations of NF1 in many of the series was similar to those in 2 smaller population-based studies. Lisch nodules were said to be present in 57% of probands and 69.9% of affected relatives.

McGaughran et al. (1999) reported a study of 523 individuals from 304 families with NF1. More than 6 cafe-au-lait patches were seen in 383 of 442 (86.7%); 310 of 370 (83.8%) had axillary freckling; 151 of 357 (42.3%) had inguinal freckling; and 157 of 249 (63%) had Lisch nodules. Cutaneous neurofibromas were seen in 217 of 365 (59.4%), and subcutaneous tumors were present in 150 of 330 (45.5%). A positive family history of NF1 was found in 327 of 459 (71.2%). Learning disabilities of varying severity were seen in 186 of 300 (62%), and 49 (9.4%) of patients had CNS tumors, 25 of which were optic gliomas. Scoliosis was seen in 11.7%; 1.9% had pseudoarthrosis; 4.3% had epilepsy; and 2.1% had spinal neurofibromas.

Macrocephaly and short stature have been reported in several clinical studies of NF1. Clementi et al. (1999) studied growth in 528 NF1 patients obtained from a population-based registry in northeast Italy. Although macrocephaly was a consistent and common finding in NF1, short stature was less prominent and less frequent than previously reported. No differences in height were apparent between NF1 and normal subjects up to 7 years of age in girls and 12 years of age in boys. Clementi et al. (1999) presented growth charts for use by physicians following NF1 patients to assist in the identification of the effects of secondary growth disorders, for growth prognosis, and for evaluation of the effects of therapy.

Szudek et al. (2000) presented growth charts derived from study of 569 white North American children with NF1. They found that stature and occipitofrontal circumference (OFC) measurements were shifted and unimodal, with 13% of children being at or more than 2 SD below mean and 24% having OFC at or more than 2 SD above mean.

Rasmussen et al. (2001) used Multiple-Cause Mortality Files, compiled from U.S. death certificates by the National Center for Health Statistics for 1983-1997, to obtain information on mortality in NF1. They identified 3,770 cases among 32,722,122 deaths in the United States, a frequency of 1 in 8,700, which is one-third to one-half the estimated prevalence. Mean and median ages at death for persons with NF1 were 54.4 and 59 years, respectively, compared with 70.1 and 74 years in the general population. Results of proportionate mortality ratio (PMR) analyses showed that persons with NF1 were 34 times more likely to have a malignant connective or other soft-tissue neoplasm listed on their death certificates than were persons without NF1. Overall, persons with NF1 were 1.2 times more likely than expected to have a malignant neoplasm listed on their death certificates, but the PMR was 6.07 for persons who died at 10 to 19 years of age and was 4.93 for those who died at 20 to 29 years of age. Similarly, vascular disease was recorded more often than expected on death certificates of persons with NF1 who died before 30 years of age, but not in older persons.

Szudek et al. (2003) studied statistical associations among 13 of the most common or significant clinical features of NF1 in data from 4 large sets of NF1 patients comprising about 3,000 patients. The results suggested grouping 9 of the clinical features into 3 sets: (1) cafe-au-lait spots, intertriginous freckling, and Lisch nodules; (2) cutaneous, subcutaneous, and plexiform neurofibromas; (3) macrocephaly, optic glioma, other neoplasms. In addition, 3-way interactions among cafe-au-lait spots, intertriginous freckling, and subcutaneous neurofibromas indicated that the first 2 groups are not independent. Cafe-au-lait spots, intertriginous freckles, and Lisch nodules are all derived from cells of melanocytic origin, which derive from the embryonic neural crest. Thus, NF1 can be considered a neurocristopathy. The common thread between optic gliomas, other neoplasms, and macrocephaly may be glial hyperplasia. There was an observed association between pseudarthrosis and other neoplasms, which was more difficult to understand. Szudek et al. (2003) noted that these results cannot be used to predict which NF1 patients will get which particular features, but suggest that some affected individuals may be more likely than others to develop certain features of the disease.

Khosrotehrani et al. (2003) performed a cohort study among 378 NF1 patients receiving more than 1 year of follow-up care at an NF1 referral center in France. Clinical features, especially dermatologic, were evaluated as potential factors associated with mortality. Factors associated independently with mortality were the presence of subcutaneous neurofibromas (odds ratio, 10.8; 95% CI, 2.1-56.7; p less than 0.001), the absence of cutaneous neurofibromas (odds ratio, 5.3; 95% CI, 1.2-25.0; p = 0.03), and facial asymmetry (odds ratio, 11.4; 95% CI, 2.6-50.2; p less than 0.01). The absence of cutaneous neurofibromas in adulthood associated with high mortality may correspond to a subtype of NF1, familial spinal neurofibromatosis (162210). Khosrotehrani et al. (2003) concluded that features that can be found by a routine clinical examination are associated with mortality in patients with NF1, and that clinical follow-up should be focused on patients with subcutaneous neurofibromas, absence of cutaneous neurofibromas, and/or facial asymmetry. In a parallel study of a cohort of 703 NF1 patients in North America, Khosrotehrani et al. (2005) validated the observation that subcutaneous neurofibromas were associated with mortality.

Skin Manifestations

Variable numbers of hyperpigmented cafe-au-lait spots usually develop in the first years of life, but may be present at birth, and are often the first apparent feature of NF1. The quantity and size of these macules has not been linked to disease severity, and they show no tendency to malignant degeneration. The presence of 6 or more cafe-au-lait macules with diameter 0.5 cm before puberty or 1.5 cm after puberty is a diagnostic feature (see DIAGNOSIS below). Axillary and inguinal freckling ('Crowe sign') are usually noted between 3 and 5 years of age. Freckling can also occur above the eyelids, around the neck, and under the breasts. (reviews by Ferner et al., 2007 and Williams et al., 2009).

Neurofibromas are benign Schwann cell tumors that are classified according to their appearance and location: focal or diffuse cutaneous, subcutaneous, nodular or diffuse plexiform, and spinal. Focal cutaneous or dermal neurofibromas typically appear in late childhood or early adolescence, rarely cause pain or neurologic deficits, and do not transform into malignant tumors. Subcutaneous lesions can be noted on palpation of the skin and may present with tenderness or tingling distributed along the affected nerve. Plexiform neurofibromas arise from nerve fascicles, tend to grow along the length of the nerve, may involve multiple nerve branches and plexuses, and can cause significant morbidity. The growth rate is unpredictable, and soft tissue hypertrophy is often noted. Only the plexiform type of neurofibromas have a potential for transformation into malignant peripheral nerve sheath tumors (MPNST, see below) (reviews by Rosser and Packer, 2002; Ferner et al., 2007, and Williams et al., 2009).

Waggoner et al. (2000) conducted a retrospective review of neurofibromas among NF1 patients seen in a tertiary care referral center. Sixty-eight (16.8%) of 405 patients with NF1 had plexiform neurofibromas, which were located on the trunk (43%), the head and neck region (42%), and the extremities (15%). About 44% of these tumors were detected by 5 years of age. Presenting symptoms were most often related to the increasing size of the tumor, a loss of function (usually weakness), or pain. Only 2 patients (3%) developed malignant peripheral nerve sheath tumors in their preexisting plexiform neurofibromas. No specific NF1 features were associated with plexiform tumors.

To analyze growth rate and prognostic factors for progression of postoperative plexiform neurofibromas in patients with NF1, Nguyen et al. (2013) studied 52 patients (mean age 25 years, range 3-64 years) with 56 plexiform neurofibromas and looked at postoperative tumor volume change per year on MRI. Initial median tumor volume was 40.3 mL. Surgical indications included disfigurement in 21 patients, pain in 20 patients, and functional deficits in 16 patients. Sixteen percent of all cases experienced acute surgical complication, and 13% showed late complication. Eight patients (19%; 6 children, 2 adults) with residual tumor had repeat surgery for tumor progression. Median tumor progression was 0.6% change per year and 2.9% from baseline. Patients aged 21 years and younger had the highest progression rate (p less than 0.01). For every year of age the mean growth rate decreased by -0.463 mean percent (p = 0.03). With age as a continuous variable, age, the site of the tumor, and depth were the only factors associated with tumor progression. Fourteen plexiform neurofibromas (10 nodular and 4 diffuse) in 13 patients (5 children and 8 adults) were completely resected by visualization and did not relapse during observation (mean: 2.9 years; range: 1.1-5.8 years). Nguyen et al. (2013) concluded that age, tumor type, location, and depth are helpful to estimate the progression of plexiform neurofibromas after surgery and that patients benefit from elective surgery of small and completely removable plexiform neurofibromas.

Ophthalmologic Manifestations

Williams et al. (2009) noted that Lisch nodules, melanocytic iris hamartomas that do not affect vision, are pathognomonic of NF1.

Perry and Font (1982) used electron microscopic studies to demonstrate that the spindle-shaped cells within Lisch nodules are of melanocytic origin and represent melanocytic hamartomas. Thus, Lisch nodules are true tumors, not merely hyperpigmented patches.

Zehavi et al. (1986) found Lisch nodules in 73% of 30 NF1 cases, and found that their presence correlated directly with the severity of skin manifestations. Lisch nodules appeared as smooth, well-defined, gelatinous masses protruding from the surface of the iris on slit-lamp examination.

Ragge et al. (1993) provided a comprehensive discussion of Lisch nodules accompanied by colored photographs in irides of different colors. They pointed out that iris nodules were reported by several workers in the decade before the paper by Lisch (1937). In particular, Sakurai (1935) published a beautifully illustrated paper linking characteristic iris nodules with von Recklinghausen neurofibromatosis. Ragge et al. (1993) suggested that the lesions be renamed Sakurai-Lisch nodules in her honor.

On rare occasions, fibromas may occur in the iris, and glaucoma may occur (Grant and Walton, 1968). Westerhof et al. (1983) found hypertelorism in 24% of patients with neurofibromatosis.

Yasunari et al. (2000) studied 33 eyes of 17 consecutive NF1 patients diagnosed with NF1 by conventional ophthalmoscopy and by noninvasive infrared monochromatic light with confocal scanning laser ophthalmoscopy (SLO). Twenty-one digital fluorescein and indocyanine-green angiographies were obtained from 11 adult patients, and 77 angiograms were obtained from age-matched controls. Infrared monochromatic light examination by confocal SLO showed multiple bright patchy regions at and around the entire posterior pole of all 33 NF1 eyes. All bright patchy regions seen in adult patients corresponded to hypofluorescent areas on their indocyanine-green angiograms; however, no abnormalities were noted in any patient at corresponding areas under conventional ophthalmoscopic examination or fluorescein angiography. Control patients and their angiograms showed no choroidal abnormalities. Iris nodules were noted in 25 eyes (76%) of 14 patients (82%) and eyelid neurofibroma in 5 patients (29%). Since choroidal abnormalities were detected in 100% of NF1 patients examined, Yasunari et al. (2000) suggested that this abnormality be included in the diagnostic criteria for NF1.

Otsuka et al. (2001) performed serial ophthalmologic exams on 70 patients of various ages with NF1. Lisch nodules were found in 80% of patients of all ages and in two-thirds of patients younger than 10 years. Only 2 of 45 individuals older than age 10 years did not have Lisch nodules. Lisch nodules were more frequent in familial cases than in sporadic cases. Cutaneous neurofibromas developed at the average +/- SD age of 15.1 +/- 3.6 years in patients who had more than 10 Lisch nodules and at 21.8 +/- 3.9 years in those who had fewer than 10 Lisch nodules. The former group was significantly younger than the latter.

Lee et al. (2004) classified the periorbital deformities of adult orbitotemporal NF, reported previously undescribed clinical findings, and recommended guidelines for surgical treatment as well as management of surgical complications. They proposed a new classification for periorbital deformities: (1) brow ptosis; (2) upper eyelid infiltration with ptosis; (3) lower eyelid infiltration; (4) lateral canthal disinsertion; and (5) conjunctival and lacrimal gland infiltration. Of 33 patients over age 16 years with orbitotemporal NF, 2 (6%) had bilateral involvement whereas 31 (94%) had unilateral orbitotemporal NF. Previously undescribed findings included severe brow infiltration, lacrimal gland involvement, and functional nasolacrimal duct obstruction.

Optic Pathway Gliomas

Optic pathway gliomas (OPGs) are typically low-grade pilocytic astrocytomas that involve some combination of the optic nerves, chiasm, or optic tracts that occur in about 15% of children with NF1. OPGs are the most common intracranial malignancy in NF1. While most have a benign course, some may manifest as precocious puberty (reviews by Ferner et al., 2007 and Williams et al., 2009).

A longitudinal study of 219 patients with NF1 reported that clinical precocious puberty developed in 7 children, all of whom had optic chiasmal tumors (Listernick et al. (1994, 1995)).

Parazzini et al. (1995) documented spontaneous regression of optic pathway lesions in 4 NF1 patients and cautioned against diagnosis of optic nerve glioma without evidence of progression.

Parsa et al. (2001) observed spontaneous regression of large, clinically symptomatic optic gliomas in 13 patients, 5 with and 8 without NF1. Regression manifested as an overall shrinkage in tumor size or as a signal change on serial MRI. A variable degree of improvement in visual function accompanied regression. The authors concluded that the possibility of spontaneous regression of an optic glioma should be considered in planning the treatment of patients with these tumors.

Balcer et al. (2001) examined the neuroophthalmologic records and brain/orbital MRI scans from 43 consecutive pediatric NF1 patients with optic pathway gliomas. Involvement of the optic tracts and other postchiasmal structures was associated with a significantly higher probability of visual acuity loss. Visual loss was noted in 47% of patients at a median age of 4 years. However, 7% of patients developed initial visual loss during adolescence. The authors recommended close follow-up beyond the early childhood years, particularly for those children with postchiasmal tumor.

Singhal et al. (2002) compared the natural history of sporadic and NF1-associated optic gliomas in a series of 52 patients from northwest Britain. Ages at presentation were similar, but those associated with NF1 were less likely to present with impaired vision. Although NF1 optic gliomas were less aggressive, there was little difference in 5- and 10-year mortality rates between the 2 tumor groups. NF1 optic glioma cases were also at risk of a second primary central nervous system tumor; in 2 of 5 cases this occurred following radiotherapy, suggesting an etiologic link.

Thiagalingam et al. (2004) reviewed the natural history of optic pathway gliomas in 54 patients with NF1. The mean age at the time of diagnosis was 5.2 years, with 32 patients having signs or symptoms at the time of diagnosis. Seventeen patients were diagnosed after the age of 6 years. Twenty-two patients had tumor progression within 1 year of diagnosis and 6 patients showed progression after 1 year. Most conditions were managed conservatively (68.5%). At follow-up, 17 patients (31.5%) had severe visual impairment in their worse eye and 16.7% had bilateral moderate to severe visual impairment. Contrary to previous reports (e.g., Balcer et al., 2001), these results showed that optic pathway gliomas in patients with NF1 often presented in older children and might progress some time after diagnosis. Given the potential for serious visual consequences, the authors stressed the need for regular ophthalmologic monitoring of patients with NF1 for a long duration.

Liu et al. (2004) described the clinical and radiologic features of 7 children with NF1 with optic pathway gliomas involving the pregeniculate optic pathway in addition to the optic radiations. Two of the patients had expanding mass lesions within the white matter of the temporal or parietal lobes, which were histopathologically demonstrated to be pilocytic astrocytomas; the other 5 had radiographic involvement of the optic radiations but did not undergo biopsy. In 3 of the cases, the visual acuity was 20/200 or worse in each eye. Liu et al. (2004) concluded that optic pathway gliomas in NF1 rarely involve the optic radiations, but that optic radiation involvement might signal a more aggressive optic pathway glioma in patients with NF1.

Malignant Peripheral Nerve Sheath Tumors

One of the most clinically aggressive cancers associated with NF1 is the malignant peripheral nerve sheath tumor (MPNST), estimated to occur in 3 to 15% of patients over a lifetime (Knight et al., 1973).

King et al. (2000) reviewed 1,475 individuals with NF1 from a cohort of patients examined by a single investigator, Vincent M. Riccardi, between 1977 and 1996. MPNST was identified in 34 individuals (2%), yielding a relative risk value of 113. Lesions occurred in the limbs in 18 patients (53%), and those with limb lesions survived longer than those with nonlimb MPNSTs. Pain associated with a mass was the strongest suggestion of MPNST development.

Leroy et al. (2001) performed a retrospective study of MPNST in a cohort of 395 patients with NF1 followed for 11 years in a teaching hospital setting. Seventeen patients (4.3%) developed tumors, with a mean age at diagnosis of 32 years (SD = 14 years). Twelve patients had high-grade tumors; all tumors except 1 developed on preexisting nodular or plexiform neurofibromas. Pain and enlarging mass were the first and predominant signs. None of the benign tumors displayed significant p53 (TP53; 191170) staining or p53 mutations. Six of 12 malignant tumors significantly overexpressed p53, and 4 of 6 harbored p53 missense mutations. Median survival was 18 months overall, 53 months in peripheral locations, and 21 months in axial locations. Leroy et al. (2001) concluded that investigations and deep biopsy of painful and enlarging nodular or plexiform neurofibromas should be considered in patients with NF1, and that late appearance of p53 mutations and overexpression precludes their use as predictive markers for malignant transformation.

Evans et al. (2002) ascertained NF1 patients with MPNST in an attempt to assess lifetime risk. They found 21 NF1 patients who developed MPNST, equivalent to an annual incidence of 1.6 per 1,000 and a lifetime risk of 8 to 13%. There were 37 patients with sporadic MPNST. The median age at diagnosis of MPNST in NF1 patients was 26 years, compared to 62 years in patients with sporadic MPNST. In Kaplan-Meier analyses, the 5-year survival after diagnosis was 21% for NF1 patients with MPNST, compared to 42% for sporadic cases. One NF1 patient developed 2 separate MPNSTs in the radiation field of a previous optic glioma.

McCaughan et al. (2007) surveyed Scottish medical records across a 10-year period and identified 14 NF1 patients with a coexistent diagnosis of MPNST. The lifetime risk of developing MPNST was calculated to be 5.9 to 10.3%, and the mean age at diagnosis of the tumors was 42.1 years. Five-year survival after diagnosis of MPNST was significantly lower in NF1 patients compared to patients without NF1 (0% vs 54%, p less than 0.01).

Susceptibility to Other Malignancies

Crowe et al. (1956) found 6 secondary malignant lesions in 168 patients with neurofibromatosis. D'Agostino et al. (1963) discovered 21 cases of secondary neoplasms in his study of 678 cases of neurofibromatosis.

Knight et al. (1973) reviewed 69 patients with single and 45 patients with multiple neurofibromas. Five patients in the group were found to have a total of 11 secondary malignant lesions including 3 fibrosarcomas, 3 squamous cell carcinomas, and 1 neurofibrosarcoma, among other forms. Some earlier studies have reported mainly sarcomas associated with neurofibromatosis.

Clark and Hutter (1982) reported an apparent association between the rare entity juvenile chronic myelogenous leukemia and neurofibromatosis. They suggested that other types of nonlymphocytic leukemia have an increased frequency, but Riccardi (1982) raised the question as to whether these are families with only cafe-au-lait spots.

Kalff et al. (1982) found pheochromocytoma in 10 of 18 NF1 patients with hypertension. Age at diagnosis ranged from 15 to 62 years. The clinical characteristics of the neurofibromatosis did not predict the presence of pheochromocytoma. One patient without pheochromocytoma had coarctation of the aorta and 1 had renal artery stenosis; this patient was described as having the Turner phenotype. At least 2 of the pheochromocytoma patients had renal artery stenosis, and 3 had small bowel and/or stomach neurofibromata. One patient with pheochromocytoma also had hypernephroma with metastases and another had disseminated metastases from an undifferentiated leiomyosarcoma thought to originate from her upper gastrointestinal tract.

Voutsinas and Wynne-Davies (1983) suggested that the risk of malignancy in NF1 had been exaggerated and that the true value was 2.0% (or 4.2% of those over 21 years).

Crawford (1986) reported on a study of 116 NF1 patients under 12 years of age and reviewed the literature. Among the unusual presentations was rhabdomyosarcoma projecting from the urethra in a girl who also had congenital pseudarthrosis of the tibia. Crawford (1986) stated that 'most of the rhabdomyosarcomas associated with neurofibromatosis involve the genitourinary tract.'

Sayed et al. (1987) described malignant schwannoma in 3 brothers who had inherited neurofibromatosis from their mother. Two of the brothers had been reported by Herrmann (1950).

Griffiths et al. (1987) reported 9 cases of NF1 with a carcinoid tumor in the duodenum that had widespread somatostatin (SST; 182450) immunoreactivity. The duodenum was also the primary site in 18 of 20 published NF1 cases with carcinoid tumor. Pheochromocytoma was also present in 6 of the 27 cases with NF1 and duodenal carcinoid tumor. In cases of von Hippel-Lindau syndrome (193300), with which pheochromocytoma also occurs, Griffiths et al. (1987) found no carcinoid tumors, but did find islet cell tumor in association with pheochromocytoma. Swinburn et al. (1988) reported 2 patients with neurofibromatosis and duodenal carcinoid tumor, bringing the total number of cases of this association to 18. Their 2 cases as well as 5 others were positively identified as somatostatinomas. The histologic finding of psammoma bodies is important in the diagnosis of duodenal somatostatinomas. One patient also had a parathyroid adenoma, which was found postmortem.

Although NF1 has been called 'peripheral neurofibromatosis,' it has been associated with tumors of the central nervous system, which include astrocytomas of the visual pathways, ependymomas, meningiomas, and some primitive neuroectodermal tumors. The most common neuroimaging abnormality in NF1 is a high signal intensity lesion in the basal ganglia, thalamus, brainstem, cerebellum, or subcortical white matter referred to as an 'unidentified bright object' (UBO). These UBOs are thought to represent sites of vacuolar change. Molloy et al. (1995) studied 17 NF1 patients with brainstem tumors, which also presented increased T2 signal abnormality on MRI scanning. Fifteen of these 17 patients had neurologic signs and symptoms indicative of brainstem dysfunction and 35% of them had evidence of radiographic tumor progression. In the 2 patients that had partial surgical resection, pathology demonstrated either a fibrillary or anaplastic astrocytoma. As 15 of these 17 patients remained alive after a 52-month follow-up, this suggested that these are much less aggressive than typical pontine tumors which should be distinguished from the UBOs seen elsewhere in the brains of neurofibromatosis patients.

Hunerbein et al. (1996) described a 56-year-old man with NF1 who had had a 6-month history of recurrent epigastric pain and was found to have a multifocal malignant schwannoma of the duodenum causing biliary obstruction.

Sakaguchi et al. (1996) described a 48-year-old man with NF1 and paroxysmal hypertension in progressive respiratory insufficiency. Clinical investigation displayed calcified tumors in the anterior mediastinum and perirenal region. Histologic examination at autopsy revealed composite tumors consisting of pheochromocytoma and malignant peripheral nerve sheath tumor at 2 sites: the left adrenal gland and the region surrounding the inferior vena cava, probably corresponding to the right adrenal gland. In addition, the gastrointestinal tract was involved with mesenchymal tumors showing neurogenic differentiation.

Coffin et al. (2004) reviewed information indicating that children and young adults with NF1 have a higher risk for non-neurogenic sarcomas than the general population, in addition to an increased risk for malignant peripheral nerve sheath tumor. When non-neurogenic sarcomas occur in early childhood, a subsequent malignant peripheral nerve sheath tumor can occur as a second malignant neoplasm, especially after alkylating agent chemotherapy and irradiation. Coffin et al. (2004) presented 4 patients. In 1, embryonal rhabdomyosarcoma was diagnosed at the age of 2 years, and was treated by surgery, radiation, and chemotherapy. A malignant peripheral nerve sheath tumor was detected at the age of 13 years. A second patient likewise had the diagnosis of embryonal rhabdomyosarcoma at the age of 2 years and had the same therapy followed by T-cell lymphoblastic lymphoma at the age of 7 years.

Oguzkan et al. (2006) described 2 cases of NF1 with rhabdomyosarcoma. The first was that of an infant with overlapping phenotypic features of neurofibromatosis and Noonan syndrome (NS1; 163950) (see NFNS, 601321) who presented with rhabdomyosarcoma of the bladder. The second infant likewise exhibited NF1 features and was also associated with bladder rhabdomyosarcoma. Loss of heterozygosity (LOH) analysis of the NF1 gene using 7 intragenic markers and 1 extragenic polymorphic marker detected a deletion in the NF1 gene in the NFNS case associated with bladder rhabdomyosarcoma.

Bausch et al. (2006) reported that 15 (3%) of 565 pheochromocytoma cases in a pheochromocytoma registry had an NF1 mutation. In 10 additional cases contributed specifically for a study of pheochromocytoma in NF1, they found 92% had germline NF1 mutations. The 25 patients with NF1 were compared with patients with other syndromes associated with pheochromocytoma: 31 patients with multiple endocrine neoplasia type 2 (MEN2; 171400) due to mutation in the RET gene (164761); 21 patients with paragangliomas-1 (168000) due to mutation in the SDHD gene (602690); 33 patients with paragangliomas-4 (115310) due to mutation in the SDHB gene (185470); 75 patients with von Hippel-Lindau disease (193300) due to mutation in the VHL gene (608537); and 380 patients with pheochromocytoma as a sporadic disease. The characteristics of patients with pheochromocytoma related to NF1 were similar to those of patients with sporadic pheochromocytoma. There were significant differences between the NF1 group and the other respective groups in the age at diagnosis (von Hippel-Lindau disease and paragangliomas-1); in the extent of multifocal tumors (MEN2, von Hippel-Lindau disease, and paragangliomas-1); and in the extent of extraadrenal tumors (MEN2, von Hippel-Lindau disease, paragangliomas-1, and paragangliomas-4). Patients with NF1 had a relatively high (but not significant) prevalence of malignant disease (12%), second only to that among patients with paragangliomas-4 who had a germline mutation in the SDHB gene (24%). Taken together, 33% of all symptomatic patients with pheochromocytoma in the multicenter, multinational registry carried germline mutations in 1 of the 5 genes, including the NF1 gene.

Vascular Manifestations

Renal artery stenosis due to 'vascular neurofibromatosis' is a relatively common cause of hypertension in patients with NF1. Reubi (1945) first described vascular NF1. Involvement of the heart in neurofibromatosis was described and reviewed by Rosenquist et al. (1970), who also reviewed involvement of the abdominal aorta and renal, carotid, and other arteries.

Salyer and Salyer (1974) found peculiar arterial lesions in 7 of 18 autopsy cases of NF1 at the Johns Hopkins Hospital. They proposed that the pathogenesis of the arterial lesions was proliferation of Schwann cells within arteries with secondary degenerative changes, e.g., fibrosis, resulting in lesions with various appearances.

Among 40 pediatric patients (16 girls and 24 boys), aged 22 months to 17 years, undergoing operation for renovascular hypertension, Stanley and Fry (1981) found that 10 had neurofibromatosis, including 3 with abdominal aortic anomalies. Abdominal aortic coarctation affected 5 other children. Cure of the hypertension was achieved in 34 patients (85%); the condition was improved in 5; and one case was classified as a therapeutic failure. Single cases of renovascular hypertension in neurofibromatosis were reported by Allan and Davies (1970), Finley and Dabbs (1988), and others.

Brunner et al. (1974) described an unusual case of chronic mesenteric arterial insufficiency caused by vascular neurofibromatosis in a 50-year-old man with a 30-year history of chronic malabsorption and chronic small intestinal paralysis. He was said to have no signs of systemic disease or cafe-au-lait spots. Pigmentation of the perioral area and lips of the patient were attributed to longstanding malabsorption syndrome.

Zochodne (1984) reported a 16-year-old NF1 girl with aneurysm of the superior mesenteric artery complicating renovascular hypertension associated with coarctation of the abdominal aorta from above the celiac trunk to above the origin of the inferior mesenteric artery. The coarctation was associated with stenosis of the renal, celiac, and superior mesenteric arteries. The patient had typical skin signs of neurofibromatosis and had had a right below-knee amputation at age 5 for nonunion of a tibial fracture. The mother and 2 sibs were affected. A very similar patient with neurofibromatosis vasculopathy, or vascular neurofibromatosis, was reported by Lehrnbecher et al. (1994). The 4-year-old boy presented with congenital pseudarthrosis of the right tibia, suggesting the vascular origin of this well-known complication, multiple cafe-au-lait spots, short stature, and mild systemic arterial hypertension. The mother and grandmother had NF1. Subsequent complications of the vasculopathy were hypertension, septic infection of an aneurysm in the deltoid muscle, infarction of a segment of colon, sudden appearance of multiple arterial aneurysms, and venous thrombosis. Histologic examination of the bowel specimen confirmed the clinical diagnosis of vascular NF1: the proliferating cells seemed to have originated from myoblasts or myofibroblasts, and not from Schwann cells.

Craddock et al. (1988) reported a 24-year-old white woman with NF1 who had renovascular hypertension resulting from a proximal renal artery stenosis and poststenotic aneurysmal degeneration. Her sister, aged 38 years, presented similarly but without clinical evidence of neurofibromatosis.

Uren et al. (1988) found a congenital left atrial wall aneurysm in a patient with neurofibromatosis; however, the association may have been coincidental. Fitzpatrick and Emanuel (1988) observed the association of typical NF1 with hypertrophic cardiomyopathy in a brother and sister.

Kousseff and Gilbert-Barness (1989) reported what they referred to as 'vascular neurofibromatosis' in 2 patients who as infants developed idiopathic gangrene with vascular changes resembling those of NF1. An additional review of 105 patients uncovered a 27-month-old boy with NF1 and extensive vascular changes with renal hypertension. They discussed the possible relationship to arterial fibromuscular dysplasia. Stanley (1975) found that 5 of 25 children with arterial fibromuscular dysplasia had NF1 as well.

Nopajaroonsri and Lurie (1996) described venous aneurysm, arterial dysplasia, and near-fatal hemorrhages in a 62-year-old who was said to have familial neurofibromatosis (no family history was given). The patient presented with an aneurysm of the internal jugular vein which was associated with dysplasia of cervical arteries. Neurofibromatous tissue was found in the wall of the aneurysm as well as in small veins. During and after surgical excision of the aneurysm, the patient developed massive hemorrhages that required reexploration and evacuation of cervical hematomas. During surgery, bleeding was difficult to control because of excessive friability of blood vessels. Despite the vascular invasion by a tumor, there was no evidence of malignancy or malignant transformation in the patient after a 10-year follow-up.

Because neurofibromin is expressed in blood vessel endothelial and smooth muscle cells, Hamilton and Friedman (2000) suggested that NF1 vasculopathy may result from an alteration of neurofibromin function in these cells.

Riccardi (2000) supported the view that endothelial injury and its repair, which appear to be important in the pathogenesis of atherosclerosis, may also play a role in NF1 vasculopathy. He recommended a regimen of aggressive antihypertensive treatment of children with NF1 in whom either episodic or persistent systemic hypertension is documented. The goal would be to decrease intravascular trauma, based on the supposition that such trauma is directly related to the evolution of the vascular disease in patients with NF1.

Lin et al. (2000) reviewed cases of NF1 and cardiovascular malformations among 2,322 patient records in the National Neurofibromatosis Foundation International Database, collected between 1991 and 1998. Cardiovascular malformations were reported in 54 (2.3%) of the NF1 patients, 4 of whom had Watson syndrome (193520) or neurofibromatosis-Noonan syndrome (NFNS; 601321). Of the 54 patients, 25 had pulmonic stenosis, and 5 had coarctation of the aorta, representing a higher proportion of all cardiovascular malformations than expected. The authors recommended that all individuals with NF1 have careful cardiac auscultation and blood pressure monitoring as part of every NF-related examination.

Hamilton et al. (2001) reported a previously healthy 33-year-old man with NF1 who died suddenly. Autopsy revealed multiple cardiac abnormalities, including evidence of an intramyocardial vasculopathy characteristic of the vascular pathology found in NF1. Other cardiac findings included nonspecific cardiomyopathic changes, myocardial fibrosis, and a floppy mitral valve. The authors emphasized the importance of recognition of vascular lesions in patients with NF1 so that appropriate management can be provided.

Friedman et al. (2002) reviewed cardiovascular disease in NF1. The NF1 Cardiovascular Task Force suggested that all patients with NF1, especially those with Watson or NF1-Noonan phenotypes, have a careful cardiac examination with auscultation and blood pressure measurement.

Tomsick et al. (1976) reported intracranial arterial occlusive disease in NF1. Erickson et al. (1980) described 2 sisters with neurofibromatosis and intracranial arterial occlusive disease leading to the moyamoya pattern of collateral circulation (MYMY1; 252350). Four other members of their sibship of 8, and members of 2 previous generations, including the mother, had neurofibromatosis. Yamauchi et al. (2000) stated that more than 50 cases of the association of NF1 and moyamoya disease had been described, including the cases reported by Woody et al. (1992) and Barrall and Summers (1996). See MYMY2 (607151) for a form of moyamoya disease showing linkage to chromosome 17q25.

Benatar (1994) described a 27-year-old man with neurofibromatosis who presented with 3 intracranial fusiform aneurysms. He referred to 3 previous descriptions of large intracranial fusiform aneurysms in patients with NF1, which he considered to be considerably less common than renal and gastrointestinal vascular lesions in this disorder.

Schievink et al. (2005) detected incidental intracranial aneurysms in 2 (5%) of 39 patients with NF1 who were hospitalized for other reasons. Limiting the patient population to the 22 patients who had a brain MRI resulted in a significantly higher detection rate of 9% compared to 0% in 526 control patients with primary or metastatic brain tumors who underwent brain MRI. The findings suggested that patients with NF1 are at an increased risk of developing intracranial aneurysms as a vascular manifestation of NF1.

Central Nervous System Abnormalities

Adornato and Berg (1977) observed the diencephalic syndrome in 2 infants who had neurofibromatosis and hypothalamic tumors.

Horwich et al. (1983) presented evidence that aqueductal stenosis occurs in neurofibromatosis.

Senveli et al. (1989) reported 6 patients with NF1 who had aqueductal stenosis and hydrocephalus requiring surgical intervention. Ages varied from 14 to 24 years. Twenty-two similar cases were found in the literature.

Winter (1991) described dural ectasia in neurofibromatosis causing bony erosion that was sufficiently severe to destroy spinal stability. Eichhorn et al. (1995) described dural ectasia in a 20-year-old woman with NF1 who presented with back and leg pain. Increasingly severe back pain led to investigations which showed multiple fractures of the pedicles of L1 to L4 with dural ectasia penetrating the body of L2. The transverse diameter of the dura was twice that of the vertebral body at that level, reaching and lifting the psoas.

Mukonoweshuro et al. (1999) reviewed the central nervous system manifestations and neuroradiologic findings in NF1.

Skeletal Manifestations

Skeletal abnormalities in NF1 include short stature, scoliosis, sphenoid wing dysplasia, and tibial pseudarthrosis, a bowing of the long bone that looks like a false joint (reviews by Ferner et al., 2007 and Williams et al., 2009).

Konishi et al. (1991) described a 40-year-old woman with NF1 and typical hypophosphatemic osteomalacia. Bone pain, multiple pseudofractures, marked increase in osteoid by bone biopsy, and hypophosphatemia with renal phosphate wasting were features. Treatment with oral phosphate and vitamin D was effective. They found reports of 34 similar cases and pointed out that of the 67 patients collected by Dent (1952), 2 had neurofibromatosis.

In a father and 3 children by 2 different women, Schotland et al. (1992) described cosegregation of NF1 and osseous fibrous dysplasia. In the 4 individuals with NF1, cafe-au-lait spots and neurofibromata were present in all 4, Lisch nodules and macrocrania in 3, and scoliosis and curvature of the long bones in 2. Schotland et al. (1992) found at least 8 reports of NF1 and osseous fibrous dysplasia associated in individuals but no previous description of a familial association. The osseous dysplasia consisted of multiple lesions at the distal ends of the shafts of the femurs and in the tibias and fibulas, with bowing of the fibulas.

Stevenson et al. (1999) reported a descriptive analysis of tibial pseudarthrosis in a large series of NF1 patients. A male predominance was observed among patients with pseudarthrosis, leading the authors to suggest that male gender may be a susceptibility factor. Examination of the natural history of pseudarthrosis showed that half of the patients who had a fracture sustained it before age 2 years, and that approximately 16% of the pseudarthrosis patients had an amputation.

Long bone dysplasia, seen in 2% (Ferner et al., 2007) to 5% (Stevenson et al., 2006) of patients with NF1, typically involves the tibia and frequently presents with anterolateral bowing that may progress to fracture and nonunion. Tibial dysplasia is most often unilateral, evident in the first year of life, and usually not associated with a neurofibroma at the site, suggesting a random molecular event. Stevenson et al. (2006) documented double inactivation of the NF1 gene in pseudarthrosis tissue, and suggested involvement of the neurofibromin-Ras signal transduction pathway. Prospectively acquired tissue from the pseudarthrosis site of 2 individuals with NF1 did not show typical immunohistochemical features of neurofibroma, but genetic markers spanning the NF1 locus demonstrated loss of heterozygosity. Patient 1 of Stevenson et al. (2006) was a 42-year-old man with a father with NF1 and a brother with NF1 associated with lower limb pseudarthrosis requiring amputation. Patient 2 was a 2-year-old boy whose tibial and fibular bowing presented at birth, with subsequent fibular fracture at age 2 weeks. Clinical findings consistent with NF1 included more than 5 cafe-au-lait macules and tibial pseudarthrosis. The mother had NF1.

Lammert et al. (2006) found significantly lower mean serum levels of 25-hydroxyvitamin D in 55 NF1 patients compared to controls (14.0 ng/ml in patients, 31.4 ng/ml in controls). Among the NF1 patients, there was a highly significant inverse correlation between serum vitamin D concentration and the number of dermal neurofibromas. Lammert et al. (2006) noted that focal osseous abnormalities and decreased bone mineral density are observed in patients with NF1, which may be related to inadequate circulating vitamin D. The relationship of serum vitamin D to neurofibromas was unclear.

Cognitive and Neuropsychologic Manifestations

Twelve NF1 families with 1 affected child, an unaffected sib, and both natural parents were studied by Hofman et al. (1994) to assess the presence of cognitive deficits or learning disability. NF1 children with known intracranial problems were excluded, but family members with known learning disabilities or hyperactivity disorders were not, making some of the results difficult to interpret. Full scale IQs ranged from 70 to 130 among children with NF1 and from 99 to 139 among unaffected sibs. Scores of parents with NF1 ranged from 85 to 114 compared to 80 to 134 in unaffected parents. Children with NF1 showed significant deficits in language and reading abilities compared to sibs, but not in mathematics. They also had impaired visuospatial and neuromotor skills. In 11 of 12 NF1 children but in none of the unaffected sibs, foci of high signal intensity on T2-weighted MRI scan images were observed. A statistically significant correlation was found between lowering of IQ and visuospatial deficits and the number of foci seen on scan.

Legius et al. (1994) studied the neuropsychologic profiles of 46 children with NF1. They found a reduction in total IQ, but a significantly better verbal rating than performance rating in all age groups. Concentration problems were especially significant in children with a higher IQ. Legius et al. (1994) suggested that these children may benefit from the use of Ritalin.

T2 'unidentified bright objects' are seen in 50 to 75% of children with NF1, most frequently in the basal ganglia, corpus cerebellum, and brainstem. Legius et al. (1995) found no difference in the mean intelligence of 18 children with such lesions and 10 neurofibromatosis children who did not show such lesions.

Silva et al. (1997) stated that learning disabilities are said to occur in 30 to 45% of patients with NF1, even in the absence of any apparent neuropathology. The learning disabilities may include a depression in mean IQ scores, visuoperceptual problems, and impairment in spatial cognitive abilities.

Schrimsher et al. (2003) found an association between visuospatial performance deficits and attention deficit-hyperactivity disorder (ADHD; 143465) in patients with NF1.

Unusual Features

Neurofibromata of the intestine are a recognized, though rare, feature of von Recklinghausen neurofibromatosis. Hashemian (1953) reported patients with mild skin changes of NF1 who had intestinal fibromatosis. Neurofibromata of the bowel leading to gastrointestinal bleeding were described by Manley and Skyring (1961) in a patient with striking skin changes. Chu et al. (1999) described a 10-year-old girl with NF1 who had a a 9-month history of anemia and low gastrointestinal bleeding associated with a jejunal leiomyoma.

Massaro and Katz (1966) established the association of interstitial pulmonary fibrosis (fibrosing alveolitis) with NF1 on the basis of studies of 76 patients. Porterfield et al. (1986) described pulmonary hypertension secondary to interstitial pulmonary fibrosis.

Hayes et al. (1961) reported hypoglycemia associated with massive intraperitoneal tumor of mesodermal origin in a patient with typical cutaneous lesions. Unusual clinical manifestations were described by Diekmann et al. (1967): hypertension due to renal artery stenosis, and hypertrophy of the clitoris. Sutphen et al. (1995) described clitoromegaly in 4 patients with NF1 and reviewed the literature documenting 26 NF1 patients with clitoral involvement.

Kurotaki et al. (1993) described the case of a 13-year-old Japanese boy who was found to have small nodules in the lung on chest radiography. He was asymptomatic. Although there was no family history of NF1, he had multiple cafe-au-lait spots over the whole body since birth, and soft subcutaneous tumors of the forehead and back were noticed from the age of 7 years. On biopsy the lung lesions were found to be papillary adenomas of type II pneumocytes. The patient had remained asymptomatic for 6 years thereafter.

Zacharin (1997) reported the unusual occurrence of precocious puberty in a 5-year-old girl and 8-year-old boy with NF1 in whom imaging studies failed to demonstrate any abnormality of the optic tracts or optic chiasm. Previous studies have indicated that optic tract lesions develop at a mean age of 3.6 years, and longitudinal studies have failed to demonstrate symptomatic optic tract tumors occurring after age 6 years. The 2 patients of Zacharin (1997) were aged 11 and 14.7 years at the time of the report.

Bahuau et al. (2001) reported a family with neurofibromatosis type I, in which 2 female children had congenital megacolon due to intestinal neuronal dysplasia type B (601223). The affected infants were found be doubly heterozygous for a mutation in the NF1 gene and in the GDNF gene (600837).

Neurofibromatous Neuropathy

Neurofibromatous neuropathy, a common feature of NF2 but an unusual complication of NF1, is characterized by a distal sensorimotor neuropathy associated with diffuse neurofibromatous change in thickened peripheral nerves (Thomas et al., 1990). NF2-associated neurofibromatous neuropathy is entirely different clinically and histologically from NF1-associated neurofibromatous neuropathy (Sperfeld et al., 2002).

Ferner et al. (2004) described 8 patients with NF1 and neurofibromatous neuropathy among 600 NF1 patients from 1 clinic, thus demonstrating a frequency of 1.3%. The patients had an indolent symmetric predominantly sensory axonal neuropathy and unusual early development of large numbers of neurofibromas. The biopsied nerves showed diffuse neurofibromatous change and disruption of the perineurium. Two patients developed a high grade malignant peripheral nerve sheath tumor. Ferner et al. (2004) pictured the side of the neck of a patient with a thickened greater auricular nerve. They also pictured studies of the lumbar spine showing neurofibromas involving all the nerve roots but not causing cord compression. Disease-causing mutations were identified in 2 individuals (613113.0040-613113.0041) and molecular studies did not reveal any whole gene deletions. Ferner et al. (2004) suggested that the cause of neurofibromatous neuropathy may be a diffuse neuropathic process arising from inappropriate signaling between Schwann cells, fibroblasts, and perineurial cells.

Segmental Neurofibromatosis

Nicolls (1969) described 2 cases of sectorial (or segmental) neurofibromatosis, which he plausibly interpreted as representing somatic mutation. One had a mediastinal neurofibroma and, in the skin area corresponding segmentally to the site of the internal lesion, 5 small neurofibromas. Miller and Sparkes (1977) also reported on this phenomenon.

Zonana and Weleber (1984) illustrated a patient who had multiple cafe-au-lait spots of von Recklinghausen type only on the right side of the body. Iris hamartomata (Lisch nodules) were present in the right eye only. The findings were consistent with a segmental form of NF1.

Riccardi and Eichner (1986) referred to the segmental form as neurofibromatosis type V. Combemale et al. (1994) presented 2 new cases of segmental NF1 and reviewed reports concerning 88 cases. One of their patients was a 71-year-old woman with multiple cutaneous tumors limited to the left side of the trunk, which were present since the age of 41 years.

In a survey of 56,183 young men, aged 17 and 18 years, Ingordo et al. (1995) found 11 cases of NF1 and 1 case of segmental NF. In this group, the relative frequency was 0.02% for NF and 0.0018% for segmental NF. From November 1988 through August 1995, Wolkenstein et al. (1995) saw 308 patients with NF1 according to the criteria of the National Institutes of Health Consensus Development Conference (1988) and 9 patients with segmental NF according to the classification of Riccardi (1982). These findings and those of Ingordo et al. (1995) suggested that segmental NF is about 30 times less frequent than NF type I.

Tinschert et al. (2000) provided molecular confirmation that segmental neurofibromatosis represents a postzygotic NF1 gene mutation. Using FISH, they identified an NF1 microdeletion in a patient with segmental NF in whom cafe-au-lait spots and freckles were limited to a single body region. The mutant allele was present in a mosaic pattern in cultured fibroblasts from a cafe-au-lait spot lesion, but was absent in fibroblasts from normal skin as well as in peripheral blood leukocytes.

Crowe et al. (1956) estimated that about 50% of NF1 patients have new mutations. Crowe et al. (1956) estimated the relative fertility of affected males and females to be 0.41 and 0.75, respectively. Samuelsson and Akesson (1988) estimated that the relative fertility of neurofibromatosis cases is 78% and the mutation rate somewhere between 2.4 and 4.3 x 10(-5).

Miller and Hall (1978, 1978) reported a possible maternal effect on the severity of NF1. They found that patients born of affected mothers had more severe disease than those born of affected fathers. In their series of 62 patients from 54 families, only 16 were new mutations.

Ritter and Riccardi (1985) studied 111 3-generation families with NF1 and found no instance of skipped generation. They suggested that penetrance of autosomal dominant NF1 is complete and that previous impressions to the contrary failed to recognize heterogeneity, minimal clinical expression, and nonpaternity.

Clementi et al. (1990) used the methods of classic segregation analysis to test whether there was a deviation from the expected mendelian segregation rate for NF1 in a sample of 129 Italian sibships. With this approach, they obtained a maximum likelihood estimate of the proportion of sporadic cases, and estimated the mutation rate for NF1 to be 6.5 x 10(-5) gametes per generation.

Jadayel et al. (1990) used molecular methods to identify the parental origin of new mutations in NF1. They found that the new mutation was of paternal origin in 12 of 14 families with NF1. The estimated mutation rate, 1 in 10,000 gametes, was one of the highest for a human disorder (Huson et al., 1989) and suggested that the NF1 gene is large or has some other structural peculiarity. The same bias toward paternal origin of new mutations had been demonstrated for retinoblastoma (180200). Neither disorder shows a paternal age effect in the incidence of mutations. (Riccardi et al. (1984), however, reported an increased paternal age effect.)

In 10 families with an NF1 mutation, Stephens et al. (1992) found that the mutation had occurred in the paternally derived chromosome 17. The probability of observing this result by chance was estimated as less than 0.001, assuming an equal frequency of mutation of paternal and maternal NF1 genes. They suggested a role for genomic imprinting that may either enhance mutation of the paternal NF1 gene or confer protection from mutation to the maternal NF1 gene.

Easton et al. (1993) studied variation in expression of 3 quantitative traits (number of cafe-au-lait patches, number of cutaneous neurofibromas, and head circumference) and 5 binary traits (presence or absence of plexiform neurofibromas, optic gliomas, scoliosis, epilepsy, and referral for remedial education). For cafe-au-lait patches and neurofibromas, correlation was highest between MZ twins, less high between first-degree relatives, and lower still between more distant relatives. The higher correlation between MZ twins suggested a strong genetic component in variation of expression, but the low correlation between distant relatives suggested that the type of mutation at the NF1 locus itself plays only a minor role. All 5 binary traits, with the exception of plexiform neurofibromas, also showed significant familial clustering. The familial effects for these traits were consistent with polygenic effects, but there were insufficient data to rule out other models, including a significant effect of NF1 mutations. There was no evidence of any association between different traits in affected individuals. Easton et al. (1993) concluded that the phenotypic expression of NF1 is to a large extent determined by the genotype at other 'modifying' loci and that these modifying genes are trait specific.

Lazaro et al. (1994) observed a family in which completely normal parents had a son and daughter with a clinically severe form of NF1. The sibs showed no inheritance of paternal alleles for a marker in intron 38 of the NF1 gene, whereas they received alleles from both parents for other NF1 markers. Analysis with probes from this region of the NF1 gene showed a 12-kb deletion involving exons 32 to 39, in the affected offspring. In the father's spermatozoa, 10% were found to carry the same NF1 deletion, but the abnormality was not detected in DNA from his lymphocytes. Thus, this appeared to be an example of gonadal mosaicism.

In a retrospective data analysis of Czech children with NF1, Snajderova et al. (2012) found that the frequency of sporadic cases was 35.6%. The mean NF1 sporadic case paternal age at birth was 32.0 years, compared to 28.8 years in the general population (p less than 0.001). The maternal age of patients was also higher than in the general population, but was not as significant as the paternal age difference. The findings confirmed the association of advanced paternal age with occurrence of sporadic NF1.

Based on the 1988 National Institutes of Health Consensus Development Conference on Neurofibromatosis, the diagnosis of NF1 is made in an individual with any 2 of the following clinical features: (1) 6 or more cafe-au-lait spots, (2) axillary or groin freckling, (3) 2 or more Lisch nodules, (4) 2 or more neurofibromas, (5) optic pathway gliomas (OPGs), (6) bone dysplasia, and (7) a first-degree family relative with NF1 (reviews by Ferner et al., 2007 and Williams et al., 2009).

Crowe et al. (1956) suggested that the presence of 6 spots, each more than 1.5 cm in diameter, is necessary for the diagnosis of neurofibromatosis. Crowe (1964) considered axillary freckling to be an especially useful diagnostic clue. Occasional features included scoliosis, pseudarthrosis of the tibia, pheochromocytoma, meningioma, glioma, acoustic neuroma, optic neuroma, mental retardation, hypertension, and hypoglycemia.

Johnson and Charneco (1970) suggested that the cafe-au-lait spot of neurofibromatosis can be distinguished from the innocent spot that occurs in normal persons and from the pigmented areas of McCune-Albright syndrome (MAS; 174800) by the presence of a large number of DOPA-positive melanocytes that have giant pigment granules in the cytoplasm. The plexiform neuroma is specific to von Recklinghausen disease: only on this feature can the histopathologist make a definitive diagnosis.

Ward et al. (1990) estimated that tightly linked, flanking DNA markers available permitted prediction of NF1 in a child with greater than 98% accuracy. They predicted that even after the NF1 gene is cloned, linkage testing would probably remain important. Linked markers may remain more cost-effective than screening for 1 genetic event among a large number of possible mutations that could be responsible for NF1 in a particular family.

Gutmann et al. (1997) provided guidelines for the diagnostic evaluation and multidisciplinary management of both NF1 and NF2.

Cnossen et al. (1998) reported a 10-year prospective follow-up study of 209 children suspected of having NF1, 150 of whom were ultimately given this diagnosis. Minor disease features of macrocephaly, short stature, hypertelorism, and thorax abnormalities were highly prevalent in children with NF1 and significantly associated with the diagnosis of NF1 at 6 years of age. In addition, children with 3 or more minor disease features were all diagnosed with NF1 under the age of 6 years. Cnossen et al. (1998) concluded that in children aged less than 6 years with insufficient diagnostic criteria, documentation of minor disease features may be helpful in predicting the diagnosis of NF1.

Park and Pivnick (1998) used a protein truncation assay to screen for mutations in 15 NF1 patients and obtained positive results in 11 (73%) of them. Sequencing of cDNA and genomic DNA yielded identification of 10 different mutations. No correlations between genotype and phenotype were apparent.

Ablon (2000) interviewed 18 unaffected parents of an affected child to document their experiences in receiving their child's diagnosis of NF1. The author found that methods of disclosure were often at variance with suggestions made in recent years for conveying 'bad news.' She also found that certain factors assist parents in receiving and more positively adapting to their child's diagnosis. These factors include physicians' attention to the setting and style of disclosure, imparting appropriate and positive information, allowing additional time for careful explanation, and scheduling a follow-up appointment.

DeBella et al. (2000) studied 1,893 NF1 patients under 21 years of age from the National Neurofibromatosis Foundation International Database to determine the age at which the features included in the NIH Diagnostic Criteria appear. Approximately 46% of sporadic NF1 cases failed to meet the NIH Diagnostic Criteria by 1 year of age. Nearly all (97%; 95% CI, 94-98) NF1 patients met the criteria for diagnosis by 8 years of age, and all did so by 20 years of age. The usual order of appearance of the clinical features listed as NIH criteria was cafe-au-lait macules, axillary freckling, Lisch nodules, and neurofibromas. Symptomatic optic glioma was usually diagnosed by 3 years of age, and characteristic osseous lesions were usually apparent within the first year of life.

Ferner et al. (2007) provided guidelines for the diagnosis and management of NF1 according to organ system, as well as suggestions for genetic counseling.

To evaluate the diagnostic performance of NF1-related choroidal abnormalities as a diagnostic criterion of the disease in comparison with the standard NIH diagnostic criteria, Parrozzani et al. (2015) used near-infrared imaging in a study of 140 consecutive pediatric patients (0-16 years of age) with NF1 (at least 2 diagnostic criteria), 50 suspected of having NF1 (1 diagnostic criterion), and 42 healthy individuals. NF1-related abnormalities were detected in 72 affected individuals (60.5%) and in 1 child suspected of having NF1 (2.4%); the feasibility rate of this sign was 82%. No healthy individuals had choroidal abnormalities. Sensitivity, specificity, and positive and negative predictive values of NF1-related choroidal abnormalities were 0.60, 0.97, 0.98, and 0.45, respectively. Compared with standard NIH criteria, the presence of NF1-related choroidal abnormalities was the third parameter for positive predictive value and the fourth for sensitivity, specificity, and negative predictive value. Compared with Lisch nodules, the choroidal abnormalities were characterized by higher specificity and positive predictive value. Parrozzani et al. (2015) concluded that choroidal abnormalities represent a diagnostic sign for NF1 in children.

By linkage analysis of 15 Utah kindreds with NF1, Barker et al. (1987, 1987) found a locus on chromosome 17, about 4 cM from the centromere (lod score of 4.21 at theta = 0.04). There was no evidence for genetic heterogeneity. Seizinger et al. (1987, 1987) presented evidence that the NF1 gene is linked to the nerve growth factor receptor gene (NGFR; 162010) on 17q12-q22 (peak lod score of 4.41 at theta of 0.14). However, crossovers between the 2 loci suggested that a mutation in NGFR was not the fundamental defect.

On the basis of the occurrence of neurofibromatosis and galactokinase deficiency (230200) in a family reported by Fanconi (1933), Stambolian and Zackai (1988) suggested that the NF1 locus may be closely linked to the GALK1 gene (604313) on 17q24. One of the affected sibs in this family was the first enzymatically identified case of galactokinase deficiency (Gitzelmann, 1965). The parents of this sibship were first cousins and the mother had NF1.

Ledbetter et al. (1989) described a patient with NF1 who had a balanced translocation between chromosome 17q11.2 and chromosome 22. Human-mouse somatic cell hybridization studies allowed localization of ERBA1 (THRA; 190120), ERBB2 (164870), and (CSF3; 138970) distal to the 17q11.2 breakpoint, and HHH202 (D17S33) and beta-crystallin (CRYB1; 123610) proximal to the 17q11.2 breakpoint. Schmidt et al. (1987) reported a family in which a mother and 2 children with NF1 had a balanced translocation t(1;17)(p34.3;q11.2). Menon et al. (1989) studied further the translocation t(1;17) described by Schmidt et al. (1987). In a somatic cell hybrid line containing only the derivative chromosome 1, they showed that the breakpoint occurred between SRC2 (164940) and D1S57, which are separated by 14 cM. The translocation breakpoint was located on chromosome 17 between D17S33 and D17S58, markers that also flank NF1 within a region of 4 cM.

Vance et al. (1989) reported linkage studies in 6 multigenerational families with NF1 using 9 markers known to map in the pericentromeric region of chromosome 17. The closest marker was HHH202 (lod score of 3.86). Two-point lod scores for NF1 versus all the markers studied were presented, and the most likely gene order determined. Similar studies were reported by Seizinger et al. (1989), who performed a multipoint linkage analysis using 6 closely linked markers on chromosome 17 (lod score of 3.83 at HHH202). The authors concluded, on the basis of the linkage data, that the NF1 gene maps to the long arm rather than the short arm of chromosome 17.

Further linkage studies involving the NF1 locus and pericentromeric markers on chromosome 17 were reported by Diehl et al. (1989), Mathew et al. (1989), Upadhyaya et al. (1989), Kittur et al. (1989), Goldgar et al. (1989), and Stephens et al. (1989). Goldgar et al. (1989) summarized the results of the international consortium for NF1 linkage. The 8 teams of researchers studied 142 NF1 families with more than 700 affected persons, using 31 markers in the pericentric region of chromosome 17. The best gene order derived from these studies was pter--pA10-41--EW301--cen--pHHH202--NF1--EW206--EW207--EW203-- CRI-L581--CRI-L946--HOX2--NGFR--qter.

Physical mapping data concerning the NF1 region on chromosome 17 were reported by O'Connell et al. (1989), Fountain et al. (1989), Fain et al. (1989), and Upadhyaya et al. (1989).

Wallace et al. (1989) described a NotI fragment from human chromosome 17q11.2 which detected breakpoints in the 2 NF1 patients with translocations involving 17q (Ledbetter et al., 1989; Schmidt et al., 1987). Fountain et al. (1989) mapped a series of chromosome 17 NotI-linking clones to proximal 17q and studied them by pulsed field gel electrophoresis in order to define the region of the breakpoint involved in a 17q11.2 balanced translocation present in 2 NF1 patients. One clone, D17S133, identified the breakpoint in 1 of the 2 patients. A pulsed field map indicated that the breakpoint was within 10 to 240 kb of the cloned segment. Similarly, O'Connell et al. (1989) isolated human cosmids and mapped them to the immediate vicinity of NF1. One cosmid probe demonstrated that the breakpoint in both patients, and presumably the NF1 gene, was contained within a 600-kb NruI fragment.

Yagle et al. (1990) isolated 5 cosmids that mapped directly proximal to 2 NF1 translocations and 11 cosmids that mapped directly distal to them. Of these, 2 cosmids in each region were found to be linked to the disease locus and 3 of these 4 cosmids showed no recombination. One distal cosmid detected the 2 NF1 translocations by pulsed field gel analysis and was used by Yagle et al. (1990) to produce a long-range restriction map that covered the translocations.

Germline Mutations in the NF1 Gene

Wallace et al. (1990) identified a large transcript from the candidate NF1 region on chromosome 17q11.2 that was disrupted in 3 patients with NF1. Two of the patients had previously been reported by Ledbetter et al. (1989) and Schmidt et al. (1987) as having translocations involving t(17;22) and t(1;17), respectively. The third patient was found to have a 0.5-kb insertion. The changes disrupted expression of the NF1 transcript in all 3 patients, consistent with the hypothesis that it acts as a tumor suppressor.

Using pulsed field gel electrophoresis, Upadhyaya et al. (1990) identified a 90-kb deletion in the proximal portion of 17q in 1 of 90 unrelated patients with NF1. Viskochil et al. (1990) detected deletions of 190, 40, and 11 kb in the gene located at the 17q translocation breakpoint in 3 patients with NF1.

In an NF1 patient, Wallace et al. (1991) identified an insertion of an Alu sequence in an intron of the NF1 gene, resulting in deletion of the downstream exon during splicing and a frameshift (613113.0001). Cawthon et al. (1990) identified 2 different point mutations in the NF1 gene (L348P, 613113.0003 and R365X, 613113.0004) in patients with NF1. Upadhyaya et al. (1992) identified multiple germline NF1 mutations (see, e.g., 613113.0006-613113.0009) in patients with NF1.

Heim et al. (1994) stated that although mutations had been sought in several hundred NF1 patients, by August 1994, only 70 germline mutations had been reported in a total of 78 individuals; only the R1947X (613113.0012) mutation had been seen in as many as 6 unrelated patients. NF1 mutations that had been identified included 14 large (more than 25 bp) deletions, 3 large insertions, 18 small (less than 25 bp) deletions, 8 small insertions, 6 nonsense mutations, 14 missense mutations, and 7 intronic mutations. At least 56 (80%) of the 70 mutations potentially encode a truncated protein because of premature translation termination.

Heim et al. (1995) used a protein truncation assay to identify abnormal polypeptides synthesized in vitro from 5 RT-PCR products that represented the entire NF1 coding region. Truncated polypeptides were observed in 14 of 20 patients with familial or sporadic NF1 diagnosed clinically and in 1 patient with only cafe-au-lait spots and no other diagnostic criterion. Mutations responsible for the generation of abnormal polypeptides were characterized by DNA sequencing; 13 previously unpublished mutations were characterized in the 14 individuals. Because the entire NF1 coding region was spanned in each individual, the distribution of NF1 truncating mutations was discerned for the first time: the mutations were relatively evenly distributed throughout the coding region. Upadhyaya et al. (1995) stated that fewer than 90 mutations had been reported to the NF1 mutation analysis consortium and details of only 76 of these mutations had been published. They described 5 new mutations identified by SSCP analysis and heteroduplex analysis, as well as 3 intragenic deletions identified by analyzing families with intron-specific microsatellite markers.

Upadhyaya et al. (1997) screened 320 unrelated NF1 patients for mutations in the GAP (RASA1; 139150)-related region of the NF1 gene, which is encoded by exons 20-27a and has a known biologic function. Sixteen different lesions in the NF1-GRD region were identified in a total of 20 patients. Of these lesions, 14 were novel and together comprised 3 missense, 2 nonsense, and 3 splice site mutations plus 6 deletions of between 1 and 4 bp.

Klose et al. (1998) identified a missense mutation in the NF1 gene (R1276P; 613113.0022) in a patient with a classic multisymptomatic NF1 phenotype, including a malignant schwannoma. The mutation specifically abolished the Ras-GTPase-activating function of neurofibromin, providing direct evidence that failure of neurofibromin GAP activity is a critical element in NF1 pathogenesis. The findings also suggested that therapeutic approaches aimed at the reduction of the Ras-GTP levels in neural crest-derived cells would be effective.

Upadhyaya et al. (2003) described a Portuguese family in which 3 members had clinical features of NF1, but each had a different underlying defect in the NF1 gene; see 613113.0030-613113.0032. The authors speculated about the mechanism of this unusual situation.

Kluwe et al. (2003) examined 20 patients with spinal tumors from 17 families for clinical symptoms associated with NF1 and for NF1 mutations. Typical NF1 features were found in 12 patients from 11 families, and NF1 mutations were identified in 10 of the 11 index patients in this group, including 8 truncating mutations, 1 missense mutation, and 1 deletion of the entire NF1 gene. Eight patients from 6 families had no or only a few additional NF1-associated symptoms besides multiple spinal tumors, which were distributed symmetrically in all cases and affected all 38 nerve roots in 6 patients. Only mild NF1 mutations were found in 4 of the 6 index patients in the latter group, including 1 splicing mutation, 2 missense mutations, and 1 nonsense mutation in exon 47 at the 3-prime end of the gene. The data indicated that patients with spinal tumors can have various NF1 symptoms and NF1 mutations; however, patients with no or only a few additional NF1 symptoms may be a subgroup or may have a distinct form of NF1, probably associated with milder NF1 mutations or other genetic alterations.

In a girl with aniridia (106210), microphthalmia, microcephaly, and cafe-au-lait macules, Henderson et al. (2007) identified heterozygous mutations in the PAX6 (R38W; 607108.0026), NF1 (R192X; 613113.0046), and OTX2 (Y179X; 600037.0004) genes. Her mother, who carried the NF1 and PAX6 mutations, had NF1 with typical eye defects; in addition, although her eyes were of normal size, she had small corneas, and also had cataracts, optic nerve hypoplasia, nystagmus, and mild iris stromal hypoplasia with normal-sized pupils. The proband's father, who had multiple ocular defects (MCOPS5; 610125), had previously been studied by Ragge et al. (2005) and was heterozygous for the OTX2 nonsense mutation. Henderson et al. (2007) noted that the proband's phenotype was surprisingly mild, given that mutations in PAX6, OTX2, or NF1 can cause a variety of severe developmental defects.

Somatic Mosaicism

Colman et al. (1996) reported an NF1 patient who was somatically mosaic for a large maternally derived deletion in the NF1 gene region. The deletion extended at least from exon 4 near the 5-prime end of the gene to intron 39 near the 3-prime end, and included at least 100 kb. Colman et al. (1996) suggested that the deletion occurred at a relatively early developmental time point, since signs of NF1 in this patient were not segmental and both mesodermally and ectodermally derived cells were affected.

Vandenbroucke et al. (2004) described a patient with NF1 manifestations throughout the body, but leaving a few sharply delineated segments of the skin unaffected, suggestive of revertant mosaicism. A large intragenic deletion was found by mutation analysis using long-range RT-PCR. The intra-exonic breakpoints were identified in exon 13 and exon 28, resulting in a deletion of 99,571 bp at the genome level. Analysis of several tissues demonstrated the presence of 2 genetically distinct cell populations, confirming somatic mosaicism for this NF1 mutation. Revertant mosaicism was excluded by demonstrating heterozygosity for markers residing in the deletion region. The findings were significant because the patient was severely and generally affected and could not be distinguished clinically from classic NF1 patients, but showed genetic mosaicism at the cellular level.

Somatic Mutations in the NF1 Gene: Loss of Heterozygosity

Skuse et al. (1989) observed loss of DNA markers from the NF1 region of 17q in DNA from malignant tumors from patients with NF1, compared with DNA from nontumor tissue from the same patients. In hereditary cases, the NF1 allele remaining in the tumor was derived from the affected parent. The findings suggested that malignant tumors in NF1 arise as a result of homozygous deficiency of a tumor-suppressor gene, i.e., loss of heterozygosity. However, LOH was not detected in benign neurofibromas. This finding suggested that neurofibromas are either polyclonal or monoclonal in origin, but arise by a mechanism different from that of NF1 malignancies.

Menon et al. (1990) found no deletions in the proximal region of 17q in NF1-derived benign tumor specimens. However, neurofibrosarcomas from NF1 patients displayed loss of alleles for polymorphic DNA markers on 17p outside the area of mapping of NF1. Since the common region of deletion included the site of the p53 gene (191170), they searched for p53 alterations in neurofibrosarcomas by direct sequencing of PCR-amplified DNA. In 2 of 7 neurofibrosarcomas they found point mutations in exon 4 of the p53 gene.

Shannon et al. (1992) reviewed the occurrence of leukemia in NF1. In 16 of 21 cases of juvenile chronic myelogenous leukemia in children with familial NF1, the genetic disorder was inherited from the mother. Of the 21 children, 17 were boys. Myeloid leukemia developed in 12 boys and 4 girls who inherited NF1 from their mothers, and in 5 boys who inherited the disease from their fathers. Father-to-daughter transmission was not observed. Shannon et al. (1992) found that among 5 children with bone marrow monosomy 7 (Mo7), 3 had NF1 and 2 others had suggestive evidence of NF1. Taken together, the presence of chromosome 7 deletions in the leukemias of children with NF1, a pattern of inheritance favoring maternal transmission of NF1, and the marked predilection for boys to develop leukemia monosomy 7 suggested a multistep mechanism of oncogenesis in which epigenetic factors might play a role.

In a neural fibrosarcoma from a patient with NF1, Legius et al. (1993) found a somatic deletion of the NF1 gene on one chromosome and LOH for all chromosome 17 polymorphisms. Thus, homozygous inactivation of NF1 was demonstrated at the molecular level, providing strong support for the view that NF1 is a tumor suppressor gene.

Shannon et al. (1994) found LOH for NF1 in the bone marrow from 5 of 11 NF1 children with malignant myeloid disorders. In each case, the NF1 allele was inherited from a parent with NF1 and the normal allele was deleted. Loss of constitutional heterozygosity had not been reported in the benign tumors associated with NF1 and had been detected only in a few malignant neural crest tumors and in some tumor-derived cell lines. The data provided evidence that NF1 may function as a tumor-suppressor allele in malignant myeloid diseases and that neurofibromin is a regulator of RAS in early myelopoiesis.

Colman et al. (1995) examined the '2-hit' hypothesis in relation to benign neurofibromas in NF1. Using both NF1 intragenic polymorphisms as well as markers from flanking and more distal regions of chromosome 17, they investigated loss of heterozygosity in 22 neurofibromas from 5 unrelated NF1 patients. Eight of these tumors revealed somatic deletions involving NF1, indicating that inactivation of NF1 is associated with at least some neurofibromas. On the other hand, Stark et al. (1995) found single-cell PCR on neurofibroma Schwann cells and found that both alleles of the NF1 gene were present; i.e., there was no evidence of loss of heterozygosity by a nondisjunction, large deletions, or somatic recombination. They granted that small mutations inactivating the wildtype allele could not be excluded.

Shen et al. (1996) speculated that it may be that a second mutation in another gene is required for genesis of neurofibromas, or that they may arise because of the loss of 1 allele. Another possibility was that a second mutation in the NF1 gene was required.

Although several observations support the contention that the NF1 gene product is a tumor suppressor involved in the RAS signal transduction pathway, mutations had not been identified in both NF1 alleles in dermal neurofibromas until the report by Sawada et al. (1996). Their patient was previously shown to have large submicroscopic deletions of the NF1 locus by both somatic cell hybrid analysis (Kayes et al., 1994) and FISH of lymphoblastoid cells (Leppig et al., 1996). The deletion extended at least 125 kb centromeric and 135 kb telomeric to NF1. As part of her medical care, the patient electively had a scalp neurofibroma removed surgically. Sawada et al. (1996) showed that the tumor DNA harbored a 4-bp deletion in NF1 exon 4b in the other allele. The authors stated that this was the first reported definitive identification of a somatic mutation limited to the NF1 locus in a benign neurofibroma from an NF1 individual in whom the constitutional NF1 mutation was known.

The risk of malignant myeloid disorders in young children with NF1 is 200 to 500 times the normal risk. Side et al. (1997) found that NF1 alleles were inactivated in bone marrow cells from children with NF1 complicated by malignant myeloid disorders. Using an in vitro transcription and translation system, they screened bone marrow samples from 18 such children for NF1 mutations that cause a truncated protein. Mutations were confirmed by direct sequencing of genomic DNA from the patients, and from the affected parents in cases of familial NF1. The normal NF1 allele was absent in bone marrow samples from 5 of 8 children who had truncating mutations of the NF1 gene.

Submalignant tumors occurring in NF1 patients have been found to show loss of heterozygosity consistent with the 2-hit hypothesis of Knudson, with 1 allele constitutionally inactivated and the other somatically mutated. Somatic NF1 deletions in benign neurofibromas were described by Colman et al. (1995) and mutations in both copies of the NF1 gene in a dermal neurofibroma were reported by Sawada et al. (1996). Serra et al. (1997) performed LOH analysis on 60 neurofibromas derived from 17 patients, of whom 9 had a family history of the disease and 8 represented sporadic cases. LOH was found in 25% of the neurofibromas (corresponding to 53% of the patients). In addition, they found that in the neurofibromas of patients from familial cases, the deletions occurred in the allele that was not transmitted with the disease, indicating that both copies of the NF1 gene were inactivated in these tumors. The authors concluded that there appears to be double inactivation of the NF1 gene in benign neurofibromas.

Skuse and Cappione (1997) reviewed the possible molecular basis of the wide clinical variability in NF1 observed even among affected members of the same family (Huson et al., 1989). The complexities of alternative splicing and RNA editing may be involved. Skuse and Cappione (1997) suggested that the classical 2-hit model for tumor suppressor inactivation used to explain NF1 tumorigenesis could be expanded to include post-transcriptional mechanisms that regulate NF1 gene expression. Aberrations in these mechanisms may play a role in the observed clinical variability.

Kluwe et al. (1999) stated that plexiform neurofibroma can be found in about 30% of NF1 patients, often causing severe clinical symptoms. Using 4 intragenic polymorphic markers, they identified LOH in 8 of 14 plexiform neurofibromas from NF1 patients. The findings suggested that loss of the second allele, and thus inactivation of both alleles of the NF1 gene, is associated with the development of plexiform neurofibromas. The 14 plexiform neurofibromas were also examined for mutation in the TP53 gene; no mutations were found.

Eisenbarth et al. (2000) described a systematic approach of searching for somatic inactivation of the NF1 gene in neurofibromas. In the course of these studies, they identified 2 novel intragenic polymorphisms: a tetranucleotide repeat and a 21-bp duplication. Among 7 neurofibromas from 4 different NF1 patients, they detected 3 tumor-specific point mutations and 2 LOH events. The results suggested that small subtle mutations occur with similar frequency to that of LOH in benign neurofibromas and that somatic inactivation of the NF1 gene is a general event in these tumors. Eisenbarth et al. (2000) concluded that the spectrum of somatic mutations occurring in various tumors from individual NF1 patients may contribute to the understanding of variable expressivity of the NF1 phenotype.

Neurofibromas presumably arise from NF1 inactivation in S100+ Schwann cells. Rutkowski et al. (2000) demonstrated that fibroblasts isolated from neurofibromas carried at least 1 normal NF1 allele and expressed both NF1 mRNA and protein, whereas the S100+ cells from 4 of 7 of these same tumors lacked the NF1 transcript completely. The authors were unable to document second NF1 mutations in the S100+ cell lines from tumors, and speculated that additional molecular events aside from NF1 inactivation in Schwann cells and/or other neural crest derivatives contribute to neurofibroma formation.

To identify somatic mutations responsible for tumorigenesis in NF1, John et al. (2000) studied DNA from 82 tumors and blood from 45 patients with NF1. Loss of heterozygosity (LOH) was found in 10 (12%) of 82 tumors studied, and SSCP/heteroduplex analysis identified 2 somatic mutations and 5 novel germline mutations. John et al. (2000) suggested that the low detection rate of somatic mutations might indicate that an alternative mechanism such as methylation is involved in tumor formation in NF1. However, they also acknowledged that mutations might be present but not identified by reason of size, location, or sensitivity of screening method.

Serra et al. (2000) cultured pure populations of Schwann cells (SCs) and fibroblasts derived from 10 neurofibromas with characterized NF1 mutations and found that SCs, but not fibroblasts, harbored a somatic mutation at the NF1 locus in all studied tumors. By culturing neurofibroma-derived SCs under different in vitro conditions, 2 genetically distinct SC subpopulations were obtained: NF1 -/- and NF1 +/-. The authors hypothesized that NF1 mutations in SCs, but not in fibroblasts, correlate with neurofibroma formation and that only a portion of SCs in neurofibromas have mutations in both NF1 alleles.

Serra et al. (2001) pointed out that the large size of the NF1 gene, together with the multicellular composition of neurofibromas, greatly hampers the characterization of the second hit, a somatic NF1 mutation, in these tumors. They presented the somatic NF1 mutation analysis of the whole set of neurofibromas studied by their group and consisting of 126 tumors derived from 32 NF1 patients. They identified 45 independent somatic NF1 mutations, 20 of which were reported for the first time. Among point mutations, those affecting the correct splicing of the NF1 gene were common, coinciding with results reported on germline NF1 mutations. In most cases, they were able to confirm that both copies of the NF1 gene were inactivated. The study of more than 1 tumor derived from the same patient was useful for the identification of the germline mutation, and the culture of neurofibromas with clearing of fibroblasts facilitated LOH detection in cases in which it had otherwise been difficult to determine.

Wiest et al. (2003) performed a mutation screen of numerous neurofibromas from 2 NF1 patients and found a predominance of point mutations, small deletions, and insertions as second hit mutations in both patients. Seven novel mutations were reported. Together with the results of studies that showed LOH as the predominant second hit in neurofibromas of other patients, these results suggested that different factors may influence the somatic mutation rate and thereby the severity of the disease in different patients.

Maertens et al. (2007) reported 3 unrelated patients with NF1: 1 was mildly affected with both neurofibromas and pigmentary skin changes, and 2 had isolated neurofibromas and pigmentary skin changes, respectively, consistent with segmental disease. Detailed molecular analysis of various tissues and cell types showed biallelic NF1 inactivation in Schwann cells from neurofibromas and biallelic NF1 inactivation in melanocytes from cafe-au-lait nodules. The data provided molecular evidence that the distinct clinical picture of the patients was due to somatic mosaicism for the NF1 mutations and that the mosaic phenotype reflected embryonic timing.

Bausch et al. (2007) performed mutation scanning of the NF1 gene and loss-of-heterozygosity analysis using markers in and around the NF1 gene in 37 patients, aged 14 to 70 years, with pheochromocytoma and NF1. Of 21 patients with corresponding tumor available, 67% showed somatic loss of the nonmutated allele at the NF1 locus versus 0 of 12 sporadic tumors (p = 0.0002). Overall, 86% of the 37 patients had exonic or splice site mutations, and 14% had large deletions or duplications; 79% of the mutations were novel. The cysteine-serine rich domain (CSR) was affected in 35%, but the RAS GTPase activating protein domain (RGD) in only 13%. There did not appear to be an association between any clinical features, particularly pheochromocytoma presentation and severity, and NF1 mutation genotype.

Jaffe-Campanacci Syndrome

The Jaffe-Campanacci syndrome (JCS) describes a complex of multiple nonossifying fibromas of the long bones, mandibular giant cell lesions, and cafe-au-lait macules in individuals without neurofibromas. The possibility that JCS is a form of neurofibromatosis type 1 had been suggested. Stewart et al. (2014) performed germline NF1, SPRED1 (609291), and GNAS1 (139320) mutation testing on patients with Jaffe-Campanacci syndrome or Jaffe-Campanacci syndrome-related features. Stewart et al. (2014) also performed somatic NF1 mutation testing on nonossifying fibromas and giant cell lesions. Pathogenic germline NF1 mutations were identified in 13 of 14 patients with multiple cafe-au-lait macules and multiple nonossifying fibromas or giant cell lesions ('classical' Jaffe-Campanacci syndrome); all 13 also fulfilled the NIH diagnostic criteria for NF1. Somatic NF1 mutations were detected in 2 giant cell lesions but not in 2 nonossifying fibromas. No SPRED1 or GNAS1 (exon 8) mutations were detected in the 7 NF1-negative patients with Jaffe-Campanacci syndrome, nonossifying fibromas, or giant cell lesions. Stewart et al. (2014) concluded that their study provided the first proof of specific somatic second-hit mutations affecting NF1 in 2 giant cell lesions from 2 unrelated patients, establishing these as NF1-associated tumors.

Benedict et al. (1968) studied the pigmentary anomaly of neurofibromatosis in relation to that of Albright polyostotic fibrous dysplasia. Gross appearance of the pigmented areas was not always reliable. However, special microscopic studies showed giant pigment granules in malpighian cells or melanocytes of normal skin and of neurofibromatosis spots, but rarely in Albright syndrome.

Fialkow et al. (1971) concluded from analysis of neurofibromas from G6PD A-B heterozygotes with von Recklinghausen disease that each tumor must originate in many cells, perhaps at least 150. Although the benign tumors of neurofibromatosis are multiclonal in nature, the malignant lesion (neurofibrosarcoma) is monoclonal (Friedman et al., 1982).

Schenkein et al. (1974) reported increased nerve growth stimulating activity in the serum of patients with von Recklinghausen disease. Kanter et al. (1980) showed an increase only in antigenic activity of nerve growth factor in central neurofibromatosis and only in functional activity in peripheral neurofibromatosis.

In 8 of 30 unrelated females with NF1, Skuse et al. (1989) found heterozygosity for a PGK (311800) RFLP, which could be used to test for clonality. In all 8 cases the neurofibromas appeared to be monoclonal in origin. These results supported the suggestion that benign neurofibromas in NF1 arise by a mechanism that is different from that of the malignant tumors.

All the lesions of NF1, the benign and malignant tumors, the cafe-au-lait spots, the Lisch iris nodules, etc., are presumably the result of 2 mutations, the inherited mutation and a second mutation on the normal homolog. Collins (1993) suggested that the wide variability of clinical manifestations in members of the same family is related to the element of chance in determining what cells are involved by the second mutation and at what stage of development. The progressive nature of the disorder is also indicated.

Using polyclonal antibodies to the NF1 protein, Koivunen et al. (2000) found increased expression of NF1 protein in cultured human keratinocytes when induced to differentiate in high calcium media. The NF1 protein appeared to be associated with the intermediate filament cytoskeleton and was expressed at highest levels during the period of desmosome formation. Cultured keratinocytes from patients with NF1 showed increased variability in cell size and morphology in comparison to control keratinocytes, suggesting that NF1 mutations may alter the organization of the cytoskeleton. The authors proposed that the NF1 tumor suppressor gene exerts its effects in part by controlling organization of the cytoskeleton during the formation of cellular contacts.

Cook et al. (1998) presented the hypothesis that some haploinsufficiency diseases result from an increased susceptibility to stochastic delays of gene initiation or interruptions of gene expression, events that are normally buffered by increased gene copy number and relatively insensitive to dosage compensation. Kemkemer et al. (2002) applied this line of thought to the tumor suppressor gene NF1 and demonstrated that haploinsufficiency of the gene results in an increased variation of dendrite formation in cultured NF1 melanocytes. These morphologic differences between NF1 and control melanocytes were described by a mathematical model in which the cell is considered to be a self-organized automaton. The model described the adjustment of the cells to a set point and included a noise term that allowed for stochastic processes. It described the experimental data of control and NF1 melanocytes. In the cells haploinsufficient for NF1, Kemkemer et al. (2002) found an altered signal-to-noise ratio detectable as increased variation in dendrite formation in 2 of 3 investigated morphologic parameters. They suggested that in vivo NF1 haploinsufficiency results in an increased noise in cellular regulation and that this effect of haploinsufficiency may also be found in other tumor suppressors.

Littler and Morton (1990) reviewed data from 4 studies on NF1, with the following results: the carrier incidence at birth was 0.0004; the gene frequency was 0.0002; and the proportion of cases due to fresh mutation was 0.56. Lazaro et al. (1994) gave the incidence of NF1 as approximately 1 in 3,500 and stated that about half of cases are the result of new mutations.

Garty et al. (1994) found an unusually high frequency of NF1 in young Israeli adults. They surveyed 374,440 17-year-old Jewish recruits for military service and concluded that 390 of them had NF1. The prevalence was 1.04/1,000 (0.94/1,000 for males and 1.19/1,000 for females), which was 2 to 5 times greater than the previously reported prevalence. NF1 was more common in young adults whose parents were of North African and Asian origin (1.81/1,000 and 0.95/1,000, respectively) and less common in those of European and North American origin (0.64/1,000). All these differences were statistically significant; Garty et al. (1994) suggested that they may be partially explained by the more advanced parental age of the NF group (as suggested by the larger number of children in the North African and Asian families) or by founder effect or both.

Poyhonen et al. (2000) studied the epidemiology of NF1 in northern Finland. The observed overall prevalence of NF1 was 1 in 4,436 and the incidence 1 in 3,647. There was no evidence of geographic clustering of NF1, nor was there any sign of linkage disequilibrium in DNA studies.

Hinrichs et al. (1987) showed that the TAT gene of human T-lymphotropic virus type 1 (HTLV-1) under control of its own long terminal repeat was capable of inducing tumors in transgenic mice. The morphologic and biologic properties of these tumors indicated a close resemblance to NF1. Multiple tumors developed simultaneously in the transgenic tat mice at approximately 3 months of age, and the phenotype was successfully passed through 3 generations. The tumors arose from the nerve sheaths of peripheral nerves and were composed of perineural cells and fibroblasts. However, evidence of HTLV-1 infection in patients with neural and other soft tissue tumors would be needed in order to establish a link between infection by this human retrovirus and von Recklinghausen disease.

Silva et al. (1997) found that heterozygous Nf1-knockout mouse (Nf1+/-) showed a deficit of learning and memory similar to humans with NF1. The deficits were restricted to specific types of learning, were fully penetrant, could be compensated for with extended training, and did not involve deficits in simple associative learning.

Vogel et al. (1999) found that 100% of mice harboring null Nf1 and p53 (191170) alleles in cis developed soft tissue sarcomas between 3 and 7 months of age. The sarcomas exhibited loss of heterozygosity (LOH) at both gene loci, and expressed phenotypic traits characteristic of neural crest derivatives and human NF1 malignancies. Vogel et al. (1999) concluded that their data and those of Cichowski et al. (1999) indicated that an additional mutation in the p53 tumor suppressor gene is required to predispose Nf1+/- mouse neural crest-derived cells to malignant transformation. Vogel et al. (1999) stated that their analyses provided evidence that NF1-associated rhabdomyosarcomas and leiomyosarcomas may be of neural crest origin and provided a possible explanation for the development of malignant Triton tumors, or MTTs. Cell lines isolated from MTTs express both Schwann cell and smooth muscle markers, often in the same tumor cell. The phenotype of these tumors is consistent with immortalization of a pluripotent neural crest stem cell, which under normal circumstances adopts a glial, smooth muscle, or neuronal fate. Unlike humans, mice that are heterozygous for a mutation in Nf1 do not develop neurofibromas.

Cichowski et al. (1999) demonstrated that chimeric mice composed in part of Nf1-/- cells do develop neurofibromas, which demonstrated that loss of the wildtype NF1 allele is rate-limiting in tumor formation. In addition, Cichowski et al. (1999) showed that mice that carry linked germline mutations in Nf1 and p53 develop malignant peripheral nerve sheath tumors, which supported a cooperative and causal role for p53 mutations in malignant peripheral nerve sheath tumor development. Cichowski et al. (1999) concluded that the 2 mouse models, either chimeric for complete loss of Nf1 or carrying Nf1 and p53 LOH, provide the means to address fundamental aspects of disease development and to test therapeutic strategies.

Humans with NF1 have an increased risk of optic gliomas, astrocytomas, and glioblastomas. The TP53 tumor suppressor is often mutated in a subset of astrocytomas that develop at a young age and progress slowly to glioblastoma (termed secondary glioblastomas, in contrast to primary glioblastomas that develop rapidly de novo). Reilly et al. (2000) presented a mouse model of astrocytoma involving mutation of 2 tumor-suppressor genes, NF1 and Trp53 (TP53). that showed a range of astrocytoma stages, from low-grade astrocytoma to glioblastoma multiforme, and thus may accurately model human secondary glioblastomas involving TP53 loss. This was the first reported mouse model of astrocytoma initiated by loss of tumor suppressors, rather than overexpression of transgenic oncogenes.

Costa et al. (2001) generated mice lacking the alternatively spliced exon 23a, which modifies the GTPase-activating protein (GAP) domain of NF1, by targeted disruption. Nf1(23a) -/- mice were viable and physically normal and did not have increased tumor disposition, but showed specific learning impairments. These mice specifically lacked the neurofibromin type II isoform. Costa et al. (2001) found that spatial learning was impaired in Nf1(23a) -/- mice but that additional training alleviated learning deficits. Nf1(23a) -/- mice were impaired in contextual discrimination and had delayed acquisition of motor skills. The Nf1(23a) -/- mutation did not affect all forms of learning. Costa et al. (2001) demonstrated that the type II isoform of neurofibromin is important for brain function, but not for embryologic development or tumor suppression. Their data indicated that the learning deficits caused by mutations that inactivate NF1 in mice and humans are not the result of developmental deficits or undetected tumors. Instead, they suggested that learning deficits in individuals with NF1 are caused by the disruption of neurofibromin function in the adult brain, a finding with important implications for treatment of the learning disabilities associated with NF1. Exon 23a modifies the GAP domain of NF1, indicating that modulation of the RAS pathway is important to learning and memory.

Although approximately 10% of Nf1 +/- mice are prone to the development of juvenile myelomonocytic leukemia, they do not manifest pigmentary abnormalities or develop neurofibromas. Neurofibromin negatively regulates Ras activity in mouse hemopoietic cells through the Kit (164920) receptor tyrosine kinase, which is encoded by the dominant white spotting (W) locus. Ingram et al. (2000) generated mice with mutations at both the W locus (val831 to met, termed W41, which results in an abnormal mottled, white coat color) and the Nf1 gene. Mice homozygous for the W41 mutation and heterozygous at Nf1 had 60 to 70% restoration of coat color. However, Nf1 haploinsufficiency increased peritoneal and cutaneous mast cell numbers in wildtype and W41 mice, and it increased wildtype and W41/W41 bone marrow mast cells in in vitro cultures containing Steel factor (184745), the mouse Kit ligand and a mast cell mitogen. Ingram et al. (2000) proposed that increasing the neurofibromin-specific GAP for Ras activity could be a strategy for preventing or treating the complications of NF1.

Gutmann et al. (1999) reported that astrocytes from mice heterozygous for a targeted mutation in the Nf1 gene (Nf1 +/- astrocytes) showed a cell autonomous growth advantage associated with increased RAS pathway activation. In addition, Gutmann et al. (2001) demonstrated that Nf1 astrocytes exhibit decreased cell attachment, actin cytoskeletal abnormalities during the initial phases of cell spreading, and increased cell motility. Whereas these cytoskeletal abnormalities were also observed in Nf1 -/- astrocytes, astrocytes expressing a constitutively active RAS molecule showed increased cell motility and abnormal actin cytoskeleton organization during cell spreading, but exhibited normal cell attachment. Increased expression of 2 proteins implicated in cell attachment, spreading, and motility were seen in Nf1 +/- and Nf1 -/- astrocytes: GAP43 (162060) and T-cadherin (CDH13; 601364). The authors hypothesized that tumor suppressor gene heterozygosity may result in abnormalities in cell function that may contribute to the pathogenesis of nontumor phenotypes in NF1.

Costa et al. (2002) crossed Nf1 heterozygote mice with mice heterozygous for a null mutation in the Kras gene (190070) and tested the Nf1 descendants. They found that the double heterozygotes with decreased Ras function had improved learning relative to Nf1 heterozygote mice. Costa et al. (2002) also showed that the Nf1 +/- mice have increased GABA-mediated inhibition and specific deficits in long-term potentiation, both of which can be reversed by decreasing Ras function. Costa et al. (2002) concluded that learning deficits associated with Nf1 may be caused by excessive Ras activity, which leads to impairments in long-term potentiation caused by increased GABA-mediated inhibition.

Through use of a conditional (cre/lox) allele, Zhu et al. (2002) demonstrated that loss of NF1 in the Schwann cell lineage is necessary, but not sufficient, to generate tumors. In addition, complete NF1-mediated tumorigenicity requires both a loss of NF1 in cells destined to become neoplastic as well as heterozygosity in nonneoplastic cells, particularly mast cells. Zhu et al. (2002) concluded that the requirement for a permissive haploinsufficient environment to allow tumorigenesis may have therapeutic implications for NF1 and other familial cancers. Zhu et al. (2002) identified a non-cell-autonomous role for the development of tumors in NF1. The onset, growth potential, and multicellular nature of the NF1 -/- neurofibromas was suppressed when the cellular environment retained both functional NF1 alleles. Zhu et al. (2002) ruled out trivial explanations for the observed difference in tumor incidence that relate to the potential relative inefficiency of the Cre transgene. The fact that NF1 +/- mast cells invade preneoplastic nerves and remain present throughout the development of the tumor is in stark contrast to the absence of NF1 +/+ mast cells in the NF1 flox/flox;Krox20-cre hyperplasias that fail to form frank neurofibromas. Zhu et al. (2002) suggested that sensitized heterozygous mast cells homing to nullizygous NF1 Schwann cells in peripheral nerves would create a cytokine-rich microenvironment that is apparently permissive for tumor growth.

Although NF1 is characterized by proliferation and malignant transformation of neural-crest derivatives, affected individuals often have disorders that seem unrelated to the neural crest, including hypertension, renal artery stenosis, increased incidence of congenital heart disease (Friedman et al., 2002), especially valvular pulmonic stenosis, and vascular abnormalities in the CNS known as moyamoya (252350). Attempts to produce animal models of NF1 have been hampered by the fact that inactivation of Nf1 in mice leads to midgestation lethality from cardiovascular abnormalities. These defects include structural malformations of the outflow tract of the heart and enlarged endocardial cushions, which are the anlage of cardiac valves. Using tissue-specific gene inactivation, Gitler et al. (2003) showed that endothelial-specific inactivation of Nf1 recapitulates key aspects of the complete null phenotype, including multiple cardiovascular abnormalities involving the endocardial cushions and myocardium. This phenotype is associated with an elevated level of Ras signaling in Nf1 -/- endothelial cells and greater nuclear localization of the transcription factor NFATC1 (600489). Inactivation of NF1 in the neural crest does not cause cardiac defects but results in tumors of neural-crest origin resembling those seen in humans with NF1. These results established a new and essential role for NF1 in endothelial cells and confirmed the requirement for neurofibromin in the neural crest.

Somatic inactivation of murine Nf1 in Schwann cells is necessary, but not sufficient, to initiate neurofibroma formation (Zhu et al., 2002). Neurofibromas occur with high penetrance in mice in which Nf1 is ablated in Schwann cells in the context of a heterozygous mutant (Nf1 +/-) microenvironment. Mast cells infiltrate neurofibromas, where they secrete proteins that remodel the extracellular matrix and initiate angiogenesis. Yang et al. (2003) showed that homozygous Nf1 mutant (Nf1 -/-) Schwann cells secrete Kit ligand (KITLG; 184745), also known as mast cell growth factor (MGF), which stimulates mast cell migration. They also showed that Nf1 +/- mast cells are hypermotile in response to Kit ligand. Thus, these studies identified a novel interaction between Schwann cells carrying a homozygous Nf1-null mutation and mast cells heterozygous for the Nf1 mutation.

Viskochil (2003) pointed out that Riccardi (1981) had presented an 'NF cellular interaction hypothesis,' implicating that the mast cell is a major player in neurofibroma formation. He posited that 'the mast cell now is seen not as a secondary arrival in a developing neurofibroma but as an inciting factor contributing in a primary, direct fashion to tumor development.'

Tong et al. (2007) investigated the pathophysiology of NF1 in Drosophila melanogaster by inactivation or overexpression of the NF1 gene. NF1 gene mutants had shortened life spans and increased vulnerability to heat and oxidative stress in association with reduced mitochondrial respiration and elevated production of reactive oxygen species (ROS). Flies overexpressing NF1 had increased life spans, improved reproductive fitness, increased resistance to oxidative and heat stress in association with increased mitochondrial respiration, and a 60% reduction in ROS production. These phenotypic effects proved to be modulated by the adenylyl cyclase/cyclic AMP (cAMP) protein kinase A (see 176911) pathway, not the Ras/Raf pathway. Treatment of wildtype D. melanogaster with cAMP analogs increased their life span, and treatment of NF1 mutants with metalloporphyrin catalytic antioxidant compounds restored their life span. Thus, Tong et al. (2007) concluded that neurofibromin regulates longevity and stress resistance through cAMP regulation of mitochondrial respiration and ROS production. They suggested that NF1 may be treatable using catalytic antioxidants.

Yan et al. (2008) stated that osteoclasts from NF1 patients and Nf1 +/- mice show abnormal Ras (see 190020)-dependent bone resorption. They found that Nf1 +/- osteoclast progenitors had elevated Rac1 (602048) GTPase activation. Knockdown of Rac1 in Nf1 +/- mice corrected the osteoclast defects and normalized Erk (see MAPK3; 601795) activation in Nf1 +/- osteoclasts.

Skeletal anomalies, such as short stature or bowing/pseudoarthrosis of the tibia, are relatively common in neurofibromatosis type I. Kolanczyk et al. (2007) created mice with Nf1 knockout directed to undifferentiated mesenchymal cells of developing limbs. Inactivation of Nf1 in limbs resulted in bowing of the tibia, diminished growth, and abnormal vascularization of skeletal tissues, consistent with findings in patients with neurofibromatosis type I. However, fusion of the hip joints and other joint abnormalities were also observed in mutant mice, a finding that had not been reported in patients with neurofibromatosis type I. Tibial bowing was caused by decreased stability of the cortical bone due to a high degree of porosity, decreased stiffness, and reduction in the mineral content, as well as hyperosteoidosis. Accordingly, cultured osteoblasts showed increased proliferation and decreased ability to differentiate and mineralize. The reduced growth in Nf1-knockout mice was due to reduced proliferation and differentiation of chondrocytes.

Using an Nf1 optic glioma (OPG) genetically engineered mouse model, Brown et al. (2010) reported novel defects in nonselective and selective attention without an accompanying hyperactivity phenotype. Specifically, Nf1 OPG mice exhibited reduced rearing in response to novel objects and environmental stimuli. Similar to children with NF1, the attention system dysfunction in these mice was reversed by treatment with methylphenidate (MPH), suggesting a defect in brain catecholamine homeostasis. The attention system abnormality was the consequence of reduced dopamine (DA) levels in the striatum, which was normalized following either MPH or l-dopa administration. The reduction in striatal DA levels in Nf1 OPG mice was associated with reduced striatal expression of tyrosine hydroxylase (TH; 191290), the rate-limiting enzyme in DA synthesis, without any associated dopaminergic cell loss in the substantia nigra. There was a cell-autonomous defect in Nf1 +/- dopaminergic neuron growth cone areas and neurite extension in vitro, which resulted in decreased dopaminergic cell projections to the striatum in Nf1 OPG mice in vivo. The authors concluded that abnormal DA homeostasis is the primary biochemical defect underlying the attention system dysfunction in Nf1 genetically engineered mice relevant to children with NF1.

Although the Elephant Man (Howell and Ford, 1980) has often been thought to have had von Recklinghausen disease, it has been suggested (Pyeritz, 1987) that Proteus syndrome (176920) is a more likely diagnosis. After considering several diagnostic possibilities, Cohen (1988) also concluded that the skeletal findings in Joseph Merrick are most consistent with Proteus syndrome. He pointed out that the 'moccasin' lesions of the feet are particularly characteristic of that disorder. See the study of the case of Joseph Merrick by Graham and Oehlschlaeger (1992).

Ruggieri and Polizzi (2003) found several historical examples of what they interpreted as mosaicism in neurofibromatosis. They suggested that the segmental lesions can be limited either to the affected area showing the same degree of severity as that found in the corresponding nonmosaic trait (type 1 segmental involvement) or may be markedly more pronounced and superimposed on a milder, nonsegmental, heterozygous manifestation of the same trait (type 2 segmental involvement).

Exclusion Mapping Studies

Using RFLPs, Darby et al. (1985) excluded the gene for nerve growth factor-beta (NGFB; 162030) on chromosome 1p13 as the site of the mutation in 4 families with neurofibromatosis type 1.

Family studies by Dunn et al. (1985) excluded close linkage of NF1 (lod score less than -2.0) with 8 markers (ABO, Rh, MNSs, GC, PGP, ACP, GPT, and HP). Negative lod scores at all values of theta were obtained with both GC (on 4) and Se (on 19), which others had proposed were linked to NF. Dietz et al. (1985) excluded linkage of NF with GC. Findings of DiLiberti et al. (1982) brought the total lod score over 3.0 for linkage of NF with myotonic dystrophy (DM1; 160900). However, Huson et al. (1986) excluded linkage with chromosome 19 markers linked to myotonic dystrophy. Thus, the reports of coinheritance of DM and NF could be not be explained by close linkage of the 2 loci.

Korenberg et al. (1989) and Pulst et al. (1990, 1991) studied markers flanking the NF1 locus in multiplex families with achondroplasia (ACH; 100800). By linkage analysis, they excluded the achondroplasia locus from the region between the 2 groups of markers flanking NF1. Thus, the concurrence of achondroplasia and NF1 is a single patient was a matter of chance.