Approach to patients with the neoplasms associated with neurofibromatosis type 1, neurofibromatosis type 2, and schwannomatosis

Introduction

Neurofibromatosis type 1 (NF1), neurofibromatosis type 2 (NF2), and schwannomatosis (SWN) are rare tumor suppressor conditions with estimated incidences of 1:3,000, 1:60,000, and 1:100,000, respectively. Although all three conditions are autosomal dominant and predispose to tumors throughout the central and peripheral nervous systems, the natural history for all of these conditions is highly variable both across and within families ( Table 16.1 ).

For example, NF1 manifests with skin findings, peripheral nerve tumors (e.g., neurofibromas), and, in some cases, malignancy. Skin findings include café-au-lait macules (CALMs) and axillary or inguinal freckling. Tumors may involve the skin (e.g., cutaneous neurofibroma [cNF]), superficial and deep nerves (e.g., diffuse infiltrating, nodular, and [pNFs]), or astrocytes (e.g., optic pathway gliomas [OPGs]). Malignant cancers include malignant peripheral nerve sheath tumor (MPNST), malignant glioma, gastrointestinal stromal tumor (GIST), and juvenile myelomonocytic leukemia (JMML). There is a remarkable range in the frequency of each of these manifestations. It is estimated that up to 99% of adults with NF1 will have cNF, but less than 1% of children with NF1 develop JMML. ^, There is also variability in presentation and manifestations experienced across and within families such that one family member may have a heavy burden of cNF and another may have a paucity of skin findings. Differences even exist in manifestations across stages of development within a given person with NF1.

Like NF1, in NF2 there is variability in the frequency and severity of manifestations; however, there is more consistency within families and more consistent genotype-phenotype relationships. For example, the full complement of tumors seen in NF2 including bilateral vestibular schwannoma (VS), intracranial and spinal meningioma, and spinal schwannomas are found in 77%, 52%, and 65%, respectively, in children with a constitutional truncating mutation in exon 2–13 of the NF2 gene in population studies in England. This was compared with bilateral VS in only 52% of children with “mild” mutations (including certain splice site mutations, small in-frame deletions, and missense mutations) and only 4% having meningioma or paraspinal schwannomas.

Schwannomatosis (SWN) is the most recently identified and least well defined of these tumor predisposition syndromes. SWN is defined by multiple schwannomas throughout the peripheral nervous system most often associated with regional and generalized pain. Importantly, the pain may or may not be associated with a particular neoplasm and the pain is far more prominent than any other neurologic symptoms. People with SWN may have unilateral VS, nonvestibular cranial schwannomas, and meningiomas, but not bilateral VS or ependymomas. By far the most common presentation is paraspinal and peripheral schwannomas. The predominance of peripheral schwannomas, some intracranial schwannomas, and meningiomas can create diagnostic confusion due to the clinical overlap with NF2; however, there are distinct genes associated with SWN (e.g., LZTR1 and SMARCB1 ) and germline NF2 testing is negative in people with SWN.

The high frequency of these conditions as well as the common, progressive, and often severe neurologic and oncologic disability that results from pathognomonic tumors (e.g., gliomas, schwannomas, meningiomas) warrants awareness from neurologists, neurosurgeons, oncologists, and related clinical communities ( Table 16.1 ).

Genetics

NF1, NF2, and SWN are all genetic conditions with autosomal dominant transmission. NF1 is by far the most common of the three syndromes. The condition NF1 is due to a germline mutation in one of the two alleles of the tumor suppressor gene NF1 on chromosome 17q11.2. ^, NF1 is a true tumor predisposition syndrome—germline mutation in NF1 is sufficient to cause the condition, and a “second hit” resulting in loss of function in the second NF1 allele is required for tumor formation. The protein product, neurofibromin, is a regulator of Ras. In the absence of normal neurofibromin, the proto-oncogene Ras is dysregulated and contributes to excessive cell growth. Roughly 50% of people with NF1 have inherited the gene from an affected parent; in the other 50% of patients, the germline mutation is de novo . There is complete penetrance such that all people who have a germline NF1 mutation have the condition NF1. However, the manifestations can vary dramatically even among family members with the same mutation (e.g., variable expressivity).

NF2 is similar to NF1 in that: (1) it is autosomal dominant; (2) heterozygous loss of NF2 results in the condition, but biallelic mutations are required for tumor formation; (3) roughly 50% of people with NF2 have a familial mutation and 50% have a de novo mutation; (4) the location of the NF2 gene (chromosome 22q12.2) and protein product (merlin) are known; and (5) there is complete penetrance but variable expression. Important differences in the genetics of NF1 and NF2 are that NF2 is comparatively rare and the existence of a genotype-phenotype correlation in NF2. Specifically, the type of mutation in the NF2 gene is associated with disease severity to a degree that influences clinical management as detailed above. Hence, genetic testing is warranted if feasible in people who meet the clinical criteria for NF2, as the type of mutation influences management.

SWN is unique from both NF1 and NF2, as the genetic condition can be caused by at least two different genes on chromosome 22q11: LZTR1 and SMARCB1 . It is known that there are people and families who meet the clinical criteria for SWN who test negative for germline mutations or deletions in NF2, LZTR1, and SMARCB1 indicating that additional genes are likely involved in SWN.

Importantly, only roughly 86% of people who meet the clinical criteria for familial SWN test positive for a LZTR1 gene mutation and 48% of people with the familial form of SMARCB1 SWN test positive for the germline mutation. Far fewer people with the de novo form of SWN test positive for a germline mutation of LZTR1 or SMARCB1 mutation. Also unique from NF1 and NF2, there appears to be incomplete penetrance and germline mosaicism with SWN. ^{, ,}

Adding to the complexity is that there is significant overlap between NF2 and schwannomatosis in the types of tumors seen in each condition (e.g., schwannomas and meningiomas; ependymomas are not known to occur in schwannomatosis) and the mutations found within the tumor. ^{, ,} Specifically, the schwannomas in SWN commonly have either three or four molecular losses underlying the tumor including biallelic NF2 loss of function (hits 1 and 2) with loss of function in one or both LZTR1 or SMARCB1 alleles. However, schwannomas in people with NF2 may also have loss of LZTR1 or SMARCB1. ^{, , ,}

Pathophysiology

Neurofibromatosis type 1 . NF1 is the most common neurocutaneous tumor predisposition condition. In fact, it is one of the most common autosomal dominant diseases of the nervous system, with estimates of incidence as high as 1:2600 people, affecting all races, ethnicities, and both sexes equally. The protein product of NF1 , neurofibromin, is a large protein that is a GTPase-activating protein (GAP). ^, The GAP-related domain is the most intensely studied portion of the protein, as this is the portion that acts as a tumor suppressor via reduction of Ras activity. Ras is an oncogene, and in the absence of neurofibromin, it is constitutively activated resulting in excessive stimulation of multiple pro-growth pathways. There are additional important functional regions of the Ras protein including a portion that regulates cyclic adenosine monophosphate (cAMP). cAMP is a ubiquitous mediator of intracellular signaling activated by a wide variety of pathways via the G protein–coupled receptor. Intracellular cAMP activity has been linked to both tumor growth and senescence in NF1.

As a result of abnormal cellular regulation, people with NF1 are prone to multiple tumors. The most common tumors are peripheral nerve sheath tumors (PNSTs), including cNFs that involve the skin and soft tissue or nodular and pNF that originate from deeper nerves. An estimated 30–50% of patients with NF1 develop pNFs. ^{, ,} These tumors can both cause significant neurologic disability and transform into malignant sarcomas (MPNST). ^, Hallmark cutaneous findings include CALMs, skinfold freckling, and cNFs. Additional common manifestations of NF1 include central nervous system (CNS) tumors such as gliomas, which frequently occur along the optic track (OPGs) or in the posterior fossa. Many patients with NF1 have cognitive deficits, bone dysplasias including scoliosis, and ophthalmologic abnormalities. There is also an increased risk of vascular abnormalities and non–nervous system malignancies, such as GISTs, leukemia, and neuroendocrine tumors including pheochromocytomas ( Table 16.1 ). ^,

Neurofibromatosis type 2. NF2 is caused by mutations in the NF2 gene on chromosome 22q11. The NF2 gene encodes the protein Merlin (moesin-ezrin-radixin-like protein) that is best known for its activity as a membrane-cytoskeleton scaffolding protein that operates as a tumor suppressor. In the absence of normal Merlin, there is a predisposition to development of multiple nervous system tumors. The pathognomonic presentation for NF2 is bilateral VS. This is present in the vast majority of people with NF2 by the age of 30 years and often results in hearing loss by early adulthood.

Schwannomas of cranial nerves are not unique to NF2 and can happen sporadically or as part of other conditions such as schwannomatosis. It is when schwannomas involve the bilateral vestibular nerves that NF2 is considered. That said, bilateral VS can be seen in people >60 years old as a matter of chance occurrence of two unique schwannomas given the frequency of sporadic VS in the general population. Furthermore, in some cases, lesions seen involving the vestibular nerve are secondary to other processes such as the sequelae of prior radiation therapy or related to leptomeningeal disease. Hence, a full history and examination is needed before confirming that bilateral VS are due to NF2. In addition to schwannomas of the cranial nerves, people with NF2 have a propensity for schwannomas throughout the peripheral nervous system including the spinal nerves and involvement of more distal peripheral nerves. Intracranial and paraspinal meningiomas are also common in people with NF2. Meningiomas are the most common tumor of the cranium in all populations and can be seen in other conditions such as schwannomatosis and forms of meningiomatosis that are independent, but phenotypically overlap with NF2. Finally, ependymomas, most commonly of the spinal cord, are common in NF2. No one of these tumors is particularly challenging in terms of sequelae or management options in and of itself. However, the multiplicity of the schwannomas, meningiomas, and ependymomas and their progressive nature across a lifetime in NF2 result in dire consequences including progressive severe neurologic disability and ultimately death secondary to the complications of these tumors. ^,

Schwannomatosis . As mentioned earlier, there is significant overlap between schwannomatosis and NF2 in that multiple schwannomas are part of the diagnostic criteria and presentation of both conditions. Furthermore, it has been recently shown that people with confirmed schwannomatosis can develop unilateral VS as well as meningiomas. ^{, ,} Hence, the presence of these lesions is no longer an accurate reason to conclude that a person has NF2 versus schwannomatosis. Increasing evidence indicates that mosaic NF2 is more common than previously suspected and can be very suggestive of schwannomatosis clinically. Finally, although pain is the most common manifestation of schwannomatosis, pain can be a part of NF2 as well. ^, Indeed, there is so much overlap in clinical presentation between these conditions that the diagnostic criteria has undergone revision and clarification many times in the last several years and is currently under review in the hopes of improving clarity about the diagnostic criteria for each condition. However, as above, given that there are people who fulfill current clinical diagnostic criteria who test negative for NF2 , SMARCB1, and LZTR1 , additional criteria are likely to be needed in the future as additional genes associated with schwannomatosis are discovered. ^{, , ,}

Chromosome 22q11. What is shared across NF1, NF2, and schwannomatosis is that the tenants of Knudson’s two-hit hypothesis are met: there is a germline mutation that serves as the “first hit,” but neoplasms form when there is a “second hit” causing loss of heterozygosity. As discussed above, in the case of SWN, there is an additional requirement for somatic loss of both alleles of NF2 , so there is a three or four hit hypothesis for schwannomatosis. Given this clinical and molecular overlap, molecular testing on at least two independent tumors and blood is the ideal approach for confirming the diagnosis of either mosaic or germline NF2 or schwannomatosis.

Clinical cases

The diagnoses of NF1, NF2, and SWN can very often be made clinically, although as there is greater data available about the genotype-phenotype relationships that may influence clinical management and given the increasingly recognized overlap between NF2 and SWN, molecular testing is playing an increasingly important role. In this chapter, we present a series of cases that highlight some of the most common clinical scenarios encountered in the care of people with NF1, NF2, and SWN including management of symptomatic and growing pNF in an adult with NF1, OPG in a child with NF1, non-OPG in a child with NF1, bilateral VS with variable hearing loss in a young adult with NF2, and the diagnosis and management of painful tumors in adults with SWN.

Case 16.1

Approach to an Adult With NF1 Associated Plexiform Neurofibroma

Case . A 21-year-old man with NF1 presented to clinic with increasing pain-limiting activities of daily living. He was diagnosed with NF1 at age 13 based on the presence of >6 CALMs and >10 cNFs. He was subsequently confirmed to have Lisch nodules (i.e., iris hamartomas) and several deep lesions consistent with neurofibroma of the right forearm, bilateral neck, back, and jaw on imaging as well as some lesions palpable on examination beneath the skin ( Fig. 16.1C,D ). By the age of 19 years, he developed progressive pain in the neck, bilateral upper extremities, head, abdomen/pelvis, and at sites of his cNFs and felt that there had been visible growth of several of the deep nodular and cNF. The increased pain had prompted two emergency room visits and the addition of gabapentin and oxycodone as needed for severe breakthrough pain not adequate to manage the multifocal pain. In addition, he reported new symptoms of urinary frequency and hesitancy. On examination, there were no signs of myelopathy or radiculopathy with normal motor, sensory, and reflex examinations. There were many palpable nodules deep to the cutaneous and subcutaneous regions in the extremities, some of which were tender to palpation. There were also >100 skin lesions consistent with cNFs. In the setting of known pNFs and the clinical presentation of increasing pain prompting acute care visits despite daily medication for neuropathic pain, fluorodeoxyglucose positron emission tomography (FDG-PET) was ordered to assess for any evidence of malignant conversion.

FDG-PET showed an area with standardized uptake value maximum (SUVmax) of 4.3 in a lesion within the left neck ( Fig. 16.1A,B ). Whole body MRI (WBMRI) was also recommended, as there were no localizing signs on examination, increasing size of palpable lesions on examination, and the FDG-PET was suggestive but not specific for potential conversion to MPNST. WBMRI is an MRI technique that encompasses head, neck, chest, abdomen, pelvis, and lower extremities MRI (e.g., to the lower leg, Fig. 16.1D ). It allows key anatomic and functional sequences to be performed across a large body area in a single acquisition session. The WBMRI showed extensive replacement of the right masseter muscle by pNF, masses in bilateral foramen at every level of the cervical spine, in the prevertebral space narrowing the airway, throughout bilateral brachial plexi, within the mediastinum, along the intercostal neurovascular bundles, within the retroperitoneum, mesentery, throughout the paraspinal soft tissues, within every visualized sacral neural foramen, extensive intrapelvic plexiform tumor burden particularly along the lumbosacral plexus and smaller pelvic nerves that encompasses the rectum, bladder, and cervix with displacement of the urinary bladder, and the rectosigmoid ( Fig. 16.1D ). Throughout all of these lesions, the target sign was maintained, there was no distinct nodular lesion or suspicious imaging features such as restricted diffusion to indicate malignant degeneration. ^, He was ultimately referred for needle biopsy of the neck lesion with the highest SUVmax on FDG-PET with conscious sedation and MRI-guided biopsy. Pathology confirmed pNF. Based on this reassuring pathology, but ongoing pain despite multiple pain medications and radiographic progression, he was enrolled on a clinical trial with a mitogen-activated protein kinase inhibitor (MEKi). Within 1 month of starting drug, he developed extensive acneiform rash and erythema nodosum-type rash. Oral doxycycline and topical clindamycin and topical corticosteroids were started and the dose of the MEKi was reduced by 25%. Rash was manageable thereafter. Within 2 months of starting the MEKi, the diffuse sites of pain reduced from roughly 8/10 to 3/10 based on self-report. By 8 months of starting the MEKi, MRI showed reduction in size of many of the lesions and no evidence of progression.

Teaching Points . This case highlights several important aspects of NF1 including that (1) new unexplained pain should prompt evaluation for malignant degeneration of a pNF; (2) in patients where MRI with diffusion-weighted sequences and FDG-PET are inconclusive, biopsy may be needed to confirm pathology; and (3) new systemic treatments are being studied and currently under review by the Food and Drug Administration for treating progressive pNF in NF1.

Plexiform neurofibromas . pNFs are common in people with NF1 and may be congenital or start growth early in life, but grow rapidly in childhood. ^, Although they are histologically benign, these tumors often result in disfigurement, pain, and focal and diffuse loss of motor and sensory function, as well as compression of organs, great vessels, and the airway. ^, These functional consequences can be associated with severe morbidity and even mortality. Furthermore, they carry a risk of transformation into MPNSTs for which there is no cure outside of aggressive resection, which is rarely feasible given the deep and infiltrative nature of these tumors. Because of this, MPNST are the leading cause of death for people with NF1. There is an association between burden of neurofibroma and MPNST, and hence, people with a high burden of neurofibroma (as in this case) warrant close clinical and radiographic monitoring. ^, Clinically, changes in the severity or nature of pain or rapid growth warrants evaluation for possible malignant conversion with imaging and biopsy. As illustrated in this case, pain is one of the most common symptoms associated with pNFs, although progressive neurologic dysfunction or compression of critical organs or regions (i.e., airway) related to the tumor location are also indications for treatment. The approach to pain management is multifaceted and includes medications for neuropathic pain, antiinflammatories, and nonpharmacologic approaches. ^{, ,}

Neuroimaging of pNF . MRI is the recommended imaging modality for the diagnosis and surveillance of pNFs. Biologic sequences such as apparent diffusion coefficient (ADC) and diffusion-weighted imaging are increasingly used to distinguish between benign and malignant peripheral nerve sheath tumors. FDG-PET has been the imaging modality recommended for distinguishing between pNF and MPNST; however, there is a relatively high false-positive rate as indicated in this case in which there is high SUVmax (i.e., SUVmax >3) seen in a tumor that is histologically proven to be a be benign neurofibroma.

Classification of benign, atypical, and malignant pNF . There is a recently recognized entity of atypical neurofibroma or atypical neurofibromatous neoplasms of uncertain biologic potential (ANNUBP) that may account for some of the imaging heterogeneity seen in NF1-associated PNSTs. ^, Neurofibromas are notoriously heterogeneous and may have multiple different histologic regions and somatic mutations within a given tumor. There is an increasing understanding of an evolution of benign to malignant neurofibroma that can be occurring within different regions of a single tumor or across different tumors within a given patient with NF1. ^{, , ,} Hence, atypical neurofibromas or ANNUBPs are considered to be precursors to MPNST and, although clear standards for management have not yet been set for these tumors, they warrant consideration of surgical resection when feasible or at the very least close surveillance for rapid growth. ^{, , ,} It is possible that atypical neurofibromas are represented on MRI as distinct nodular lesions with a round/ovoid and well demarcated margin that is distinct from the surrounding pNF; however, work is ongoing to validate this observation. Currently, in a person with NF1 with a growing pNF that is increasingly symptomatic and has findings that could be consistent with malignant degeneration based on a distinct nodular appearance on anatomical MRI, low ADC on diffusion MRI, and elevated SUVmax on FDG-PET, biopsy of the most atypical region on imaging should be pursued if feasible. ^{, ,}

Plexiform vs cutaneous neurofibromas . It is important to distinguish pNF from cutaneous and subcutaneous neurofibromas that can involve the skin. The majority of adults, and many children, with NF1 will develop cNFs ( Fig. 16.2 ). These are tumors that are limited to the skin and, unlike pNFs, have virtually no chance of becoming malignant. As such, cNFs do not require regular surveillance; however, given that they are a major cause of poor quality of life and affect almost all adults with NF1, there are major research efforts underway to identify the most effective currently available therapies and develop new therapies for these tumors.

Treatment options for pNF . The standard of care for symptomatic or growing pNFs has long been surgical resection as this can result in complete resection for some tumors and resolution of symptoms even with minor debulking in others. However, due to the intrinsic involvement of nerve, diffuse infiltration, deep locations, and excessive perfusion, surgery is often challenging or not feasible. There is also no defined role for radiation therapy for these tumors and it is avoided due to concern for secondary malignancies. As such, there is a long history of preclinical and clinical efforts to develop promising medical therapies for pNFs. ^, There are several clinical trials that are showing early signs of reducing tumor volume and potentially improving tumor-associated symptoms for both children and adults with morbid pNFs (NCT03962543, NCT03231306, NCT02407405, NCT02101736, NCT03326388). Given these early successes, there has been enthusiasm by many patients for enrollment on clinical trials and off-label use of various targeted therapies for pNFs. This is reasonable in the right clinical context, and there are increasingly good medical options for both adults and children with pNF. However, each of these agents carries with it potentially serious complications and many low-grade toxicities that require close surveillance and management to make these therapies feasible—as was the case with this patient. ^, We also do not have data about the long-term safety and consequences of these therapies at this time. Furthermore, there is no evidence to date that the drugs developed for pNF are effective for sporadic PNSTs.

Clinical Pearls

• 1.

  In NF1 patients with pNF, changes in the severity or nature of pain or rapid growth warrants evaluation with imaging for possible malignant conversion.

• 2.

  Imaging findings that suggest malignant degeneration include a distinct nodular appearance on anatomical MRI, low ADC on diffusion MRI, and elevated SUVmax on FDG-PET; in such lesions, biopsy of the most atypical region should be pursued if feasible.

• 3.

  It is important to distinguish pNF from cutaneous and subcutaneous neurofibromas that can involve the skin; cNFs have no malignant potential and do not require regular surveillance.

• 4.

  Increasingly, medical options are being investigated for both adults and children with pNF but neurofibroma specialty care centers are recommended for these considerations.

Case 16.2

Diagnosis and Management of NF1 Associated Optic Pathway Glioma

Case . A 17-month-old girl with NF1 underwent an MRI of brain and spine for evaluation of motor developmental delay. Although no cause for motor delay was found, evaluation of her orbits showed tortuosity of bilateral optic nerves, asymmetric enlargement of the intraorbital optic nerves (right greater than left), and bilateral optic tracts along with contrast enhancement in the optic nerves ( Fig. 16.3A, D ). Laboratory evaluation for endocrinopathy (e.g., thyroid stimulating hormone, free thyroxine, growth hormone evaluation) revealed no abnormalities, and a vision test using Teller Acuity Cards showed normal visual acuity for age (20/94 right eye, 20/130 left eye). No surgeries were performed and no therapies undertaken. Three months later, a repeat MRI showed tumor growth in the left optic nerve and chiasm/hypothalamus ( Fig. 16.3B, E ) but vision had improved (visual acuity: 20/63 right eye, 20/94 left eye). At the latest follow-up, 6 months from initial examination, the tumor had continued to grow in the left optic nerve and chiasm/hypothalamus ( Fig. 16.3C, F ) but vision was unaffected (visual acuity: 20/63 right eye, 20/63 left eye). Motor delays improved with physical therapy. Despite tumor growth, she remains unaffected by her tumor and has not started therapy for her OPG, although she remains under close surveillance.

Teaching Points . This case highlights the management options and natural history of OPGs in children with NF1, including the importance of clinical assessment to guide treatment decisions. OPGs are common tumors that threaten vision in NF1. Many of these tumors remain asymptomatic, and routine screening with MRI is not recommended because early identification of tumor does not appear to affect outcomes and may serve only to heighten parental anxiety. ^, Instead, annual surveillance with ophthalmology in children with NF1 is important to identify symptomatic OPG. Once they are identified, children with OPG should be followed closely for tumor growth and new symptoms with MRI and ophthalmology evaluations. Careful surveillance with both MRI and vision testing is imperative to reduce the functional impact of these tumors. Decisions to treat are complex and should be undertaken by a team of experts including oncologists and ophthalmologists with expertise in NF1.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here