NF2-Related Schwannomatosis (Formerly Neurofibromatosis Type 2) and Other Schwannomatoses

Predisposition syndromes for benign nerve sheath and meningeal tumours caused by pathogenic variants in NF2, SMARCB1 or LZTR1, unified since 2022 under the term “schwannomatosis” and distinguished by the gene and molecular tumour pattern.

Literature and regulatory information updated: 6 October 2026.

0. Initial Reference Table

1. Introduction

NF2-related schwannomatosis (NF2-SWN) is an autosomal dominant tumour predisposition syndrome characterised by bilateral vestibular schwannomas, schwannomas of other cranial, spinal and peripheral nerves, meningiomas, ependymomas and specific ocular abnormalities, caused by inactivation of the tumour suppressor gene NF2 [2, 7]. The tumours are histologically benign, but their location and multiplicity explain morbidity and early mortality: bilateral deafness, brainstem compression, hydrocephalus, cranial nerve deficits and myelopathy [1, 2]. “Schwannomatosis” now denotes a group of entities with overlapping phenotypes, the most frequent after NF2-SWN being SMARCB1- and LZTR1-related forms, dominated by painful peripheral and spinal schwannomas without bilateral vestibular schwannomas [3, 4].

The history of these entities is one of nosological confusion. Eduard Sandifort, Professor of Anatomy in Leiden, is believed to have described the first vestibular schwannoma in 1777, and the Edinburgh surgeon Wishart described a man with multiple intracranial nerve tumours in 1822, whose seventh-nerve tumours were reinterpreted by Worster-Drought in 1937 as eighth-nerve tumours [22]. After von Recklinghausen delineated neurofibromatosis type 1 (NF1) in 1882, patients with bilateral acoustic nerve tumours were included in von Recklinghausen disease, largely because Harvey Cushing asserted in 1917 that bilateral nervus acusticus tumours formed part of that disease [22]. The description in 1930 of a large family with mild manifestations and no skin abnormalities was insufficient to separate the conditions but gave rise to “Gardner type” for the mild phenotype, in contrast to severe “Wishart type” [1, 22].

Formal separation occurred in 1987. That year, genetic linkage localised bilateral acoustic neurofibromatosis to a chromosome 22 marker distinct from the NF1 locus on chromosome 17, and the National Institutes of Health (NIH) consensus conference separated NF1 and NF2 [22, 23]. In 1993, two groups independently cloned NF2: Rouleau's group identified a gene encoding a putative membrane-organising protein, with germline mutations in patients and somatic mutations in tumours, while Trofatter's group described a gene resembling moesin, ezrin and radixin, encoding a 587-amino-acid protein with non-overlapping deletions in two families [24, 25]. The UK clinical criteria, known as the Manchester criteria, arose from the clinical study of 150 patients published by Evans and colleagues in 1992 and were revised in 2019 on the basis of 2777 molecularly investigated individuals [26, 27].

Schwannomatosis was proposed as an entity distinct from NF2 in 1996–1997, following descriptions of patients, mostly isolated cases, with painful non-cranial, non-intradermal schwannomas and no vestibular schwannomas on imaging; exclusion of the NF2 locus in families led to identification of SMARCB1 in 2007, in a father–daughter family with an inactivating exon 1 variant, and LZTR1 in 2014, in patients whose schwannomas showed loss of a 22q allele and a different somatic NF2 variant in each tumour [4, 28, 29]. In 2022, an international consensus sponsored by the Children's Tumor Foundation, using a Delphi process and patient participation, updated the criteria and removed “neurofibromatosis 2”, because neurofibromas are not part of the disease and the new name reduces confusion with NF1 and highlights overlap with the other schwannomatoses [4]. GeneReviews adopted “NF2-related schwannomatosis” in its 2023 revision, emphasising the importance of molecular testing to distinguish NF2-SWN from LZTR1-SWN in patients with unilateral vestibular schwannoma and other schwannomas [2].

2. Molecular Genetics and Pathophysiology

2.1. The NF2 gene and merlin

NF2 lies at 22q12.2 and encodes merlin (also called schwannomin), an ERM-family protein with a FERM domain (4.1 protein, ezrin, radixin, moesin) linking the cortical cytoskeleton to membrane proteins [2, 6]. Merlin coordinates growth factor receptor signalling with cell adhesion and restricts contact-dependent proliferation; it acts on several pathways, including PI3K-AKT, RAC-PAK and EGFR-RAS-ERK, and has a critical role in Hippo signalling, mediated by suppression of YAP and TAZ [2]. GeneGuidance curation adds, from OMIM, cooperation with WWC1 in LATS1/LATS2 phosphorylation, possible PI3 kinase inhibition through AGAP2, and suppression of the CUL4A-RBX1-DDB1-DCAF1 ubiquitin ligase complex [6]. The MANE Select reference transcript is NM_000268.4 (ENST00000338641.10), with coding sequence between mRNA positions 367 and 2154, confirmed in VariantValidator [5].

The disease mechanism is loss of function. NF2-SWN tumours follow the two-hit model: a first germline event (or postzygotic event in mosaicism) inactivates one NF2 allele, and a second somatic event inactivates the allele in trans, through loss of heterozygosity of part of 22q or a second point variant differing between anatomically independent tumours [4, 15]. ClinGen dosage curation assigns NF2 haploinsufficiency score 3 and triplosensitivity score 0, so exon or whole-gene deletions are an established mechanism [6]. Attempts to identify the truncated product have been unsuccessful; nonsense-mediated mRNA decay explains the absence of protein but does not explain why truncating variants produce more severe phenotypes than complete deletions, leaving a dominant-negative effect unresolved [2].

NF2-SWN tumours derive from Schwann cells, meningeal cells and glial cells [2]. Approximately 40% of NF2-SWN vestibular schwannomas have a lobular pattern, rare in sporadic tumours; NF2-SWN vestibular schwannomas tend to be more invasive and have a higher fraction of dividing cells, and meningiomas are usually fibroblastic, also with a higher mitotic index than sporadic tumours [2]. No histological differences have been observed between glial tumours in patients with and without NF2-SWN [2]. Somatically, biallelic NF2 inactivation is frequent in sporadic meningiomas, schwannomas and mesotheliomas; an exclusively somatic finding in an isolated tumour does not indicate a hereditary predisposition syndrome [2, 6]. In the ESMO oncology genomics handbook, NF2 alterations in meningiomas reach tier II of ESCAT (ESMO Scale for Clinical Actionability of molecular Targets), meaning an investigational target, without any tier I target [30].

2.2. Variant spectrum, mosaicism and genotype–phenotype correlations

NF2 sequencing detects approximately 75% of pathogenic variants in probands, and deletion/duplication analysis (or chromosomal microarray) detects approximately 20% [2]. The 2022 consensus describes an approach combining next-generation sequencing, multiplex ligation-dependent probe amplification (MLPA) for single- or multi-exon changes and high-resolution karyotyping for rearrangements such as translocations interrupting the locus or ring chromosome 22, the latter particularly in de novo cases with severe learning difficulties; this approach detects approximately 96% of germline variants in the second generation of families with typical NF2-SWN and bilateral vestibular schwannomas [4]. In the 2023 English epidemiological study, nonsense variants were most frequent (173/697, 24.8% of patients with an established variant), but accounted for only 18/235 (7.7%) inherited variants, and missense variants had the highest proportion of familial association (56%) [8].

Somatic mosaicism is the most important molecular feature of NF2-SWN. Kluwe and colleagues detected blood variants in 52% of 233 founders with bilateral vestibular schwannomas and estimated the mosaicism rate at 24.8% (58/233); Evans and Wallace estimated 33% in 2009 [1, 31]. With next-generation sequencing in 1055 de novo cases, proven or presumed mosaicism was 22.0% (232/1055), but non-mosaic heterozygous variants were found in only 36.7% (387/1055); applying second-generation detection rates, the probable mosaicism rate was 59.7%, ranging from 21.7% in patients with bilateral vestibular schwannomas before 20 years to 80.7% in those presenting at 60 years or older [32]. GeneReviews summarises that 25% to 50% of individuals with a de novo variant are mosaic, and the 2025 epidemiological review estimates that approximately 60% of non-familial cases have mosaicism [2, 22]. Differences between estimates reflect era, technical sensitivity (exon screening versus deep sequencing), inclusion of tumour analysis and age at presentation, and the estimates are not interchangeable. According to the latest English study, approximately 25% of living patients inherited the disease from an affected parent, 30% are de novo heterozygotes and 45% are mosaic [22].

Genotype–phenotype correlations are among the most robust in tumour predisposition syndromes, and intrafamilial variability is much lower than interfamilial variability [2]. Nonsense and frameshift variants are associated with severe disease irrespective of position, with earlier onset and more intracranial meningiomas, spinal tumours and peripheral nerve tumours; missense variants are usually associated with the mildest form; splicing variants have variable expression and tend to be milder in the 3' half of the gene; large deletions, even extensive ones, are associated with a milder phenotype and do not cause intellectual impairment [2, 33]. In the UK register of 406 patients, carriers of missense or splicing variants, large deletions or mosaicism had significantly fewer tumours than carriers of nonsense or frameshift variants [33]. Variant position determines meningioma risk: in 411 patients with a proven variant, cumulative cranial meningioma risk by 50 years was 70% for exons 1–3, 81% for 4–6, 49% for 7–9, 56% for 10–13 and 28% for 14–15, with cumulative incidence approaching 80% by 70 years in both sexes [11]. A revised genetic severity score further recognises that truncating variants in exons 1 and 14–15 are associated with less severe disease than those in exons 2–13 [22]. Mosaicism, even when detectable in lymphocytes, attenuates the phenotype of truncating variants that would cause severe disease in constitutional heterozygosity [2]. The 2021 French PNDS (French National Diagnosis and Care Protocol) uses the same genetic score to assign patients to three groups (attenuated, when the mosaic event is not detected in blood; classic, mild or moderate; and severe), emphasises that splicing variants in exons 1 to 7 are associated with more severe disease than those in exons 8 to 15 and that missense variants, in-frame deletions that preserve a protein product and whole-gene deletion tend towards the moderate form; it acknowledges, however, that the score still needs refinement for mosaics identified in children, whose course is highly heterogeneous [17].

2.3. SMARCB1- and LZTR1-related schwannomatosis: the three-step, four-hit model

SMARCB1 (22q11.23) encodes a nuclear subunit of the ATP-dependent chromatin-remodelling SWI/SNF (BAF) complex, and LZTR1 (22q11.21) encodes the substrate-recognition adaptor of a CUL3 ubiquitin ligase that ubiquitinates RAS-family GTPases (HRAS, KRAS, NRAS, MRAS) and RIT1, restraining RAS-MAPK signalling [3, 6]. Both genes lie centromeric to NF2 on chromosome 22 [4]. The classic two-hit model cannot fully explain these tumours: the first event is an inactivating germline SMARCB1 or LZTR1 variant; the second is loss of heterozygosity through a contiguous 22q deletion removing the wild-type LZTR1, SMARCB1 and NF2 alleles; the third is a somatic variant in the remaining NF2 allele, on the same chromosome as the first event, resulting in biallelic inactivation of both the predisposing gene and NF2 [3, 34]. Each schwannoma in the same patient therefore has a different somatic NF2 variant, which should not be confused with mosaicism [15]. Reduced penetrance is attributed to the requirement for several successive somatic events or as-yet-unknown modifiers [4].

Schwannomatosis-associated SMARCB1 variants are predominantly hypomorphic and non-truncating (missense, in-frame deletions or duplications, splicing), located mainly at the 5' and 3' ends, including the 3' untranslated region; the most frequent variant is NM_003073.5:c.*82C>T, in the 3' UTR [3]. Some exon 1 frameshift variants (described in the review as c.30delC and c.34C>T, corresponding in the MANE transcript to NM_003073.5:c.30del p.(Phe10LeufsTer6) and NM_003073.5:c.34C>T p.(Gln12Ter), according to VariantValidator) do not abolish the protein, because translation reinitiation at a downstream AUG codon produces an amino-terminally truncated protein expressed in schwannomas [5, 12]. By contrast, rhabdoid tumour predisposition syndrome type 1 (RTPS1) variants are truncating or deletions preferentially affecting exons 2–9 and abolishing the protein [3, 4]. SMARCB1-associated Coffin–Siris syndrome results from probably dominant-negative or gain-of-function missense variants at codons 363–377 of the carboxyl terminus [4, 35]. These three allelic phenotypes may overlap: three families with a truncating variant in which some members had rhabdoid tumours and others schwannomatosis have been reported, as have a rhabdoid tumour survivor who developed schwannomatosis and a patient with Coffin–Siris syndrome and NM_003073.5:c.1121G>A p.(Arg374Gln) who developed schwannomatosis in adulthood [4, 5]. Distal type 3 22q11.2 deletions include SMARCB1 and confer increased malignant rhabdoid tumour risk [35].

The LZTR1 mechanism in schwannomatosis is loss of function, but the gene tolerates loss-of-function variants in the population: in the compilation by Deng and colleagues, 64 of 71 loss-of-function variants reported in patients were classified as pathogenic or likely pathogenic, with a proband frequency of 17.8% (64/359) versus 0.36% potentially loss-of-function variants in gnomAD [36]. The ERN GENTURIS guideline calculates, from a frequency of approximately 1 in 310 in gnomAD and LZTR1-SWN incidence below 1 in 227,000, that fewer than 1% of carriers without a personal or family history will develop schwannomatosis [13]. For this reason, a truncating LZTR1 variant without other evidence will probably remain classified as likely pathogenic; analysis of two schwannomas showing variant retention, 22q loss in trans and different NF2 variants may support reclassification as pathogenic [4]. LZTR1 variants also cause autosomal dominant Noonan syndrome (gain-of-function variants on the Kelch surface) and autosomal recessive Noonan syndrome (biallelic loss-of-function variants), as well as isolated multiple café-au-lait macules; a family with Noonan syndrome and NM_006767.4:c.740G>A p.(Ser247Asn) was reported, in which a 53-year-old woman also had multiple schwannomas [3–5, 35]. Somatic LZTR1 variants occur in approximately one-fifth of glioblastomas, without hereditary implications [3].

2.4. Other schwannomatosis mechanisms

22q-related schwannomatosis is defined by loss of heterozygosity of the same 22q markers in two anatomically distinct schwannomas, with a different NF2 variant in each tumour, without a detectable variant in unaffected tissue or a germline DGCR8 variant [4]. DGCR8 (22q) was proposed as a fourth gene from a three-generation family and a sporadic case with multiple schwannomas and multinodular goitre carrying c.1552G>A p.(Glu518Lys), with the same three-event, four-hit mechanism; no other variants were found in 31 LZTR1-, SMARCB1- and NF2-negative cases, so it does not appear to be a major locus [4, 37]. Rare families with schwannomas due to p14 and p16 loss (CDKN2A deletion) and an uncorroborated COQ6 variant have also been described [4]. Approximately 20% to 30% of families with schwannomatosis have no germline SMARCB1 or LZTR1 variant, indicating other genes [4].

3. Clinical Spectrum and Manifestations by System

3.1. Natural history of NF2-SWN

Mean age at manifestation onset is 18 to 24 years, ranging from birth to 70 years, and almost all patients develop bilateral vestibular schwannomas by 30 years [2]. In the study of 368 patients by Baser and colleagues, mean age at first symptoms was 22 years and at diagnosis 27 years, and 31.5% were diagnosed before 20 years [1]. In the 1992 clinical study of 150 patients, mean onset age was 21.57 years (n = 110), no patient presented after 55 years, 44% of the 100 patients examined by the authors presented with deafness (unilateral in 35 of 44), 10% with deafness and tinnitus, 12% with muscle weakness or atrophy, 8% with seizures, 8% with vertigo, 2% with numbness or paraesthesia and 1% with blindness, and 11 were diagnosed asymptomatically through screening [26]. Some patients present young and have an aggressive course; others present in middle age with slow progression. The Oxford Desk Reference recognises two classic subtypes: mild (Gardner, type A), with onset after 15 years and usually after 25, symptoms mainly due to bilateral vestibular schwannomas, minimal cutaneous manifestations and rare other central nervous system tumours; and severe (Wishart, type 2B), with onset before 30 years and usually before 25, with numerous meningiomas and spinal tumours [1]. The French PNDS (French National Diagnosis and Care Protocol) ranks the presenting signs in decreasing order of frequency: unilateral or bilateral hearing loss, focal motor deficit, tinnitus, balance disturbance, epileptic seizures, sensory deficit and visual disturbance; about 18% of patients express the disease in childhood, and the absence of a family history in half of cases delays diagnosis, particularly in children and young adults [17].

The following table collates frequencies from different series; denominators and ascertainment methods are not equivalent, and percentages should neither be added nor compared directly.

3.2. Vestibular and other schwannomas

Initial symptoms of vestibular schwannoma are tinnitus, hearing loss and balance dysfunction; onset is usually insidious, although hearing loss may be sudden, presumably due to tumour-induced vascular compromise [2]. Patients frequently report difficulty using the telephone with one ear or unsteadiness when walking at night or on uneven ground; over time, tumours extend medially into the cerebellopontine angle and, untreated, compress the brainstem and cause hydrocephalus, although significant facial paralysis is rare even with large tumours [2]. Schwannomas may affect other cranial and peripheral nerves, more frequently sensory than motor [2]. Vestibular schwannoma growth rates are generally higher in younger patients but vary widely, even between relatives of similar ages [1]. Children and young adults with an apparently isolated vestibular or other cranial nerve schwannoma should be considered at risk of de novo NF2-SWN, frequently mosaic [2].

The French PNDS describes the eighth nerve tumours of NF2-SWN as cochleovestibular schwannomas, because they differ from sporadic tumours in their multicentric origin (73%), infiltration of neighbouring nerve and bone structures, very frequent invasion of the cochlear fossette and labyrinth (88%), frequent origin in the cochlear nerve (24%) and the presence of small facial nerve schwannomas; these factors explain the higher recurrence rate and the difficulty of preserving hearing and facial function after surgery [17]. Up to 40% of patients do not yet have bilateral disease at diagnosis [17]. Growth ranges from 0.4 to 10 mm per year, and about 25% of tumours grow more than 3 mm per year, mainly in younger patients, with no relationship between the growth rates of the two sides or between size and deafness [17]. Hearing loss and/or tinnitus reveal the disease in 60% of adults; deafness is present in 62% at diagnosis, but only in 34.7% of tumours smaller than 1 cm³, is sudden in 15% and preceded by tinnitus in 53%; in a large series, tumour growth was observed in 31%, 64% and 79% of tumours at 1, 2 and 3 years, but hearing deterioration in only 5%, 13% and 16%, and other series describe useful hearing maintained in 65% to 74% of patients followed for more than 6 years [17]. Balance disorders occur in 14.6%, and facial paresis or spasm usually indicate a large, rapidly growing tumour or a schwannoma of the facial nerve itself [17]. On magnetic resonance imaging, the FLAIR sequence frequently (69%) shows cochlear hypoxic injury, and labyrinthine invasion is visible in 14.6% [17].

Peripheral nerve schwannomas manifest as swelling along the course of a nerve and/or pain or paraesthesia in its territory, worse at night in 25%; the palpable tumour is firm and mobile perpendicular to the nerve, Tinel's sign is present in 65% of sporadic schwannomas and preoperative motor deficit occurs in only 12% [17]. On magnetic resonance imaging, the schwannoma is an ovoid mass centred on the nerve trunk, hypo- or isointense on T1, hyperintense on T2 and with intense, homogeneous gadolinium enhancement; multinodular forms account for 10% in NF2-SWN [17]. Ultrasound, including portable ultrasound, serves as the first-line examination in clinic, and electromyography is not systematic, although it may show associated peripheral neuropathy [17].

3.3. Spinal tumours and meningiomas

At least two-thirds of patients develop spinal tumours, frequently the most difficult to manage; the most common are intravertebral schwannomas arising from the dorsal root, extending medially and laterally in a dumbbell shape [2]. Intramedullary tumours (ependymomas and, very rarely, astrocytomas) occur in 5% to 33% of patients; most patients with spinal cord involvement have multiple tumours, many asymptomatic [2]. NF2-SWN spinal ependymomas are classically non-progressive [15]. Approximately half of patients have meningiomas in cross-sectional studies, but lifetime risk approaches 80%; most are intracranial, less frequent at the skull base than supratentorially, and fibroblastic [2]. Orbital meningiomas may compress the optic nerve and cause visual loss, and skull-base meningiomas may cause cranial neuropathies, brainstem compression and hydrocephalus [2]. Meningiomas are frequently the first tumour detected in children with early presentation [12, 15].

The French PNDS adds data on intracranial meningiomas, present in about half of patients: mean maximum diameter of 14 mm at diagnosis, location on the falx cerebri in half of cases, a mean of three meningiomas per patient and a higher proportion of high-grade tumours among operated patients than in the population without NF2-SWN (grade 2 in 29% and grade 3 in 6%) [17]. Their presence is associated with more intracranial schwannomas, spinal and cutaneous tumours, more operations over a lifetime and higher mortality [17]. Mean growth is 1.5 mm per year, but saltatory, with prolonged periods of quiescence, so that excision must be weighed according to the overall course of the disease, general condition and symptoms; meningiomas arising de novo during follow-up grow faster (6.6 mm per year), are accompanied by peritumoral oedema and more often warrant resection [17]. They present with epileptic seizures, focal deficit and intracranial hypertension, owing both to volume and to infiltration and occlusion of the venous sinuses [17]. An intracranial meningioma in a child, or multiple meningiomas in an adult, should trigger clinical, radiological and genetic assessment for NF2-SWN [17].

According to the same PNDS, spinal tumours occur in 60% of patients and mark a severe form; intramedullary ependymomas affect 33% to 53%, with a predilection for the upper cervical cord and the cervicomedullary junction, and the authors argue that they should constitute a separate diagnostic criterion, with a high positive predictive value in the presence of a unilateral vestibular schwannoma or multiple meningiomas, which the 2022 criteria went on to establish [4, 17]. Cauda equina schwannomas, sometimes arranged like a string of beads along the roots, progress little and rarely require surgery; a spinal tumour discovered in a child should systematically raise the possibility of NF2-SWN [17].

3.4. Ocular, cutaneous and non-tumour neurological manifestations

One-third of patients have reduced visual acuity in one or both eyes [2]. Posterior subcapsular lens opacity is the most common ocular finding, rarely progresses to visually significant cataract and may precede vestibular schwannoma symptoms; wedge-shaped cortical cataracts may be congenital and associated with amblyopia [2]. In children, NF2-SWN may present with embryonal, posterior subcapsular and cortical opacities, sometimes with retinal hamartomas, nerve palsies, schwannomas or café-au-lait macules [38]. Other findings include thickened optic nerves, third cranial nerve palsy, epiretinal membranes and retinal tufts on optical coherence tomography [2]. Persistent hyperplastic primary vitreous has been described in a father and son [2]. In adulthood, loss of facial, trigeminal and intermediate nerve function, particularly after surgery, causes corneal problems; intracranial and intraorbital tumours cause reduced acuity and diplopia [2]. Optic nerve sheath meningiomas may cause visual loss in early childhood, and combined deafness and visual loss in severe cases profoundly compromises communication [22]. The French PNDS attributes to juvenile (posterior subcapsular) cataract a frequency of about 90%, higher than series with specialist assessment (60% to 80%), and considers it one of the simplest means of identifying affected relatives; epiretinal membranes and retinal hamartomas, identified by optical coherence tomography, occur in 40% to 50%, mainly with onset before 20 years [17, 22]. Some children have severe amblyopia without a tumour; the tumours that compromise vision are optic nerve sheath meningiomas, spheno-orbital meningiomas (with exophthalmos), cavernous sinus tumours and oculomotor nerve schwannomas (with diplopia); in addition there is optic atrophy, exposure keratopathy due to facial palsy and papilloedema due to intracranial hypertension, often without headache, which may lead to permanent visual loss if untreated [17]. The only malignant neoplasms described in the PNDS are the rare MPNST after vestibular schwannoma radiosurgery and the rare mesotheliomas in patients exposed to asbestos [17].

Cutaneous manifestations are subtler than in NF1. Approximately 70% of patients have cutaneous schwannomas, but only 10% have ten or more; intradermal schwannomas and plaques (frequently plexiform schwannomas histologically, sometimes present from birth) may be the presenting feature in childhood [2, 22]. Café-au-lait macules occur in approximately 43% of patients but rarely number more than three [38].

Mononeuropathy is characteristic particularly in childhood, manifesting as facial paralysis with usually partial recovery, strabismus due to third-nerve palsy or foot or wrist drop, potentially mimicking poliomyelitis; in adulthood, progressive polyneuropathy not directly related to tumour masses is recognised, confirmed in sural nerve biopsies [2]. Renal vascular disease resembling that of NF1 has been reported in a single patient, probably coincidentally; renal disease observed in NF2-SWN patients is mainly associated with bevacizumab [2].

3.5. Paediatric NF2-SWN and differences by age at diagnosis

Paediatric NF2-SWN has a multisystem presentation, high tumour burden and significant physical and psychological morbidity; visual impairment and non-tumour dermatological and neurological manifestations are common, and genetic testing characteristically identifies a germline variant, frequently truncating [39]. In a US multidisciplinary clinic (2021–2024), the 32 patients diagnosed at 30 years or older were less likely to undergo genetic testing (53% versus 85%), had lower diagnostic yield (65% versus 85%), fewer anatomical regions affected (2.3 versus 3.4) and fewer tumours (7 versus 12) than the 39 diagnosed earlier, with a similar distribution of loss-of-function variants [40]. A clinical severity classification applied to 168 German patients classified 48% as mild (grade 1), 43% as moderate (grade 2) and 10% as severe (grade 3), with substantial inter-rater agreement (Fleiss' kappa 0.62) and correlation with hospital admissions, number of operations and implant dependence [41].

3.6. SMARCB1- and LZTR1-related schwannomatosis

Presentation most frequently occurs between the second and fourth decades, with localised or diffuse pain or an asymptomatic mass; focal weakness or muscle atrophy is rarely the sole sign [3]. Schwannomas mainly affect peripheral and spinal nerves, predominantly lumbar; cranial nerve involvement is rare and, when present, mainly affects the trigeminal nerve [3]. Pain is the dominant comorbidity, affecting 67% to 94% of patients with non-NF2 schwannomatosis, is not always localised to a schwannoma and may behave as systemic neuropathic pain; an association between high tumour volume and severe pain has been observed [3, 12]. In a whole-body MRI study of 51 patients, 36 (71%) had internal nerve sheath tumours, 81% of which were discrete [3]. Meningiomas have been described only in SMARCB1-SWN and are infrequent even there (approximately 5% in the AACR table); most patients with multiple meningiomas have no SMARCB1 variant [2, 3, 15]. Malignant transformation into malignant peripheral nerve sheath tumour (MPNST) is a risk in SMARCB1-SWN, with four patients reported even without radiation, and rhabdoid tumours and atypical teratoid/rhabdoid tumours have rarely been reported in SMARCB1-SWN families [3, 15]. Unilateral vestibular schwannoma occurs in approximately 10% of patients with LZTR1-SWN and does not appear to occur more frequently in other forms [13]. Spinal tumours were more frequent in LZTR1 (5/5) than SMARCB1 (6/15) in a small series [12]. Patients with LZTR1-SWN and a somatic SH3PXD2A-HTRA1 fusion had more pain [15]. Café-au-lait macules, skinfold freckling and the cutaneous schwannomas typical of other forms are uncommon [3].

The segmental form, with schwannomas limited to one limb or fewer than five contiguous spinal segments, affected 30% (27 of 87) in one series; germline LZTR1 variants were found in 33% to 40% of these patients, and surgery controls pain but with frequent recurrence (5 of 9) or consistent development of new tumours (4 of 4) [13]. Life expectancy is not usually reduced: in the Manchester region, mean age at death was 76.9 years in schwannomatosis versus 66.2 years in NF2-SWN (p = 0.004) [9]. Emotional impact is substantial, including suicide risk, justifying annual psychological assessment [13].

The French PNDS for non-NF2 schwannomatoses (2025) places diagnosis usually between 20 and 40 years, with pain (46%), a palpable mass (27%) or both (11%) as initial symptoms, sensorimotor deficits rare at presentation and schwannomas of peripheral nerves (89%), spinal (74%) and, rarely, cranial (8%) [18]. Familial forms, with at least one affected first-degree relative, account for only 15% of the total, which, together with the interval of years that may separate the first two operations when tumour burden is low, explains the diagnostic delay [18]. Chronic pain, reported by 68% of patients, is more often independent of the schwannomas, focal, multifocal or diffuse, paroxysmal or continuous, and neuropathic in character; the PNDS recommends the DN4 questionnaire for screening, the 0 to 10 numerical rating scale for intensity and the NPSI (Neuropathic Pain Symptom Inventory) to characterise symptoms, and intensity is greater in carriers of an LZTR1 variant and in patients with a high total tumour volume [18]. In the surgical series of the French reference centre, 65% of operated schwannomas involved large nerve trunks and 16% intramuscular and 16% subcutaneous branches, the lower limb was more often affected (52%) than the upper limb (30%) and the nerves most affected were the ulnar and the tibial; intradermal forms are rare and point rather towards NF2-SWN [18]. Spinal schwannomas, present in more than half of patients (33% to 100% depending on the series), predominate in the lumbar segment (56.9%, versus 31.5% thoracic and 11.6% cervical), unlike sporadic schwannomas, which are more frequent in the cervical segment (46.3%); mean age at first intervention is 44 years, segmental forms may account for one-third of operated spinal schwannomas and all patients with multiple spinal operations have non-segmental forms [18]. Intracranial schwannomas affect about 10% of patients, most often the trigeminal nerve (8%); unilateral vestibular schwannoma occurs in about 15% of patients with LZTR1-SWN, and patients with multiple intracranial schwannomas without vestibular schwannomas probably have mosaic NF2-SWN [18]. In a retrospective series of 86 patients, the median was two operations per patient, 68% had chronic pain and 30% depression [18]. Meningiomas (5%, versus 50% in NF2-SWN) arise only with SMARCB1 variants, mainly exon 2 missense variants in families, are located mostly on the falx, are usually single and fibroblastic grade 1, and appear to be of low aggressiveness, unlike sporadic meningiomas with somatic SMARCB1 alteration, which are more often atypical [18]. Glioblastoma has been described in rare families with LZTR1-SWN, and the PNDS suggests that overlap with Noonan syndrome may be underdiagnosed because it is not sought in the history [18]. Polyneuropathy confirmed by electroneuromyography is very rare (1 of 20 and 3 of 14 patients in two series), and reduced intraepidermal small-fibre density on skin biopsy (91% to 97% in possibly overlapping series from a single team) has debatable specificity and relationship with pain; for this reason, neither electroneuromyography nor skin biopsy is systematic without suggestive clinical signs [18].

Frequencies from the Manchester series (Evans 2018), as reproduced by Kehrer-Sawatzki and Cooper in 2025 [9, 12].

4. Clinical Recognition and Warning Signs

NF2-SWN should be suspected in children with two or more of the following: schwannoma at any site, including intradermal; skin plaques present at birth or in early childhood; meningioma, particularly non-meningothelial; wedge-shaped cortical cataract; retinal hamartoma; mononeuropathy, particularly with facial paralysis, foot or wrist drop or third-nerve palsy [2]. In adults, suggestive findings are bilateral vestibular schwannomas; unilateral vestibular schwannoma with two of meningioma, schwannoma, glioma, neurofibroma or cataract (subcapsular opacities or wedge-shaped cortical cataract); and multiple meningiomas with unilateral vestibular schwannoma or two of schwannoma, ependymoma or cataract in an individual under 40 years [2]. Identification of a pathogenic NF2 variant in tumour tissue and an affected first-degree relative increase the probability [2]. When evaluating adults with ataxia, MRI showing bilateral acoustic neuromas indicates NF2-SWN, and deafness with cataract should prompt consideration of this disease [38].

Age is the main risk modifier with apparently isolated tumours. A patient under 30 years with symptomatic unilateral vestibular schwannoma has a high risk of developing a contralateral tumour and should be monitored; after 30 years the risk is very low [2]. In the population analysis by Evans and colleagues, germline NF2 testing is justified in sporadic unilateral vestibular schwannoma only before 20 years unless there are other manifestations, and should ideally start with the tumour [42]. An isolated meningioma before 25 years should prompt genetic assessment [2]. In the Manchester study of young people with a solitary tumour before 25 years, 16 of 42 patients with meningioma (38%) and 27 of 135 with schwannoma (20%) had a predisposing variant; including retrospectively identified patients, 25 of 63 with solitary meningioma (40%) had a constitutional NF2 variant and 9 (14%) a SMARCE1 variant, and 24 patients with spinal schwannoma (55%) had a constitutional variant versus 20 (18%) with cranial schwannoma [43].

Warning signs include those suggesting hydrocephalus or brainstem compression by cerebellopontine angle or skull-base tumours, visual loss due to orbital meningiomas and loss of lower cranial nerve function, which may lead to aspiration pneumonia, a frequent terminal cause of death [2, 22]. In a two-centre French study of 551 patients followed between 1994 and 2026, 33 (6.0%) developed confirmed intracranial hypertension, predominantly with moderate to severe genetic phenotypes (83%), due to hydrocephalus (45.5%), venous outflow obstruction (36.4%) or tumour mass effect (15.2%); 30% had residual visual impairment, 6% blindness and 12% died, and previous vestibular schwannoma surgery was associated with less frequent headache, so classic symptoms may be absent [44]. Patients with vestibular tumours should be warned about insidious balance problems and underwater disorientation, potentially causing drowning [2]. In schwannomatosis, rapid schwannoma growth, intractable pain or functional loss should raise suspicion of malignant transformation, particularly in SMARCB1 variant carriers [3, 13].

The French PNDS (French National Diagnosis and Care Protocol) organises emergencies into two pathways, which the attending physician must recognise: neurological emergencies (intracranial hypertension, motor deficit, gait disturbance) require referral to a local neurosurgery service, with a time frame dictated by the speed of symptom onset; neurosensory emergencies (sudden deafness or hearing loss, particularly in the only functioning ear, and visual disturbance) require urgent otorhinolaryngology or ophthalmology consultation to confirm the diagnosis and start symptomatic treatment [17]. In non-NF2 schwannomatosis, the PNDS sets as a rule that the diagnosis should be considered as soon as a patient has two schwannomas, so that any patient with a peripheral schwannoma should be asked about previous schwannomas; an isolated schwannoma, intra- or extracranial, does not warrant investigation for schwannomatosis [18]. An intracranial schwannoma or meningioma in a child or young adult should also prompt radiological and genetic assessment for schwannomatosis [18].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Evolution of clinical criteria

The first diagnostic criteria derive from the 1988 NIH consensus conference and the 1992 UK study; Manchester criteria extended NIH criteria to patients without a family history with multiple schwannomas and/or meningiomas who did not yet have bilateral vestibular schwannomas, increasing sensitivity without loss of specificity [4, 22]. The version reproduced in the Oxford Desk Reference establishes diagnosis with any of the following: bilateral vestibular schwannomas confirmed histologically or by gadolinium-enhanced MRI; a parent, sibling or child with NF2 and unilateral vestibular schwannoma or two of meningioma, glioma, schwannoma, posterior subcapsular lens opacities or cerebral calcification; unilateral vestibular schwannoma with two of these tumours or findings; or multiple meningiomas with one of glioma, schwannoma, posterior subcapsular lens opacities or cerebral calcification [1]. The book itself warns that no criteria set was adequate at initial assessment to diagnose patients without bilateral vestibular schwannomas, particularly without a family history [1].

The study of 2,777 molecularly analysed individuals (1,361 meeting criteria and 1,416 tested with incomplete criteria) demonstrated the shortcomings of these criteria: the items “glioma” and “neurofibroma” were unhelpful; “ependymoma” had a positive predictive value of 100% and 67.7% confirmed diagnosis; patients with bilateral vestibular schwannomas alone aged 60 years or older had the lowest confirmation rate (6.6%) and a positive predictive value of 80%; siblings as the sole first-degree relative had a positive predictive value of 0%; and seven individuals with unilateral vestibular schwannoma and two or more non-dermal schwannomas with LZTR1-SWN reduced this combination's positive predictive value to 67% [27]. In 205 Manchester patients presenting with unilateral vestibular schwannoma, 83 (40.7%) developed a contralateral tumour; of the 70 meeting criteria with unilateral vestibular schwannoma and two non-intradermal schwannomas, 5 (7%) had an LZTR1 variant, and among the 50 not progressing to bilateral disease, germline LZTR1 variants (5) were as frequent as NF2 variants (5), with NF2 mosaicism the most frequent cause [45]. The revised Manchester criteria (2019) replaced “glioma” with “ependymoma”, removed “neurofibroma”, introduced molecular criteria and a 70-year limit for isolated bilateral vestibular schwannoma, and required negative LZTR1 testing when unilateral vestibular schwannoma accompanies two or more schwannomas [4, 22]. The international consensus rejected the age limit to avoid discouraging investigation of older patients [22]. The French NF2 PNDS (French National Diagnosis and Care Protocol) (2021), which predates the consensus, retains the 1992 Manchester criteria (with “glioma”, understood as intramedullary ependymoma, and without molecular criteria) and merely announces the ongoing revision of nomenclature; the schwannomatosis PNDS (2025) already adopts the designation “NF2-related schwannomatosis” and the 2022 molecular classification, so that, in French practice, the 2022 international criteria are the current reference [4, 17, 18].

5.2. 2022 international criteria

For SMARCB1-SWN and LZTR1-SWN, diagnosis requires at least one pathologically confirmed schwannoma or hybrid nerve sheath tumour and a pathogenic gene variant in unaffected tissue, or the same pathogenic SMARCB1 or LZTR1 variant in two schwannomas or hybrid tumours; if a likely pathogenic variant is identified, tumour analysis may aid reclassification [4]. 22q-SWN requires loss of heterozygosity of the same 22q markers in two anatomically distinct schwannomas with a different NF2 variant in each, undetectable in unaffected tissue, and exclusion of NF2-SWN, SMARCB1-SWN, LZTR1-SWN and a germline DGCR8 variant [4]. Without genetic testing, SWN-NOS is diagnosed with two or more lesions consistent with non-intradermal schwannomas on appropriate imaging and pathological confirmation of at least one schwannoma or hybrid tumour; SWN-NEC additionally requires negative testing of unaffected tissue and at least two tumours [4]. The consensus also adopted “at risk of schwannomatosis” (schwannoma or hybrid tumour with a first-degree relative with SWN-NOS, or two or more schwannomas without pathological confirmation) and “carrier” (asymptomatic individual with a germline pathogenic SMARCB1 or LZTR1 variant) [4].

Mosaicism is confirmed by an allele fraction clearly below 50% in clinically unaffected tissue, or absence of the variant in that tissue with the same variant in two or more anatomically independent lesions; laboratories should define and report quantitative orthogonal confirmation criteria distinguishing germline variants from mosaicism [4]. Detection of an isolated pathogenic variant is not a standalone diagnostic criterion: an NF2 variant found through exome or genome sequencing in an individual without manifestations should be investigated using orthogonal methods to distinguish a constitutional variant from mosaicism, a somatic variant or clonal haematopoiesis; NF2-SWN is diagnosed in an asymptomatic individual only if the variant was inherited from an affected parent [4]. An incidental SMARCB1 or LZTR1 variant in an asymptomatic individual is insufficient to diagnose schwannomatosis, and such individuals should be referred to specialist centres [4]. In variant classification, benign, likely benign and uncertain-significance variants cannot be used diagnostically; likely pathogenic variants correspond to greater than 90% certainty and, with compatible clinical findings, justify a working diagnosis of “probable schwannomatosis”, which testing relatives or two tumours may confirm [4]. A ClinGen variant curation expert panel for NF1, NF2, SMARCB1 and LZTR1 has been established [4].

5.3. Molecular testing: strategy and limitations

Blood test sensitivity depends on presentation: approximately 95% in generations following a de novo index case versus only 37% in sporadic cases [22]. A negative blood result therefore does not exclude NF2-SWN, and tumour tissue analysis may be necessary; when analysing a tumour, both NF2 alterations must be identified, assessing loss of heterozygosity, and blood can then be tested to determine which is constitutional [2]. Testing the tumour first is a reasonable option in apparently sporadic cases [1]. In non-NF2 schwannomatosis, two-tumour analysis identified identical NF2 variants in 37% to 57% of sporadic patients meeting clinical criteria with negative blood testing, confirming mosaic NF2-SWN [4]. In a French series of 22 patients with multiple peripheral schwannomas and 69 nodules analysed, 11 had LZTR1-SWN (50%), 3 possible LZTR1- or SMARCB1-SWN, 7 mosaic NF2-SWN (32%) and 1 22q-SWN; patients with multinodular schwannomas had NF2 mosaicism in 5 of 9 cases, whereas mononodular cases more frequently had LZTR1-SWN (9 of 13) [47]. In sporadic schwannomatosis, SMARCB1 and LZTR1 explain only approximately 30% to 40% of cases, and 70% to 85% of familial cases [13]. The French schwannomatosis PNDS breaks these proportions down by gene: in sporadic forms, which account for 85% of cases, SMARCB1 explains 8% to 10% and LZTR1 25% to 30%; in familial forms, SMARCB1 explains 40% to 45% and LZTR1 30% to 35% [18]. The French NF2 PNDS states that NF2 testing in blood identifies the variant in 90% of familial forms and in only 25% to 60% of sporadic forms, that about 80% of variants are point variants (nonsense, missense, frameshift or splicing) and 20% deletions or duplications of one or more exons or of the whole gene, and that next-generation sequencing in blood detects mosaics with an allele fraction above 5% [17]. In tumours, the second event is, in most cases, a 22q11.2–q12.1 deletion; analysis of a single tumour may reveal candidate variants that are then sought again in blood with sensitivity down to allele fractions below 1%, and, without an operated tumour, tissues of embryological origin different from the haematopoietic, such as buccal mucosa or urinary tract cells, can be studied [17, 18]. The French panel includes SMARCE1, for the differential diagnosis with meningiomatosis [17]. When tumour analysis finds no alteration of NF2, SMARCB1 or LZTR1, allelic imbalance of informative polymorphic variants of 22q11.21–q12.2 makes it possible to distinguish 22q-SWN (with loss of heterozygosity) from schwannomatosis not related to 22q (without loss) [18]. A negative result, even after tumour analysis, does not exclude an alteration not detected by the method used; in that case, both PNDS documents allow, within French clinical care, trio genome sequencing (sporadic index case and parents) or comparison of the blood genome with the tumour genome [17, 18]. GeneReviews table 1 attributes 60% of LZTR1/SMARCB1-SWN to SMARCB1 and 40% to LZTR1, with deletions/duplications reported in a single individual for each gene, including an approximately 7.7 kb SMARCB1 duplication containing exon 7 [3].

Genetic reports should state the gene, transcript with version (NM_000268.4 for NF2, NM_003073.5 for SMARCB1, NM_006767.4 for LZTR1, all MANE Select), HGVS variant nomenclature, zygosity and allele fraction, tissue analysed, ACMG/AMP classification and technical limitations, including sensitivity for mosaicism [4, 5]. A variant of uncertain significance neither establishes nor excludes the diagnosis [3]. The proportion of variants detectable by each method does not equal clinical test sensitivity, which depends on the prior probability of mosaicism.

5.4. Practical algorithm

Initial evaluation after NF2-SWN diagnosis includes neurological examination by a clinician experienced in the disease, brain MRI from 10–12 years, hearing assessment with brainstem auditory evoked responses, skin examination, complete ophthalmological examination and genetic counselling [2]. The Oxford Desk Reference adds a three-generation family history with targeted enquiry about neurofibromatosis, deafness and neurosurgery, skin examination for plaques, café-au-lait macules, cutaneous neurofibromas and schwannomas, ophthalmological assessment for posterior subcapsular and cortical cataracts and hamartomas, gadolinium-enhanced cranial and spinal MRI and audiometry [1]. The AACR update (2025) recommends baseline contrast-enhanced brain MRI with high-resolution internal auditory canal imaging (preferably at 3 T, 1–3 mm slices), spinal MRI, whole-body MRI, audiology and ophthalmology at diagnosis or 10 years, whichever is later [15]. In non-NF2 schwannomatosis, the ERN GENTURIS guideline recommends brain MRI with slices of at least 3 mm and preferably ≤1 mm through the internal auditory canals to exclude bilateral vestibular schwannomas, brain, spinal and whole-body MRI in SMARCB1-SWN, and high-resolution brain MRI with internal auditory canal slices below 3 mm, spinal and whole-body MRI in LZTR1-SWN, and advises against positron emission tomography for diagnosis or surveillance [13].

The French NF2 PNDS requires a complete cochleovestibular assessment, because the disease threatens cochlear and vestibular function on both sides: pure-tone and speech audiometry; auditory evoked potentials whenever hearing allows, as a marker of cochlear nerve injury and of the chances of hearing preservation; and videonystagmography with caloric tests, video head impulse test and otolithic evoked potentials, to assess the residual function of both vestibules and detect signs of central compression [17]. Caloric and otolithic tests are normal in 62.5% of small tumours, possibly arising from the cochlear nerve, and the assessment should be repeated after each therapeutic step to measure central vestibular compensation [17]. An entirely normal assessment in the presence of an internal auditory canal tumour should raise suspicion of a facial nerve schwannoma and prompt electromyography of the seventh nerve with blink reflex study, because this hypothesis changes the indication and the strategy [17]. Whole-body magnetic resonance imaging is not recommended routinely in the NF2 PNDS, but may be offered to young patients with significant disability or with truncating variants predicting a high tumour burden, since overall tumour burden has been correlated with quality of life [17].

In non-NF2 schwannomatosis, the French PNDS proposes localised magnetic resonance imaging of the schwannoma, brain magnetic resonance imaging with thin T2 and gadolinium-enhanced T1 slices through the internal auditory canals (to exclude bilateral vestibular schwannomas and look for meningiomas) and whole-spine magnetic resonance imaging down to the cauda equina; it recommends, if available and not systematically, whole-body imaging, by magnetic resonance imaging with apparent diffusion coefficient mapping, possibly combined with positron emission tomography (PET-CT or PET-MRI), although none of the three techniques has shown superiority [18]. Schwannomas are frequently hypermetabolic on PET (mean SUVmax 4.2 ± 1.9), which should not be mistaken for malignancy when PET is the first imaging performed [18]. Syndromic schwannomas cannot be distinguished from solitary ones either on imaging or on histology; multinodular forms account for about 15% of operated schwannomas; high-resolution magnetic resonance neurography shows intrafascicular microlesions in the nerves of all limbs, even in segmental forms, with no established relationship with pain, and electrophysiological studies are usually normal [18].

6. Differential Diagnosis

According to the Oxford Desk Reference, the only genetic diseases in which schwannomas occur are NF2-SWN, schwannomatosis and Carney complex; NF2-SWN is distinguished from NF1 by axillary freckling and Lisch nodules, common in NF1 and very rare in NF2-SWN, and deafness is common in NF2-SWN and rare in NF1 [1, 38]. The spinal neurofibromas of NF1 may be identical to the spinal schwannomas of NF2-SWN on imaging and at surgery, being distinct on histology, which may require review by a neuropathologist [1]. Only 1% to 2% of patients with NF2-SWN have six or more café-au-lait macules, and none of the other NF1 criteria occurs with increased frequency; conversely, some patients with schwannomatosis are wrongly diagnosed with NF1 when schwannomas or hybrid tumours are interpreted as neurofibromas [4]. NF1 is considered here only as a differential diagnosis.

The main diagnostic difficulty is the overlap between mosaic NF2-SWN, LZTR1-SWN and SMARCB1-SWN rather than distinction from NF1 [4]. In the Manchester region, NF2 mosaicism caused misdiagnosis of at least 9% of schwannomatosis cases, and LZTR1-SWN caused 1% to 2% of cases misdiagnosed as NF2-SWN because of unilateral vestibular schwannoma [9]. Multiple meningiomas usually occur in older people, so an isolated meningioma before 25 years should prompt genetic assessment; NF2-SWN is the most frequent predisposing condition with two or more meningiomas before 40 years, and commercial panels for multiple meningiomas exist [2, 22]. Somatic NF2, SMARCB1 or LZTR1 variants in isolated tumours (schwannomas, meningiomas, glioblastomas) do not confer hereditary predisposition [2, 3].

7. Treatments and Risk Management Strategies

7.1. General principles and organisation of care

There is no cure for NF2-SWN [2]. Assessment and treatment should take place at a centre experienced in the disease with a skull-base surgical team including a neurosurgeon, otorhinolaryngologist, audiologist and other specialists [1, 2]. In the UK register of 368 patients, the relative risk of death in patients treated at specialist centres was 0.34 (95% confidence interval [CI] 0.12–0.98) compared with those treated elsewhere [48]. In England, approximately 95% of patients are followed by centrally commissioned multidisciplinary teams at four National Health Service (NHS) centres [49]. The ERN GENTURIS patient journey emphasises that quality centres exist only in some countries (United Kingdom, France, Spain, Germany and Belgium), that quality of life and survival are lower outside specialist centres, that neurosurgery is the treatment of choice only when necessary (depending on location, volume and growth rate), that radiosurgery is contraindicated at young ages because of second-tumour risk, that bevacizumab is expensive and unavailable in many countries, and that “wait and scan” is the rule for multiple tumours throughout life [16]. Every new tumour requires weighing its own risk against the risk of intervention [16].

In France, the NF2 PNDS (French National Diagnosis and Care Protocol) defines as principles the preservation of function for as long as possible and the adaptation of each decision to the volumetric course and functional impact of each tumour [17]. The diagnosis of bilateral vestibular schwannomas should trigger a structured disclosure process with the patient and family, the physician, the geneticist and the psychologist; all certified teams have a multidisciplinary team meeting (RCP) and the most difficult cases can be presented at a national RCP, with the patient and family as participants in the decision [17]. The team includes oncology and radiotherapy, dermatology, clinical genetics and genetic counselling, obstetrics, paediatrics and paediatric neurology, neurology, neurosurgery, general practice, ophthalmology, otorhinolaryngology, speech and language therapy and phoniatrics (lip-reading, auditory rehabilitation, rehabilitation of facial palsy and dysphagia), radiology and physical and rehabilitation medicine with physiotherapy [17]. The decision weighs age and general condition, the extent and prognosis of the disease, tumour size, infiltration of the labyrinth and cochlear fossette that precludes hearing preservation, hearing in both ears and evoked potentials, the course of the other tumours and the sequelae of a previous contralateral resection [17]. The psychologist joins the team from disclosure, accompanies difficult decisions (genetic testing, pregnancy plans, complex therapeutic choices, end of life), adapts to communication with deaf patients, respects psychological defences in the event of refusal and can follow distant patients by teleconsultation; in familial forms, attention extends to the various affected members [17]. In non-NF2 schwannomatosis, the French PNDS accepts that follow-up need not take place in a reference centre, provided that the patient has imaging and genetic assessment, and centres the strategy on the type of pain: without pain, surveillance; with pain, search for a responsible schwannoma and first-line surgery if the radioclinical correlation is good; without a causal tumour or without sufficient correlation, medical treatment [18].

7.2. Vestibular schwannomas: surgery, radiosurgery and hearing rehabilitation

Untreated tumours may grow slowly and require no short-term intervention, and therapy remains predominantly surgical [2]. Small, entirely intracanalicular vestibular tumours can frequently be removed with preservation of hearing and facial function; larger tumours are probably best managed expectantly, with debulking or decompression only for brainstem compression, hearing deterioration and/or facial nerve dysfunction, and balancing early surgery preserving facial function against later surgery while the patient can still hear is difficult [2]. The Oxford Desk Reference states that early microsurgery of small tumours optimises hearing and facial preservation and that complete removal of a large vestibular schwannoma usually leaves the patient completely deaf on that side [1]. In Manchester, mean annual vestibular schwannoma surgery and radiotherapy rates fell from 4.66 and 1.65 per 100 patients in 1990–1999 to 2.11 and 1.01 in 2010–2020 as bevacizumab use increased (2.51 per 100 patients in 2010–2020); 153 of 353 patients (43.3%) underwent at least one neurosurgical or radiotherapy intervention within five years of diagnosis [50].

The French PNDS places surveillance by serial magnetic resonance imaging as a first-line option to be discussed, comparing tumour progression with hearing progression [17]. In sudden deafness, the patient should be seen promptly by an otorhinolaryngologist (otoscopy and confirmatory audiogram) and immediately start oral, intravenous or intratympanic corticosteroid therapy, which improves hearing in most cases (about 70%); the intravenous route is preferred if hearing does not improve quickly or if deafness is total in the only functioning ear [17]. The PNDS does not specify doses; in the analogous condition, idiopathic sudden sensorineural hearing loss in adults, the North American multicentre randomised non-inferiority trial (250 patients, 2004–2009) used oral prednisone 60 mg/day for 14 days with a 5-day taper, or four intratympanic injections of methylprednisolone 40 mg/mL over 14 days, with similar gains in mean pure-tone threshold at 2 months (30.7 dB and 28.7 dB) [51]. Transposing these regimens to sudden deafness in NF2-SWN is an extrapolation; the trial included only idiopathic unilateral deafness of at least 50 dB with less than 14 days' duration and does not establish paediatric or intravenous doses [51]. The North American guideline of the American Academy of Otolaryngology–Head and Neck Surgery Foundation (AAO-HNSF), updated in 2019 for adults (≥18 years) with sudden hearing loss, offers corticosteroids as an initial treatment option within the first 2 weeks and recommends offering intratympanic corticosteroids when recovery is incomplete 2 to 6 weeks after onset [52]. The systemic dose indicated is prednisone 1 mg/kg/day as a single dose, with a usual maximum of 60 mg/day (equivalents: methylprednisolone 48 mg/day or dexamethasone 10 mg/day), at full dose for 7–14 days and tapered over a similar period; prepackaged tapering dose packs of prednisone or methylprednisolone deliver an insufficient dose and should be avoided [52]. By the intratympanic route, the guideline indicates dexamethasone 24 mg/mL (compounded) or 10 mg/mL, or methylprednisolone 40 or 30 mg/mL, with 0.4–0.8 mL per injection and up to 4 injections over 2 weeks, keeping the head with the affected ear uppermost for 15–30 minutes [52]. Hyperbaric oxygen therapy combined with corticosteroids is an option within the first 2 weeks, or as salvage up to 1 month, and is not FDA-approved for this indication [52]. The guideline also recommends audiometry as early as possible (within 14 days of onset), investigation of retrocochlear pathology by magnetic resonance imaging or auditory brainstem evoked potentials, and audiometry at the end of treatment and up to 6 months afterwards; it advises against computed tomography and routine laboratory tests, as well as antivirals, thrombolytics, vasodilators and vasoactive agents [52]. Insulin-dependent or poorly controlled diabetes, labile hypertension, glaucoma, tuberculosis, peptic ulcer and previous psychiatric reactions to corticosteroids may preclude systemic corticosteroid therapy [52]. The oral regimen of the Rauch trial (60 mg/day for 14 days) coincides with the usual maximum in the guideline [51, 52]. The guideline cites a 2016 study in which adding intravenous methylprednisolone to oral corticosteroids did not improve hearing, whereas the PNDS prefers the intravenous route when hearing does not improve quickly or deafness is total in the only functioning ear [17, 52]. For NF2-SWN, the guideline includes NF2 among the causes of bilateral sudden deafness, states that 10–20% of patients with vestibular schwannoma report a sudden drop in hearing at some point and that recovery of hearing does not exclude a tumour, and notes that evoked potentials miss on average 20% (8–42%) of intracanalicular schwannomas, so that magnetic resonance imaging is the reference examination [52]. This guideline addresses idiopathic sudden deafness in adults; its application to NF2-SWN and to children is an extrapolation [52].

Microsurgery may be performed by the retrosigmoid, translabyrinthine, retrolabyrinthine or middle fossa (suprapetrous) approach, the latter used by some teams to resect intrameatal tumours with the aim of preserving hearing, particularly in children [17]. The PNDS distinguishes four aims: decompression of the internal auditory canal through the middle fossa, to maintain hearing; complete resection, citing the risk of recurrence of the residue; subtotal resection under monitoring, sparing the facial nerve; and partial decompressive resection, followed or not by planned radiosurgery of the residue [17]. In proactive strategies, with small tumours, decompression limited to the canal and cautious partial excision, useful hearing is preserved in up to 80% of operated tumours, but long-term control of the residual tumour remains unresolved [17]. In large tumours, postoperative deafness is almost constant and the facial risk is substantial; subtotal excision prioritising the facial nerve brings facial outcomes close to those of sporadic schwannoma of the same size: in the series of the French reference centre, resections were total in 32%, near-total in 25%, subtotal in 32% and partial in 11%, with good facial function (House-Brackmann grades I–II) in 81.5% at one year; risks of cerebrospinal fluid fistula, meningitis and cerebellar or long-tract deficits remain [17].

For radiosurgery, the PNDS refers to gamma knife as the technique with the greatest experience, and acknowledges that the risk of secondary malignant transformation is not negligible, although poorly documented, and that control is lower than in sporadic schwannoma [17]. In a 2018 Dutch series of 47 growing vestibular schwannomas, tumour control was 98%, 89%, 87% and 87% at 1, 3, 5 and 8 years, useful hearing was preserved in 95%, 82%, 59% and 33% at 1, 3, 5 and 7 years and facial palsy occurred in 2.5%; in the Marseille series (129 tumours in 103 patients), progression-free survival was 88% and 75% at 5 and 10 years and functional hearing on the treated side 47% and 34% [17]. These results do not override the British evidence of an excess of malignant neoplasms after irradiation of benign tumours, which leads other sources not to offer radiotherapy as first-line treatment [16, 53].

Stereotactic radiosurgery, particularly gamma knife, has been offered as an alternative in selected patients, but NF2-SWN results are inferior to those for sporadic vestibular schwannoma, with long-term tumour control of approximately 60% [2]. The Oxford Desk Reference reserved stereotactic radiotherapy for particularly aggressive tumours, poor surgical risk, refusal of surgery or advanced age [1]. In the UK national study of 1345 patients, 266 received radiation for benign tumours between 1969 and 2021 (median age at first treatment 32.9 years); 9 central nervous system malignancies or malignant progressions occurred in irradiated patients versus 4 in 1079 non-irradiated patients (p < 0.001), with lifetime and 20-year risk of approximately 6% in irradiated patients (4.9% after vestibular schwannoma radiotherapy) versus less than 1% in non-irradiated patients, and 30-year survival of 45.6% (95% CI 34.0–56.5) versus 66.4% (57.3–74.0) in matched controls [53]. The authors concluded that radiotherapy should not be offered as first-line treatment for benign tumours and that patients should be informed of a potential absolute excess risk of approximately 5% [53]. Transformation may occur in the treated tumour or as a new malignancy in the radiation field (e.g. glioblastoma), sometimes 15 years later; radiotherapy should be avoided in children [2]. The AACR update acknowledges that benefit may outweigh risk in rare individual cases, such as recurrent atypical meningiomas [15] The same update states that vestibular schwannoma surgery is rarely recommended in children with NF2-SWN because of the risk of additional cranial nerve injury [15].

Hearing preservation and rehabilitation are central. All patients and families should be referred to audiology; lip-reading can be taught, sign language is more easily acquired before hearing loss, hearing aids may help early, and rehabilitation with a cochlear or auditory brainstem implant should be discussed with those who have lost hearing; cochlear implantation may be useful in stable tumours with evidence of cochlear nerve function or after vascular injury to the cochlea without nerve damage [1, 2]. The French PNDS quantifies the results: with an intact cochlear nerve, cochlear implantation is effective even with the schwannoma in place, although below the usual candidates (mean sentence-in-quiet score of 69% with stable untreated tumours, 49% after radiotherapy and about 40% after surgery with nerve preservation), and may be combined with bevacizumab to control growth; the promontory test verifies nerve functionality [17]. Without a functional cochlear nerve, the auditory brainstem implant is the last resort: 13% of patients do not use it, the mean score with the implant alone is 12%, and it works mainly as an aid to lip-reading and to perception of environmental sounds; it is usually placed at excision of the second schwannoma, and a "dormant" implant at the first excision is not a routine indication [17]. Implants cause artefacts on follow-up magnetic resonance imaging and require repeated adjustments and motivated patients, and the use of bevacizumab has reduced recourse to the brainstem implant [17].

7.3. Other tumours and manifestations

Other intracranial, cranial nerve or spinal tumours grow very slowly, and surgery for a tumour with little impact may cause disability years before it would naturally occur; surgery for lesions other than vestibular schwannomas is usually performed only when symptomatic [1, 2]. Whether NF2-SWN ependymoma warrants aggressive treatment is unclear, and bevacizumab has shown clinical benefit in some patients [2]. Most cataracts require no removal, but childhood cataracts that may cause amblyopia may require removal and occlusion of the healthy eye [2]. Mononeuropathy and polyneuropathy with wrist or foot drop warrant rehabilitation medicine, physiotherapy and occupational therapy; cutaneous schwannomas are removed when causing deformity or pain, which may also aid diagnosis [2]. Cervical spine MRI should precede cranial surgery to avoid complications of manipulation under anaesthesia, and lumbosacral MRI should precede regional analgesia because spinal tumours may complicate epidural analgesia [2].

The French PNDS retains resection as the reference treatment for meningiomas that grow, increase oedema or cause symptoms, deficits or epileptic seizures; adjuvant radiotherapy is discussed in grade 2 meningiomas, although postoperative progression appears less unfavourable than in sporadic tumours of the same grade, and gamma knife radiosurgery is an option for small tumours and recurrences, with good local control but frequent distant recurrence, particularly in women, in young patients and after several treatments of other meningiomas [17]. Extramedullary spinal tumours are operated on in the presence of symptoms of spinal cord compression (fatigability on walking, sphincter, motor or sensory disturbances); symptomatic ependymomas are operated on in selected cases by experienced surgeons, preferably with intraoperative neurophysiological monitoring [17]. In peripheral schwannomas, isolated swelling or a moderate complaint warrant only clinical surveillance, and the risk of malignant transformation is so low that isolated growth does not warrant surgery; pain or significant sensory symptoms prompt surgery, usually as a day case and under regional anaesthesia, with a tumour-centred approach, intraoperative ultrasound for small or intramuscular tumours, nerve stimulation, systematic electromyographic monitoring in the brachial and lumbosacral plexuses and enucleation after opening the epineurium under optical magnification [17]. In sporadic peripheral schwannomas, pain and sensory symptoms improve in about 90%, with motor worsening in 11%, new hypoaesthesia in 11% and new pain in 2%; the two series published in NF2-SWN showed no increase in surgical risk [17].

7.4. Bevacizumab

Bevacizumab is a humanised anti-VEGF (vascular endothelial growth factor) monoclonal antibody. In 21 NF2-SWN vestibular schwannomas, VEGF was expressed in 100% of tumours and VEGFR-2 in 32% of tumour vessels, providing the rationale for the first compassionate treatment [54]. It has no approved indication for NF2-SWN or vestibular schwannoma in any label consulted (FDA, version of 6 January 2025; EMA, EPAR revision 68), so use is off-label, supported by cohort and phase 2 studies [6, 19, 20]. The 2025 epidemiological review states that no drug has been approved by regulatory agencies for NF2-SWN tumours and that those used are off-label, with endorsement by national health agencies, as in the English national programme [22, 49].

No label specifies a bevacizumab dose for NF2-SWN; comparison with the doses approved for other neoplasms puts the trial regimens in context. The European SmPC approves, in metastatic colorectal cancer, 5 mg/kg or 10 mg/kg every 2 weeks or 7.5 mg/kg or 15 mg/kg every 3 weeks, and the US label approves 5 mg/kg or 10 mg/kg every 2 weeks (colorectal cancer), 7.5 mg/kg every 3 weeks (second-line colorectal cancer) and 10 mg/kg every 2 weeks in adult recurrent glioblastoma, always by intravenous infusion, with no separate maximum dose defined in either label, 15 mg/kg every 3 weeks being the highest regimen among those listed [19, 20]. The regimens of 5 mg/kg every 2 weeks, 7.5 mg/kg every 3 weeks and 10 mg/kg every 2 weeks used in NF2-SWN thus fall within the label doses for other indications, whereas maintenance with 5 mg/kg every 3 weeks (Plotkin 2023) is lower than any label dose; in all cases this is use outside the label conditions, and neither label establishes safety and efficacy below 18 years [19, 20, 57]. The English programme criteria for initiating treatment were schwannoma growth of ≥4 mm in linear diameter or ≥60% in volume over the preceding mean 12-month period, or an imminent threat to neurological function, such as brainstem compression [49]. Age and previous growth rate predicted response [49]. GeneReviews summarises that approximately 60% to 70% of patients respond and treatment may continue for years, with concern about renal toxicity [2]. In a US cohort of 37 patients treated between 2014 and 2024, schwannomas in adolescents and young adults (12–24 years) progressed more frequently during treatment than in patients over 25 years (37.5% versus 11.9%; p = 0.026), with similar hearing outcomes [59]. In NF104, bevacizumab was more active in non-target intracranial non-vestibular schwannomas (response in 5/12, 42%) than in meningiomas (5/40, 13%), and 55% of meningiomas progressed within two years [60]. For spinal ependymoma, GeneReviews describes clinical benefit in some patients [2]. The French NF2 PNDS presents bevacizumab as an alternative to surgery and radiotherapy when these threaten hearing, with two main indications: patients with a single functioning ear, a growing vestibular schwannoma and a risk of deafness with surgery or radiotherapy; and patients with functional bilateral hearing, two growing schwannomas and a risk of unilateral or bilateral deafness with those modalities [17]. The PNDS regimen, with no distinction between adults and children, is 5 mg/kg intravenously every 2 weeks or 7.5 mg/kg every 3 weeks (maximum doses of the regimen), for an initial 6 months (it can be given through a peripheral vein); in stable or responding patients, it is continued while effective and tolerated and, with a persistent response at 12 months, the interval can be extended (for example, 5 mg/kg every 3 or every 4 weeks), to limit toxicity; it is stopped 4 weeks before planned surgery [17]. The PNDS estimates tumour reduction and hearing improvement in about one-third of patients, rare radiological improvement in children (although hearing may stabilise or improve), most frequent adverse effects of fatigue, hypertension and proteinuria, little efficacy in meningiomas and possible reduction of the cysts of cystic spinal ependymomas, delaying or avoiding surgery [17]. The PNDS induction regimens coincide with those of the English programme and with label doses for other neoplasms; extension to 5 mg/kg every 4 weeks is lower than the label doses and those tested in maintenance (5 mg/kg every 3 weeks), and the 4-week pause before surgery coincides with the 28-day minimum of the US label, but is shorter than the window of 6 weeks to 6 months proposed in the multicentre surgical series [19, 49, 57, 61]. The ERN GENTURIS non-NF2 schwannomatosis guideline suggests, with weak evidence, considering bevacizumab through multidisciplinary discussion for multiple rapidly growing tumours causing pain or neurological deficit, or inoperable tumours [13].

Toxicity is cumulative. In 33 patients receiving compassionate bevacizumab for a median of 34.1 months, 15/26 (58%) developed hypertension (median time 12.8 months) and 18/29 (62%) proteinuria (median times 23.7 months for 1+ and 31.9 months for 2+); eight required interruptions because of proteinuria, and cumulative dose was associated with mean blood pressure but not proteinuria [62]. The US label includes warnings about gastrointestinal perforations and fistulas, wound-healing complications (withhold for at least 28 days before elective surgery and do not administer for 28 days after major surgery), haemorrhage, arterial and venous thromboembolism, hypertension, posterior reversible encephalopathy syndrome, renal injury and proteinuria (withhold until proteinuria is below 2 g; discontinue for nephrotic syndrome), infusion reactions, embryofetal toxicity and ovarian failure; paediatric safety and efficacy are unestablished, non-mandibular osteonecrosis has been reported under 18 years, and the drug is not approved at this age [19]. The label requires blood pressure and urinary protein monitoring during treatment, and NF2-SWN series reinforce this surveillance during prolonged treatment [19, 62]. In two surgical series, previous bevacizumab was associated with greater haemorrhage during vestibular schwannoma excision (mean blood loss 421 mL versus 285 mL; total or near-total resection in 18% versus 52%), and haemorrhage was related to cumulative dose and interval since discontinuation; the authors of the multicentre series of 21 patients proposed a bevacizumab-free window of 6 weeks to 6 months before surgery, depending on the setting, longer than the labelled minimum of 28 days [19, 61, 63].

7.5. Brigatinib and other targeted drugs

Brigatinib is a multityrosine kinase inhibitor, including ALK, identified by the Synodos NF2 consortium in screens of approximately 2,000 compounds; NF2-deficient cells do not express ALK, and activity appears mediated by other receptors, such as FAK [64]. In substudy A of the INTUITT-NF2 platform basket trial (NCT04374305), 40 patients aged 12 years or older (median 26 years) with progressive tumours (10 vestibular schwannomas, 8 non-vestibular schwannomas, 20 meningiomas and 2 ependymomas) received oral brigatinib 180 mg/day; after a median follow-up of 10.4 months, radiological response was 10% (95% CI 3–24) in target tumours and 23% (95% CI 16–30) in all tumours, with greatest benefit in meningiomas and non-vestibular schwannomas, decreased annualised growth rates across tumour types, hearing improvement in 35% (95% CI 20–53) of eligible ears, exploratory reduction in self-reported pain and no treatment-related grade 4 or 5 adverse effects [21]. In the trial registry, the regimen is 90 mg/day for 7 days, increasing to 180 mg/day if tolerated, identical to the titration in the US label and the European SmPC for lung cancer (90 mg once daily for 7 days and then 180 mg once daily) [65, 66]; the brigatinib substudy is closed to recruitment, substudy B evaluates neratinib (200–240 mg/day) and substudy C the combination of retifanlimab (375 mg intravenously every 3 weeks) with bevacizumab (7.5 mg/kg every 3 weeks, intermittently); the trial continues recruiting, with primary completion expected in December 2029 [65, 67]. Investigators acknowledged the absence of a placebo arm, the small number of patients and the use of a quality-of-life questionnaire not specific to pain as limitations [64]. Brigatinib (Alunbrig®) is authorised only for ALK-positive non-small-cell lung cancer; the label includes interstitial lung disease or pneumonitis (particularly in the first week), hypertension, bradycardia, visual abnormalities, raised creatine kinase and pancreatic enzymes, hepatotoxicity and hyperglycaemia, without established paediatric efficacy and safety and with embryofetal toxicity in animals [65]. Its use in NF2-SWN is off-label.

The French PNDS considers everolimus an option of modest efficacy and good tolerability, which sometimes stabilises growing vestibular schwannomas when surgery, radiotherapy or bevacizumab are not required, or when these are no longer possible or effective; in the French phase 2 trial (NCT01490476), the registry provides, in patients aged 15 years or older, oral everolimus 10 mg/day (maximum protocol dose; 5 mg/day as the reduced dose) for 12 months [17, 67, 70]. For NF2-SWN meningiomas no longer amenable to surgery or radiotherapy, the PNDS proposes the combination of everolimus with octreotide, based on the phase 2 CEVOREM trial: 20 patients with recurrent meningiomas (2 grade 1, 10 grade 2 and 8 grade 3; 4 with a germline NF2 variant) received fixed doses of long-acting octreotide 30 mg on day 1 and everolimus 10 mg/day on days 1 to 28 of each cycle; progression-free survival at 6 months was 55% (95% CI 31.3–73.5), and the median tumour growth rate fell from 16.6% to 0.02% per quarter at 3 months, in a trial without a control arm [17, 73]. Both regimens come from trials and do not correspond to authorised indications for NF2-SWN, like the other drugs in this section.

A 2026 systematic review of 30 studies concluded that bevacizumab is the most studied and most active systemic therapy (hearing improvement up to 61%, hearing stability up to 84% and tumour response up to 60%), that mTOR inhibitor results were variable (one study stopped for futility), that five tyrosine kinase inhibitors showed hearing improvement of 17% to 60% and tumour responses of 10% to 23.5%, and that evidence is level 4, predominantly non-randomised [74].

Limits of trial interpretation. All studies are phase 2, with 9 to 61 patients, without a control arm (except the aspirin trial), with different response definitions (≥20% volumetric reduction according to REiNS, Response Evaluation in Neurofibromatosis and Schwannomatosis; word recognition above the critical interval) and populations differing in age and previous growth rates. Ranking drugs by indirect comparison of response percentages is not legitimate: brigatinib was evaluated mainly in meningiomas and non-vestibular schwannomas, whereas bevacizumab was studied in vestibular schwannomas with hearing loss. Decreased growth rate during treatment without placebo may reflect regression to the mean or natural growth variation. Median or subgroup hearing responses do not equal population benefit, and evidence in children is smaller and less favourable [56, 58, 59].

7.6. Regulatory status

No authorisation applications, CHMP opinions, PDUFA dates or regulatory designations specific to NF2-SWN or schwannomatosis were found in the sources consulted; nor were reimbursement or access data found for Portugal (INFARMED) or Brazil (ANVISA/CONITEC). Absence from the consulted sources does not prove non-existence.

7.7. SMARCB1- and LZTR1-related schwannomatosis

Treatment aims at quality of life, function and prevention of complications, ideally through a multidisciplinary team [3]. The ERN GENTURIS guideline recommends multidisciplinary pain management with a biopsychosocial approach; tricyclic antidepressants and gabapentinoids first-line, selective serotonin reuptake inhibitors or other antiepileptics (topiramate, carbamazepine, oxcarbazepine) second-line; avoiding chronic opioids because of poor neuropathic pain efficacy, tolerance, dependence and hyperalgesia; and, with weak evidence, TRPV1 antagonists (capsaicin and some cannabinoid receptor ligands) for intractable pain [13]. GeneReviews gives gabapentin, amitriptyline and venlafaxine as examples and states that no class appears superior [3]. None of these drugs has schwannomatosis as an indication; the reference doses are those in the labels for neuropathic pain, the analogous condition, and their application to schwannomatosis pain is an extrapolation. Gabapentin (Neurontin®) is indicated, in the Portuguese SmPC, for peripheral neuropathic pain in adults: starting at 300 mg on day 1, 300 mg twice on day 2 and 300 mg three times on day 3 (or 900 mg/day in three divided doses), with increases of 300 mg/day every 2–3 days up to a maximum of 3600 mg/day [76]; the US label restricts the indication to postherpetic neuralgia, with the same titration up to 1800 mg/day in three divided doses, with no additional benefit demonstrated above this dose [77]. Pregabalin (Lyrica®) is indicated in the European SmPC for peripheral and central neuropathic pain in adults, starting at 150 mg/day in two or three divided doses, increasing to 300 mg/day after 3–7 days and to a maximum of 600 mg/day after a further 7 days [78]; the US label approves, in adults, 300 mg/day in diabetic neuropathy (maximum 300 mg/day) and up to 600 mg/day in postherpetic neuralgia and spinal cord injury pain [79]; in both, the dose is adjusted to creatinine clearance and discontinuation is gradual over at least one week [78, 79]. Amitriptyline is indicated for neuropathic pain in adults in the harmonised Portuguese SmPC: 10–25 mg at night initially, increases of 10–25 mg every 3–7 days and a usual dose of 25–75 mg/day, with caution above 100 mg/day (above 75 mg/day in people over 65 years or with cardiovascular disease) and no use below 18 years, with no maximum dose defined in the source [80]; the US label is indicated only for depression, with 75–150 mg/day in outpatient adults (maximum of 150 mg/day in outpatients) [81]. Venlafaxine has no pain indication in any of the labels consulted; the reference dose by analogy is that for major depressive disorder in the US label of the extended-release formulation (Effexor XR®), in adults: 75 mg/day as a single dose, with a possible start at 37.5 mg/day for 4–7 days and increases of up to 75 mg/day at intervals of no less than 4 days up to a maximum of 225 mg/day, with reduction in renal and hepatic impairment [82]. Surgery is the most effective treatment: it should be offered early for painful schwannomas when possible without neurological deficit, by surgeons experienced in nerve sheath tumours, with intraoperative neurophysiological monitoring for critical nerves; repeat operations in the same area should be avoided because of diminishing benefit and potential worsening of the pain syndrome; spinal cord stimulation is an emerging option [13]. Intraspinal schwannomas over 5 mm warrant imaging and clinical surveillance and removal if symptomatic and growing; hearing preservation decreases in vestibular schwannomas over 1 cm and facial function in those over 2.5–3 cm; meningiomas are managed as sporadic tumours [3]. Radiotherapy probably increases malignant transformation risk and should be considered only for growing schwannomas untreatable by surgery or other therapies [13, 16]. In LZTR1-SWN, vestibular schwannoma management follows sporadic schwannoma recommendations, including surgery [15].

The French schwannomatosis PNDS applies to pain the 2020 French recommendations for neuropathic pain, based on a systematic review using GRADE methodology, and distinguishes focal pain from diffuse or central pain [18, 83]. In focal peripheral pain, first-line treatment is lidocaine plasters or transcutaneous electrical nerve stimulation (TENS), and second-line treatment high-concentration capsaicin patches or botulinum toxin A; in diffuse or central pain (for example, from a spinal or intracranial schwannoma), first-line treatment is tricyclic antidepressants (amitriptyline, clomipramine or imipramine), serotonin and noradrenaline reuptake inhibitors (duloxetine and, alternatively, venlafaxine, without French authorisation for analgesia) or gabapentin, and second-line treatment pregabalin, tramadol or combinations (for example, gabapentin with an antidepressant); repetitive transcranial magnetic stimulation, spinal cord stimulation and strong opioids are reserved for the last line, and cognitive behavioural therapies complement all steps [18, 83]. The reference doses of these classes, set for neuropathic pain in adults by the international NeuPSIG systematic review on which the French recommendations are based, are: gabapentin 1200–3600 mg/day in three divided doses; pregabalin 300–600 mg/day in two divided doses; duloxetine 60–120 mg/day as a single dose; extended-release venlafaxine 150–225 mg/day as a single dose; tricyclic antidepressants 25–150 mg/day in adults in one or two divided doses, not exceeding 75 mg/day in older people with tertiary amines (amitriptyline, imipramine, clomipramine) and with a risk of sudden cardiac death described above 100 mg/day; tramadol 200–400 mg/day in adults in two (extended-release) or three divided doses; lidocaine plasters, 1 to 3 over the painful area for up to 12 hours per day; 8% capsaicin patches, 1 to 4 for 30 to 60 minutes every 3 months; botulinum toxin A, 50–200 units subcutaneously in the painful area every 3 months; and strong opioids by individual titration [84]. For the drugs in common, these ranges coincide with the labels already mentioned (gabapentin up to 3600 mg/day in the Portuguese SmPC; pregabalin up to 600 mg/day in the European SmPC), and their application to schwannomatosis is an extrapolation [76, 78, 84]. Even with appropriate treatment, 27% of patients report having used at least six analgesics; benefits of cannabinoids have been described in a few patients and of bevacizumab in one isolated patient, with no data to justify their wider use [18].

According to the same PNDS, surgery for a peripheral schwannoma is indicated when pain is disabling and resistant to medical treatment, corresponds to a single radicular territory and the anatomoclinical correlation supports attribution to the tumour [18]. Pain relief (87%) and complete excision (89%) are slightly lower than in sporadic forms, and postoperative deficits are more frequent (motor in 8%, sensory in 14%), reaching 35% in multinodular forms, in excision of several tumours at the same sitting and in tumours arranged like a string of beads on the same nerve; radiosurgery of peripheral schwannomas has few published cases [18]. Radicular spinal schwannomas (foraminal, extraforaminal and intradural of the cauda equina) follow the same principles, whereas intracranial and cervical and dorsal intradural schwannomas may be operated on in the presence of significant and progressive compression of the brainstem or spinal cord, even without symptoms; intracranial schwannomas in schwannomatosis progress little, so surgery or radiosurgery is decided case by case, to avoid repeated procedures in patients already weakened by chronic pain and depression [18]. In neuropathic pain of the limbs or trunk resistant to first- and second-line treatment, spinal cord stimulation may reduce pain by more than 50%, is indicated before morphine-type drugs, requires a multidisciplinary decision and prior dorsolumbar magnetic resonance imaging to confirm canal patency at the site of the epidural electrode, and should use devices compatible with high-field magnetic resonance imaging, given the need for imaging surveillance [18]. If stimulation fails and there is no intracranial hypertension, intrathecal analgesia by implanted pump (morphine, ziconotide and local anaesthetics) may be used; in facial pain due to an intracranial tumour, spinal cord stimulation is not indicated, and stimulation of the contralateral precentral cortex may be proposed after a previous response to transcranial magnetic stimulation targeted at the same region and, in the most severe and resistant cases, deep brain stimulation [18]. Local psychological support (outpatient, in medical-psychological centres or in pain clinics) is recommended, given the recognised risk of depression [18].

Pain-targeted trials included tanezumab, an anti-nerve growth factor antibody, in a randomised placebo-controlled phase 2 trial (10 mg subcutaneously; NCT04163419), the first therapeutic schwannomatosis trial, whose registry lists 9 enrolled patients and unknown status since 2024, without results in the sources consulted; tanezumab is not authorised for this indication [6, 67, 85]. The phase 2 STARFISH screening trial (NCT05684692) evaluates siltuximab and erenumab against placebo for schwannomatosis pain in adults and is recruiting, with primary completion expected in November 2027 [67]. The French PNDS considers the place of systemic and gene therapies in non-NF2 schwannomatosis still to be evaluated [18].

7.8. Divergences between the French PNDS documents and other guidelines

7.9. Developments in the past 6 to 12 months

None of these developments constitutes an authorisation application, regulatory opinion or authorisation; they comprise observational studies, registry phase 2 results, protocols and trials not yet started.

8. Reproductive Options and Family Counselling

8.1. NF2-SWN: risks for parents, siblings and offspring

NF2-SWN is inherited in an autosomal dominant manner, and each child of a patient has up to a 50% probability of inheriting the variant: 50% for a germline variant and below 50% if the patient is mosaic [2]. Penetrance is age-dependent, almost complete by 60 years and close to 100% over an average lifespan [1, 2]. The clinical course has strong intrafamilial correlation, with considerable variability between families, although differences occur within families and even between identical twins [1, 2]. The classic assertion that approximately half of patients have an affected parent has been revised: in the 2023 English study of 1084 living patients, only 23% had inherited disease from an affected parent; excluding patients without a confirmed molecular diagnosis or bilateral vestibular schwannomas, de novo cases remained at 72%, and almost half were mosaic [8]. The 2010 estimate was 56% of cases due to a de novo variant [10].

If the proband is mosaic (variant absent in unaffected tissue or with allele fraction clearly below 50%), the parents are presumed not to carry the variant and siblings not to be at risk [2]. If the proband has a germline variant undetected in either parent's blood, sibling risk is slightly above population risk because of possible parental germline mosaicism; to date, all parents with more than one affected child had a blood-detectable variant and were clinically affected, and only one report of two affected children of apparently healthy parents exists, predating molecular testing [2]. The Oxford Desk Reference states that if parental skin, eyes and brain and spinal MRI are normal, recurrence risk is very low (below 1%) [1]. Exclusion of siblings as qualifying relatives in current criteria derives precisely from this low probability of recurrence through germline mosaicism compared with the probability of an incidental unilateral vestibular schwannoma in a sibling [2, 27]. Because onset age is usually consistent within families, offering surveillance to asymptomatic parents is generally unnecessary [2]; without a family history, the Oxford Desk Reference recommended evaluating parents under 60 years with skin examination, ophthalmological assessment and gadolinium-enhanced cranial and spinal MRI [1]. An apparently negative family history can be confirmed only if the proband is mosaic or neither parent carries the variant [2].

Risk to children of a mosaic patient depends on the blood allele fraction. Mosaic patients with bilateral vestibular schwannomas have below a 50% probability of an affected child; if the variant is detected in several tumours but not leucocytes, offspring risk is probably below 1% [2]. In the series of 1055 de novo cases, all parents of affected children with an NF2 point variant had a detectable mosaic variant, and offspring risks appear to correlate with blood allele fraction [32]. The 2023 English protocol provides estimated transmission risks for mosaic patients, derived from testing over 1000 individuals in the Manchester service and lower than previously used risks [14]. The ERN GENTURIS schwannomatosis guideline estimates transmission below 10% in isolated cases without a family history and with negative LZTR1 and SMARCB1 testing [13].

8.2. Predictive testing and family screening

Predictive testing requires prior identification of the familial variant; when the proband is mosaic, identification of the same variant in more than one tumour enables predictive testing of offspring [2]. Before identifiable variants became available, predictive testing could use genetic linkage in families with several affected individuals and a typical diagnosis [1]. Because surveillance begins in childhood and early detection changes management, predictive testing before 18 years is beneficial; many parents prefer to know their child's genetic status before screening begins, to avoid unnecessary investigations, and communication of results to parents and children should be planned [1, 2]. ClinGen classifies NF2-SWN actionability as strong in adults and children, reflecting the value of presymptomatic surveillance [6]. For at-risk offspring in whom testing is impossible or declined, the Oxford Desk Reference proposed ophthalmological assessment in early childhood, annual review of symptomatic lesions until early adolescence, vestibular schwannoma screening from early adolescence and complete cranial and spinal MRI at 15 and 30 years, with discharge from surveillance if everything is normal at 30 years and there are no other manifestations, except in late-onset families [1]. The 2023 English protocol proposes surveillance until disease risk falls below a pragmatic 1% threshold, which may be reached earlier using revised transmission risks [14]. Test consent should address socioeconomic consequences, the need for prolonged follow-up and test limitations within formal genetic counselling; DNA banking should be considered in probands without a molecular diagnosis [2]. The French NF2 PNDS (French National Diagnosis and Care Protocol) places testing for the familial variant in the children of an affected individual between 6 and 14 years, after discussion with the parents and when the child already tolerates MRI without sedation; a positive result leads to a complete assessment (brain and spinal MRI, audiometry and specialist consultation) and to follow-up adjusted to that assessment [17]. If the parent's variant is not known, it proposes brain MRI with thin slices through the internal auditory canals and, if normal, repetition every 5 years until 20 years, the age after which the appearance of vestibular schwannomas is considered exceptional [17]. At disclosure of the diagnosis, the patient should be informed of the genetic nature of the disease and the usefulness of the genetics clinic, lifelong clinical (neurological, otorhinolaryngological, ophthalmological) and radiological surveillance, the therapeutic options (surveillance, surgery, radiotherapy, bevacizumab), the fact that only tumours that warrant it are treated, the value of follow-up in a specialist centre, the auditory, neurological and visual risks, dangerous activities in the event of deafness or impaired balance, clinical trials and the patient association [17]. Genetic counselling, by a physician with specific training or a genetic counsellor, may be repeated as often as necessary, should precede prescription of the test (with information on the right not to undergo it), and the result is given in an individual consultation, with psychological support if desired [17].

8.3. SMARCB1- and LZTR1-related schwannomatosis

Inheritance is autosomal dominant with reduced penetrance. Fewer than 20% of patients have an affected parent; in simplex cases, de novo variants account for approximately 30% in LZTR1-SWN and 10% in SMARCB1-SWN; a carrier parent may be asymptomatic [3]. Penetrance is estimated at 40% to 50%, higher in SMARCB1 [3, 13]. A 2025 review, discussing SMARCB1 variant carriers surviving in rhabdoid tumour families, states that SMARCB1-SWN penetrance is high, “probably 100%”, citing the ERN GENTURIS guideline, which, however, gives a 40% to 50% estimate for schwannomatosis overall; this divergence shows that SMARCB1 penetrance is unestablished [12, 13]. Siblings of a proband with a carrier parent have a 50% risk of inheriting the variant but a lower risk of schwannomas and/or meningiomas due to reduced penetrance; one case of maternal germline SMARCB1 mosaicism has been reported [3]. Second-degree relatives have a 25% probability of carrying the variant; transmission risk may be considerably lower in mosaic individuals; cascade screening is generally not recommended for relatives of patients with confirmed 22q-SWN [16]. Counselling on 50% transmission risk in inherited variant carriers and reduced LZTR1 penetrance is a strong ERN GENTURIS guideline recommendation [13]. Predictive testing of minors may be beneficial because surveillance starts around 12 years [3]. Incidental LZTR1 carriers without a personal or family history should be informed about reduced penetrance and referred for counselling, without screening imaging [3, 13]. The French schwannomatosis PNDS establishes that presymptomatic testing is possible only with the variant identified in the index case, is discussed with the parents from adolescence and should not be offered before 12–14 years, because of incomplete penetrance and usual onset after puberty; a carrier child starts radiological surveillance from adolescence [18]. Asymptomatic relatives carrying a SMARCB1 or LZTR1 variant are offered brain MRI with study of the internal auditory canals (unilateral vestibular schwannoma possible with LZTR1, exceptional with SMARCB1, and meningiomas possible with SMARCB1), whole-spine MRI and, if possible, whole-body imaging to estimate tumour burden [18]. Pre-test counselling includes the obligation to inform the family if a genetic alteration is diagnosed [18]. As for LZTR1 variants found by genome-wide sequencing, the 2021 revision of the French bioethics law now allows communication of additional findings when they open up preventive or care measures; but, with a frequency of pathogenic variants of 1 in 281 people in gnomAD against an LZTR1-SWN prevalence of 1 in 527,000, the risk of schwannomatosis for a carrier without a personal or family history is below 1%, and the PNDS advises against radiological surveillance without schwannomas, pain or other suggestive symptoms in the person or the family [8, 18].

8.4. Prenatal diagnosis, preimplantation genetic testing and pregnancy

Once the familial variant is identified, prenatal diagnosis and preimplantation genetic testing (PGT-M) are possible; the optimal time for risk assessment and discussion is before pregnancy [2, 3]. The ERN GENTURIS guideline on reproductive counselling in cancer predisposition syndromes strongly recommends offering comprehensive, balanced, timely and non-directive reproductive counselling to all people with a predisposition syndrome and relevant relatives of reproductive age; encouraging couples considering prenatal diagnosis to reflect before conception on continuing or terminating a pregnancy; ensuring access to a multidisciplinary team (counsellor or geneticist, prenatal diagnosis specialist, fertility specialist and psychologist); and offering longitudinal counselling from diagnosis throughout reproductive life to men and women, including partners [86]. Options presented should include natural conception accepting the risk, prenatal diagnosis, PGT, gamete donation, adoption and postnatal predictive testing, with realistic information on success rates, the possibility of obtaining only affected embryos and each country's legal framework [86]. The specific schwannomatosis guideline recommends discussing transmission risks and prenatal and preimplantation diagnosis options at reproductive age or transition to adult care [13]. The ERN GENTURIS NF2-SWN journey includes availability of reliable prenatal and preimplantation diagnostic tests among its objectives [16]. In France, the NF2 PNDS allows prenatal and preimplantation diagnosis in familial and sporadic forms provided that the variant has been identified, non-invasive prenatal diagnosis by exclusion when the variant sought in the fetus is a point variant, pathogenic and of paternal origin, and prenatal diagnosis also in mosaic patients, whose transmission risk is much lower than 50% but not precisely defined [17]. Conversely, the non-NF2 schwannomatosis PNDS does not offer prenatal or preimplantation diagnosis for SMARCB1 and LZTR1, citing incomplete penetrance and variable expressivity with onset after puberty, in divergence from GeneReviews and from the European reproductive counselling guideline, which treat these options as available once the variant is known [3, 18, 86].

During pregnancy, there is no convincing evidence that schwannomas enlarge, but meningiomas may undergo hormonal effects, so intracranial hypertension risk should be assessed in women planning pregnancy [2]. Lumbosacral MRI before regional analgesia is particularly relevant during delivery, as illustrated by anaesthetic reports of parturients with NF2-SWN [2]. The French PNDS considers that pregnancy does not appear to alter the course of the disease or the tumours and that the mode of delivery is decided case by case with obstetrics: epidural analgesia and vaginal delivery with expulsive efforts, with possible cerebrospinal fluid leakage, are contraindicated in the presence of spinal tumours compressing the cord or posterior fossa tumours with a risk of herniation, which makes some deliveries high-risk [17]. Bevacizumab and brigatinib have embryofetal toxicity according to their respective labels, which recommend effective contraception and avoiding breastfeeding [19, 65].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence and incidence: what each number measures

The difference between birth incidence and diagnostic prevalence reflects diagnostic delay and reduced survival [6, 22]. NF2-SWN prevalence should not be transferred to other schwannomatoses, nor combined LZTR1/SMARCB1 frequency to either gene individually. The high de novo rate, lower reproductive fitness of heterozygotes (more marked in males) and mosaicism frequency make populations with substantially different rates unlikely, according to the 2025 review authors; no founder effects were described in the consulted sources [22]. The population frequency of loss-of-function LZTR1 variants (approximately 1 in 310 in gnomAD) contrasts with LZTR1-SWN rarity, reinforcing that variant frequency does not measure disease frequency [13, 36].

9.2. Regional data and special populations

The most robust epidemiological data come from the United Kingdom, whose nationally commissioned service has maintained a national database since 2009, and Finland [8, 22]. US data available in the consulted sources mainly concern clinical cohorts and multicentre trials (Massachusetts General Hospital, Johns Hopkins, NYU, Mayo Clinic and NFCTC consortium), rather than population prevalence; the 2025 review estimates approximately 100 new annual US cases based on 1992 incidence [22, 40, 60]. In France, the cohort from two national referral centres (551 patients) and the Paris molecular series document organisation into national centres [44, 47]. In Germany, severity classification was developed in 168 patients [41]. Japan conducted the anti-VEGFR vaccine trial, and China the icotinib trial and new trials registered in 2026 [67, 69, 71]. The 2021 and 2025 French PNDS documents do not present their own epidemiological data for France and use the British estimates for LZTR1-SWN and SMARCB1-SWN [17, 18]. No epidemiological data for Portugal or Brazil were found in the sources consulted; this absence does not prove non-existence of studies.

In special groups, children have a more severe phenotype, with high tumour burden and early ocular and cutaneous manifestations, and respond less to bevacizumab than adults [39, 56, 59]. Patients with onset after 30 years have fewer tumours and undergo less genetic testing [40]. Patients aged 70 years or older with bilateral vestibular schwannomas alone have a very low probability of a blood-detectable variant (approximately 6%), and many cases will represent coincident sporadic tumours [22, 27]. In non-NF2 schwannomatosis, chronic pain and anxiety or depression are the main causes of disability, and people with combined hearing and visual impairment in severe NF2-SWN have special communication needs [13, 22].

9.3. Referral networks and patient associations

The relevant European Reference Network is ERN GENTURIS (European Reference Network on Genetic Tumour Risk Syndromes), which published the schwannomatosis guideline (2022), reproductive counselling guideline (2026) and patient journeys for NF2-SWN (2023 text update) and non-NF2 schwannomatosis (2026), with a list of patient associations [13, 16, 86]. The list of ERN GENTURIS-authored clinical guidelines accessed on 7 October 2026 includes schwannomatosis but no specific formal NF2-SWN guideline; the European position on NF2-SWN is expressed in the patient journey and by network members co-authoring the 2022 consensus and English protocol [4, 14, 16]. Patient organisations and registries listed by GeneReviews and the Oxford Desk Reference include the Children's Tumor Foundation and NF Registry (United States), Neurofibromatosis Network, Acoustic Neuroma Association, Nerve Tumours UK, The Neuro Foundation (cited in the Oxford Desk Reference), International Schwannomatosis Database and, in the European guideline, Schwannoma Support UK, Italy's NF2 Project and NF Patients United [1–3, 13].

In France, the rare disease network is organised around a constitutive reference centre for NF2-SWN and the schwannomatoses in the neurosurgery department of the Pitié-Salpêtrière Hospital (Assistance Publique–Hôpitaux de Paris), with associated competence centres in Lille (Roger Salengro Hospital), Bordeaux (Pellegrin Hospital), Lyon (Pierre Wertheimer Neurological Hospital, Bron) and Marseille (Hôpital Nord) and, for non-NF2 schwannomatoses, also in Toulouse (CHU Purpan) [17, 18]. A national NF2-schwannomatosis RCP (multidisciplinary team meeting) meets quarterly for difficult cases [18]. The PNDS documents serve as a reference for the attending physician to establish, with the health insurance medical adviser, the care protocol for an application for exemption from co-payment (ALD "outside the list"), and they identify uses outside the marketing authorisation [17, 18]. The reference patient association is the Association Neurofibromatoses et Recklinghausen (ANR), which brings together people with NF1, NF2-SWN and non-NF2 schwannomatosis, is presented to all patients followed in the centres, publishes validated information letters, funds research, organises an annual colloquium and designates contact persons for schwannomatosis; there is also a French-speaking patient group on social media (NF2 Crew-Francophone) [17, 18].

10. Illustrative Clinical Case

In 2025, Li and Liu published in Frontiers in Medicine the case of a 10-year-old girl seen at Shenzhen Eye Hospital (China) in June 2023 for “blurred vision for more than 7 years” [89]. At 3 years, her parents noticed bilateral visual changes, and she was diagnosed with amblyopia and treated for 7 years without improvement before referral to ocular oncology [89]. Examination showed a soft skin-coloured scalp mass measuring 1.3 × 1.3 cm and three finger-sized soft abdominal masses, noted since birth and growing; visual acuity was 0.5 in the right eye and 0.3 in the left [89]. Slit-lamp examination showed posterior subcapsular opacity of the right lens; fundoscopy showed a membrane covering the right macular retina; optical coherence tomography showed a right epiretinal membrane and, on the left, ill-defined neuroepithelial thickening and a “flame-shaped” membrane projecting into the vitreous, consistent with retinal hamartoma; fluorescein angiography showed vascular tortuosity and hyperfluorescence (figure 1 of the original article) [89]. Ultrasound and MRI showed a left intraconal lesion beside the lateral rectus muscle (approximately 28 × 5.8 mm) extending into the optic canal, a left posterior cervical lesion, a sublingual lesion, lesions in both cerebellopontine angles without internal auditory canal enlargement, uptake along both trigeminal nerves and an ovoid lesion along the left oculomotor nerve's parasellar course, all interpreted as schwannomas (figure 2 of the original article) [89].

Exome sequencing of the proband, parents and sister identified an NF2 variant in the child described as c.626_635del, causing a frameshift, absent from relatives and not previously reported; biopsy of an abdominal mass in June 2023 confirmed the diagnosis (figure 3 of the original article) [89]. The article does not specify the reference transcript and describes the variant as heterozygous in the results and “homozygous” in a subsequent paragraph, so nomenclature cannot safely be normalised from the publication and the reference to homozygosity should be read as an inconsistency in the report. The authors reviewed 158 published NF2-SWN cases with ophthalmological manifestations, with visual impairment in 64%, strabismus in 38%, cataract in 25%, retinal abnormalities in 23% and retinal hamartoma in 16% [89].

Applying the 2022 criteria, bilateral cerebellopontine angle lesions interpreted as schwannomas point to bilateral vestibular schwannomas, although the report describes neither enlarged auditory canals nor audiological assessment; even without that criterion, the child had several minor criteria (multiple schwannomas, including cutaneous, juvenile subcapsular cataract, retinal hamartoma and epiretinal membrane before 40 years), and the blood frameshift variant would constitute a major criterion if classified as pathogenic, a classification the article does not provide according to ACMG/AMP [4, 89]. The case illustrates the paediatric warning signs listed by GeneReviews (cutaneous schwannomas present from birth, cataract, retinal hamartoma) and the risk of attributing visual loss to amblyopia without examining the lens and retina [2, 89]. It also illustrates that, because the variant is truncating and de novo, the child has a 50% risk of transmitting it to each child and siblings are not at increased risk if parents test negative, except for rare germline mosaicism [2]. The report describes neither treatment nor hearing and neurological follow-up; it should therefore not serve as a management model, which should follow the surveillance programme in section 11 [2, 15].

11. Prognosis and Long-Term Follow-Up

11.1. Mortality, survival and prognostic factors

NF2-SWN life expectancy is reduced. In the 1992 clinical series, mean age at death was 36.25 years in 40 cases, and all deaths except one resulted from disease complications [26]; GeneReviews gives a mean age at death of 36 years and actuarial survival after diagnosis increasing from 15 to over 30 years [2]. Before 1990, mean survival after diagnosis was only 15 years, and loss of lower cranial nerve function with aspiration pneumonia is a common terminal diagnosis [22]. In the UK register of 368 patients, relative mortality risk increased 1.13-fold for each year of decrease in diagnosis age (95% CI 1.08–1.18) and was 2.51 times higher with intracranial meningiomas (95% CI 1.38–4.57); risk was very low in missense variant carriers [48]. In the national analysis of 1192 patients (241 deaths over 10,995 patient-years), early diagnosis and intracranial meningiomas were associated with higher mortality and mosaicism with lower mortality; splicing and missense variants had lower mortality than truncating variants, splicing variants in exons 6–15 lower than those in exons 1–5 (OR 0.333; 95% CI 0.129–0.858), and mortality was lower in patients diagnosed in more recent decades [90]. In the Manchester series of 353 patients followed between 1990 and 2020 (60 deaths, 17.0%), asymptomatic patients at diagnosis had a longer interval to intervention and better survival; those presenting symptomatically before 16 years and after 40 had poorer survival than those aged 26–39, and constitutional de novo heterozygotes had poorer survival than mosaic or inherited cases [50]. Establishing four treatment centres in England from 2009 was associated with additional survival improvement, to which bevacizumab use may have contributed [22]. In non-NF2 schwannomatosis, life expectancy is not usually reduced, but quality of life is profoundly affected [9, 13].

11.2. NF2-SWN surveillance

Whether MRI before 10 years is beneficial or at what age it can safely be discontinued is unknown [2]. The French PNDS (French National Diagnosis and Care Protocol) bases annual radiological follow-up on the need to know the natural history of each tumour and to adapt treatment even without symptoms, given that the appearance of new tumours after 20–25 years is rare [17]. Differences between sources reflect era and population: the 2017 Oxford Desk Reference described a protocol for at-risk offspring without molecular testing, GeneReviews applies to affected individuals and at-risk people whose genetic status cannot be clarified, and the 2025 AACR update targets children and adds whole-body MRI and periodic spinal MRI [1, 2, 15]. Intracranial hypertension may have few symptoms, justifying systematic ophthalmological (papilloedema) and radiological surveillance [44]. In patients receiving bevacizumab, surveillance includes blood pressure and proteinuria, and response is assessed by volumetry and audiometry with word recognition [19, 49, 62]. For those undergoing planned surgery after bevacizumab, the interruption interval should be planned [61]. Assessment instruments used in studies include the disease-specific NFTI-QOL quality-of-life questionnaire and REiNS response criteria [49, 72], and three-grade severity classification is a proposed clinical communication tool [41].

11.3. SMARCB1- and LZTR1-related schwannomatosis surveillance

The European guideline, GeneReviews and the AACR agree on the 2–3-year interval and annual pain assessment but differ in starting age (10 to 14 years) and systematic whole-body MRI inclusion; cost and limited insurance coverage restrict its use [3, 13, 15]. The 2025 French PNDS acknowledges that imaging frequency is not a matter of consensus, accepts annual or biennial intervals depending on tumour burden and dispenses with systematic imaging when the brain and spinal study is normal; it considers whole-body imaging less sensitive than brain and whole-spine imaging, but justifiable in longitudinal follow-up because of the risk of malignant transformation [18]. Transition to adult care should be planned by a dedicated professional because many symptoms become more apparent after this stage [16].

12. Summary and Final Considerations

NF2-SWN is an autosomal dominant syndrome with almost complete penetrance, caused by merlin loss of function, with bilateral vestibular schwannomas, other schwannomas, meningiomas, ependymomas and juvenile ocular abnormalities, and early mortality that has improved with specialist centres [2, 22, 90]. The 2022 nomenclature grouped it with SMARCB1-, LZTR1- and 22q-related schwannomatoses, distinguished by absence of bilateral vestibular schwannomas, ependymomas and ocular abnormalities, dominant pain, reduced penetrance and the three-step, four-hit mechanism [3, 4]. Distinction is primarily molecular, and NF2 mosaicism, present in approximately half or more of de novo cases, requires testing blood and, whenever possible, at least two anatomically independent tumours [4, 32]. NF2 genotype–phenotype correlations (more severe truncating variants in exons 2–13; milder missense variants, large deletions, exons 14–15 and mosaicism) inform prognosis and counselling [2, 22].

Annual surveillance with brain MRI from approximately 10–12 years, audiometry and ophthalmological examination forms the basis of NF2-SWN care, whereas non-NF2 schwannomatosis requires imaging every 2–3 years and annual pain and psychological assessment [2, 13, 15]. Radiotherapy for benign tumours increases malignancy risk and should be avoided, particularly in children [2, 53]. No medicine is authorised for these disorders; off-label bevacizumab stabilises or reduces vestibular schwannomas and preserves hearing in some patients, with cumulative renal and hypertensive toxicity and less benefit in young people, and brigatinib showed responses mainly in meningiomas and non-vestibular schwannomas in an uncontrolled phase 2 trial [21, 49, 62]. Transmission risks range from 50% in germline carriers to below 1% in some mosaics, and incidental LZTR1 variants do not themselves justify diagnosis or surveillance [2, 13]. The 2021 and 2025 French PNDS documents (French National Diagnosis and Care Protocols) add a pathway for neurological and neurosensory emergencies, immediate corticosteroid therapy in sudden deafness, bevacizumab regimens with interval extension after one year of response and a line-based neuropathic pain algorithm, but diverge from the other sources on radiosurgery, on the intensity of surveillance of children and on the refusal of prenatal diagnosis in non-NF2 schwannomatosis [17, 18]. ERN GENTURIS, the national reference centres and patient associations form part of the care pathway, and current research focuses on platform trials, new MEK-pathway targets and immunotherapy, and ultrasensitive mosaicism detection methods [16, 46, 67].

13. Bibliographical References

1. Firth HV, Hurst JA. Neurofibromatosis type 2 (NF2). In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 596–598 (Box 4.10; Table 4.19; consultant: Evans DG).
2. Evans DG. NF2-Related Schwannomatosis. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle: University of Washington; 1993–2026. Published on 14 Oct 1998; comprehensive update on 20 Apr 2023. NBK1201. PMID:20301380.
3. Dhamija R, Plotkin S, Gomes A, Babovic-Vuksanovic D. LZTR1- and SMARCB1-Related Schwannomatosis. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle: University of Washington; 1993–2026. Published on 8 Mar 2018; comprehensive update on 27 Jul 2023; revision on 4 Dec 2025. NBK487394. PMID:29517885.
4. Plotkin SR, Messiaen L, Legius E, Pancza P, Avery RA, Blakeley JO, et al. Updated diagnostic criteria and nomenclature for neurofibromatosis type 2 and schwannomatosis: an international consensus recommendation. Genet Med. 2022;24(9):1967–77. doi:10.1016/j.gim.2022.05.007. PMID:35674741.
5. VariantValidator. MANE Select transcripts for NF2 (NM_000268.4; ENST00000338641.10), SMARCB1 (NM_003073.5) and LZTR1 (NM_006767.4) and validation of variants NM_003073.5:c.*82C>T, c.1121G>A, c.30del, c.34C>T and NM_006767.4:c.740G>A (GRCh38). VariantValidator. Accessed on 7 Oct 2026. Method reference: Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61–8. doi:10.1002/humu.23348. PMID:28967166.
6. GeneGuidance. Gene and condition curation: NF2, SMARCB1, LZTR1 [database]. Version dated 26 September 2026; accessed on 7 Oct 2026.
7. Orphanet. Full NF2-related schwannomatosis (ORPHA:637); Mosaic NF2-related schwannomatosis (ORPHA:634475); Full schwannomatosis (ORPHA:93921); Mosaic schwannomatosis (ORPHA:634492); Neurofibromatosis/schwannomatosis (ORPHA:634518) [Internet]. ORPHA:637 entry; data via API ORPHAcode dated 2 Jul 2026. Accessed on 7 Oct 2026.
8. Forde C, Smith MJ, Burghel GJ, Bowers N, Roberts N, Lavin T, et al. NF2-related schwannomatosis and other schwannomatosis: an updated genetic and epidemiological study. J Med Genet. 2024;61(9):856–60. doi:10.1136/jmg-2024-110065. PMID:38925914.
9. Evans DG, Bowers NL, Tobi S, Hartley C, Wallace AJ, King AT, et al. Schwannomatosis: a genetic and epidemiological study. J Neurol Neurosurg Psychiatry. 2018;89(11):1215–9. doi:10.1136/jnnp-2018-318538. PMID:29909380.
10. Evans DG, Howard E, Giblin C, Clancy T, Spencer H, Huson SM, et al. Birth incidence and prevalence of tumor-prone syndromes: estimates from a UK family genetic register service. Am J Med Genet A. 2010;152A(2):327–32. doi:10.1002/ajmg.a.33139. PMID:20082463.
11. Smith MJ, Higgs JE, Bowers NL, Halliday D, Paterson J, Gillespie J, et al. Cranial meningiomas in 411 neurofibromatosis type 2 (NF2) patients with proven gene mutations: clear positional effect of mutations, but absence of female severity effect on age at onset. J Med Genet. 2011;48(4):261–5. doi:10.1136/jmg.2010.085241. PMID:21278391.
12. Kehrer-Sawatzki H, Cooper DN. SMARCB1-related schwannomatosis and other SMARCB1-associated phenotypes: clinical spectrum and molecular pathogenesis. Fam Cancer. 2025;24(3):64. doi:10.1007/s10689-025-00486-4. PMID:40794298.
13. Evans DG, Mostaccioli S, Pang D, Fadzil O Connor M, Pittara M, Champollion N, et al. ERN GENTURIS clinical practice guidelines for the diagnosis, treatment, management and surveillance of people with schwannomatosis. Eur J Hum Genet. 2022;30(7):812–7. doi:10.1038/s41431-022-01086-x. PMID:35361920.
14. Halliday D, Emmanouil B, Evans DGR. Updated protocol for genetic testing, screening and clinical management of individuals at risk of NF2-related schwannomatosis. Clin Genet. 2023;103(5):540–52. doi:10.1111/cge.14310. PMID:36762955.
15. Perrino MR, Jongmans MCJ, Tomlinson GE, Greer MC, Scollon SR, Mitchell SG, et al. Update on cancer and central nervous system tumor surveillance in pediatric NF2-, SMARCB1-, and LZTR1-related schwannomatosis. Clin Cancer Res. 2025;31(8):1400–6. doi:10.1158/1078-0432.CCR-24-3278. PMID:39937237.
16. European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS). Patient journey NF2-related schwannomatosis (version 4.0, 2023 text update) and Patient journey non-NF2-related schwannomatosis (version 2.0, accepted on 3 Jun 2026) [Internet]. NF2-SWN journey; Non-NF2 SWN journey; ERN GENTURIS guidelines. Accessed on 7 Oct 2026.
17. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Neurofibromatose de type 2. Centre constitutif labellisé de neurofibromatose de type 2 (coord. Kalamarides M). Text of September 2021. Saint-Denis La Plaine: HAS; 2 Dec 2021. [https://www.has-sante.fr/jcms/p_3299797](https://www.has-sante.fr/jcms/p_3299797). Accessed on 2 Nov 2026.
18. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Schwannomatoses non NF2. Centre constitutif labellisé des schwannomatoses (coord. Peyre M). Text of December 2024. Saint-Denis La Plaine: HAS; 21 May 2025. [https://www.has-sante.fr/jcms/p_3598547](https://www.has-sante.fr/jcms/p_3598547). Accessed on 2 Nov 2026.
19. Genentech, Inc. Avastin (bevacizumab) injection, for intravenous use: US prescribing information. BLA 125085; version effective 6 Jan 2025; initial authorisation on 26 Feb 2004 (Drugs@FDA). DailyMed. Accessed on 7 Oct 2026.
20. European Medicines Agency. Avastin (bevacizumab): EPAR, EMEA/H/C/000582, revision 68; marketing authorisation issued on 12 Jan 2005; holder Roche Registration GmbH. EPAR. Accessed on 7 Oct 2026.
21. Plotkin SR, Yohay KH, Nghiemphu PL, Dinh CT, Babovic-Vuksanovic D, Merker VL, et al. Brigatinib in NF2-related schwannomatosis with progressive tumors. N Engl J Med. 2024;390(24):2284–94. doi:10.1056/NEJMoa2400985. PMID:38904277.
22. Evans DG, Blakeley JO, Plotkin SR. History and clinical epidemiology of NF2-related schwannomatosis. Fam Cancer. 2025;24(4):79. doi:10.1007/s10689-025-00504-5. PMID:41160189.
23. Rouleau GA, Wertelecki W, Haines JL, Hobbs WJ, Trofatter JA, Seizinger BR, et al. Genetic linkage of bilateral acoustic neurofibromatosis to a DNA marker on chromosome 22. Nature. 1987;329(6136):246–8. doi:10.1038/329246a0. PMID:2888021.
24. Rouleau GA, Merel P, Lutchman M, Sanson M, Zucman J, Marineau C, et al. Alteration in a new gene encoding a putative membrane-organizing protein causes neuro-fibromatosis type 2. Nature. 1993;363(6429):515–21. doi:10.1038/363515a0. PMID:8379998.
25. Trofatter JA, MacCollin MM, Rutter JL, Murrell JR, Duyao MP, Parry DM, et al. A novel moesin-, ezrin-, radixin-like gene is a candidate for the neurofibromatosis 2 tumor suppressor. Cell. 1993;72(5):791–800. doi:10.1016/0092-8674(93)90406-g. PMID:8453669. Errata: Cell. 1993;75(4):826.
26. Evans DG, Huson SM, Donnai D, Neary W, Blair V, Newton V, et al. A clinical study of type 2 neurofibromatosis. Q J Med. 1992;84(304):603–18. PMID:1484939.
27. Evans DG, King AT, Bowers NL, Tobi S, Wallace AJ, Perry M, et al. Identifying the deficiencies of current diagnostic criteria for neurofibromatosis 2 using databases of 2777 individuals with molecular testing. Genet Med. 2019;21(7):1525–33. doi:10.1038/s41436-018-0384-y. PMID:30523344.
28. Hulsebos TJ, Plomp AS, Wolterman RA, Robanus-Maandag EC, Baas F, Wesseling P. Germline mutation of INI1/SMARCB1 in familial schwannomatosis. Am J Hum Genet. 2007;80(4):805–10. doi:10.1086/513207. PMID:17357086.
29. Piotrowski A, Xie J, Liu YF, Poplawski AB, Gomes AR, Madanecki P, et al. Germline loss-of-function mutations in LZTR1 predispose to an inherited disorder of multiple schwannomas. Nat Genet. 2014;46(2):182–7. doi:10.1038/ng.2855. PMID:24362817.
30. ESMO Handbook — Cancer Genomics. Lugano: ESMO Press. Rare cancers: meningiomas. p. 100.
31. Kluwe L, Mautner V, Heinrich B, Dezube R, Jacoby LB, Friedrich RE, et al. Molecular study of frequency of mosaicism in neurofibromatosis 2 patients with bilateral vestibular schwannomas. J Med Genet. 2003;40(2):109–14. doi:10.1136/jmg.40.2.109. PMID:12566519.
32. Evans DG, Hartley CL, Smith PT, King AT, Bowers NL, Tobi S, et al. Incidence of mosaicism in 1055 de novo NF2 cases: much higher than previous estimates with high utility of next-generation sequencing. Genet Med. 2020;22(1):53–9. doi:10.1038/s41436-019-0598-7. PMID:31273341.
33. Baser ME, Kuramoto L, Joe H, Friedman JM, Wallace AJ, Gillespie JE, et al. Genotype-phenotype correlations for nervous system tumors in neurofibromatosis 2: a population-based study. Am J Hum Genet. 2004;75(2):231–9. doi:10.1086/422700. PMID:15190457.
34. Kehrer-Sawatzki H, Farschtschi S, Mautner VF, Cooper DN. The molecular pathogenesis of schwannomatosis, a paradigm for the co-involvement of multiple tumour suppressor genes in tumorigenesis. Hum Genet. 2017;136(2):129–48. doi:10.1007/s00439-016-1753-8. PMID:27921248.
35. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 149 (Noonan syndrome; LZTR1), 388 (distal 22q11.2 deletions including SMARCB1), 816 (Coffin–Siris syndrome).
36. Deng F, Evans DG, Smith MJ. Comparison of the frequency of loss-of-function LZTR1 variants between schwannomatosis patients and the general population. Hum Mutat. 2022;43(7):919–27. doi:10.1002/humu.24376. PMID:35391499.
37. Nogué C, Chong AS, Grau E, Han H, Dorca E, Roca C, et al. DGCR8 and the six hit, three-step model of schwannomatosis. Acta Neuropathol. 2022;143(1):115–7. doi:10.1007/s00401-021-02387-z. PMID:34821987.
38. Firth HV, Hurst JA. Ataxic adult; Cataract; Lumps and bumps; Patchy pigmented skin lesions (café-au-lait); Neurofibromatosis type 1 (NF1). In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 70, 88–89, 222–223, 278–279, 507.
39. Halliday D, Hanington L. The clinical spectrum of paediatric NF2-related schwannomatosis. Fam Cancer. 2025;24(4):89. doi:10.1007/s10689-025-00511-6. PMID:41286393.
40. Olson BG, Marinelli JP, Carlson ML, Neff BA, Driscoll CLW, Babovic-Vuksanovic D, et al. NF2-related schwannomatosis diagnosed before and after 30 years of age: differences in disease presentation and rates of positive genetic testing. Otol Neurotol. 2026;47(9):e1090–3. doi:10.1097/MAO.0000000000004975. PMID:42303233.
41. Lawson McLean AC, Löschner D, Farschtschi S, Dengler NF, Rosahl SK. Clinical severity grading of NF2-related schwannomatosis. Orphanet J Rare Dis. 2025;20(1):4. doi:10.1186/s13023-024-03512-3. PMID:39762911.
42. Evans DG, Ramsden RT, Gokhale C, Bowers N, Huson SM, Wallace A. Should NF2 mutation screening be undertaken in patients with an apparently isolated vestibular schwannoma? Clin Genet. 2007;71(4):354–8. doi:10.1111/j.1399-0004.2007.00778.x. PMID:17470137.
43. Pathmanaban ON, Sadler KV, Kamaly-Asl ID, King AT, Rutherford SA, Hammerbeck-Ward C, et al. Association of genetic predisposition with solitary schwannoma or meningioma in children and young adults. JAMA Neurol. 2017;74(9):1123–9. doi:10.1001/jamaneurol.2017.1406. PMID:28759666.
44. Keraudy A, Vincent C, Risoud M, Monfilliette A, Hanafi R, Peyre M, et al. Increased intracranial pressure in NF2-related schwannomatosis: an underestimated danger with serious consequences. J Neurooncol. 2026;178(1):15. doi:10.1007/s11060-026-05624-1. PMID:42168700.
45. Smith MJ, Bowers NL, Bulman M, Gokhale C, Wallace AJ, King AT, et al. Revisiting neurofibromatosis type 2 diagnostic criteria to exclude LZTR1-related schwannomatosis. Neurology. 2017;88(1):87–92. doi:10.1212/WNL.0000000000003418. PMID:27856782.
46. Horbacz M, Prokopiuk J, Castellanos E, Brems H, Blanco I, Legius E, et al. From clinical suspicion to molecular detection of low-level mosaicism in NF2-related schwannomatosis via ultra-sensitive duplex sequencing. Mol Med. 2026 [published online on 10 Jul 2026]. doi:10.1186/s10020-026-01561-x. PMID:42432501.
47. Peyre M, Barbance C, Tran S, Pacot L, Terris B, Kalamarides M, et al. High frequency of mosaic NF2-related schwannomatosis diagnosed after somatic analysis of multi-nodular schwannomas in schwannomatosis patients. Hum Genet. 2026;145(1):30. doi:10.1007/s00439-026-02825-6. PMID:41824074.
48. Baser ME, Friedman JM, Aeschliman D, Joe H, Wallace AJ, Ramsden RT, et al. Predictors of the risk of mortality in neurofibromatosis 2. Am J Hum Genet. 2002;71(4):715–23. doi:10.1086/342716. PMID:12235555.
49. Morris KA, Golding JF, Axon PR, Afridi S, Blesing C, Ferner RE, et al. Bevacizumab in neurofibromatosis type 2 (NF2) related vestibular schwannomas: a nationally coordinated approach to delivery and prospective evaluation. Neurooncol Pract. 2016;3(4):281–9. doi:10.1093/nop/npv065. PMID:29692918.
50. Forde C, King AT, Rutherford SA, Hammerbeck-Ward C, Lloyd SK, Freeman SR, et al. Disease course of neurofibromatosis type 2: a 30-year follow-up study of 353 patients seen at a single institution. Neuro Oncol. 2021;23(7):1113–24. doi:10.1093/neuonc/noaa284. PMID:33336705.
51. Rauch SD, Halpin CF, Antonelli PJ, Babu S, Carey JP, Gantz BJ, et al. Oral vs intratympanic corticosteroid therapy for idiopathic sudden sensorineural hearing loss: a randomized trial. JAMA. 2011;305(20):2071–9. doi:10.1001/jama.2011.679. PMID:21610239.
52. Chandrasekhar SS, Tsai Do BS, Schwartz SR, Bontempo LJ, Faucett EA, Finestone SA, et al. Clinical practice guideline: sudden hearing loss (update). Otolaryngol Head Neck Surg. 2019;161(1 Suppl):S1–S45. doi:10.1177/0194599819859885. PMID:31369359.
53. Evans DG, Halliday D, Obholzer R, Afridi S, Forde C, Rutherford SA, et al. Radiation treatment of benign tumors in NF2-related-schwannomatosis: a national study of 266 irradiated patients showing a significant increase in malignancy/malignant progression. Neurooncol Adv. 2023;5(1):vdad025. doi:10.1093/noajnl/vdad025. PMID:37051330.
54. Plotkin SR, Stemmer-Rachamimov AO, Barker FG 2nd, Halpin C, Padera TP, Tyrrell A, et al. Hearing improvement after bevacizumab in patients with neurofibromatosis type 2. N Engl J Med. 2009;361(4):358–67. doi:10.1056/NEJMoa0902579. PMID:19587327.
55. Blakeley JO, Ye X, Duda DG, Halpin CF, Bergner AL, Muzikansky A, et al. Efficacy and biomarker study of bevacizumab for hearing loss resulting from neurofibromatosis type 2-associated vestibular schwannomas. J Clin Oncol. 2016;34(14):1669–75. doi:10.1200/JCO.2015.64.3817. PMID:26976425. Errata: J Clin Oncol. 2026;44(8):723.
56. Plotkin SR, Duda DG, Muzikansky A, Allen J, Blakeley J, Rosser T, et al. Multicenter, prospective, phase II and biomarker study of high-dose bevacizumab as induction therapy in patients with neurofibromatosis type 2 and progressive vestibular schwannoma. J Clin Oncol. 2019;37(35):3446–54. doi:10.1200/JCO.19.01367. PMID:31626572.
57. Plotkin SR, Allen J, Dhall G, Campian JL, Clapp DW, Fisher MJ, et al. Multicenter, prospective, phase II study of maintenance bevacizumab for children and adults with NF2-related schwannomatosis and progressive vestibular schwannoma. Neuro Oncol. 2023;25(8):1498–506. doi:10.1093/neuonc/noad066. PMID:37010875.
58. Tops AL, Schopman JE, Koot RW, Gelderblom H, Putri NA, Rahmi LNA, et al. Efficacy and toxicity of bevacizumab in children with NF2-related schwannomatosis: a systematic review. Cancers (Basel). 2025;17(3):519. doi:10.3390/cancers17030519. PMID:39941885.
59. Hatley MG, Yohay KH, Roland JT Jr, Segal D. The association between age and outcomes of bevacizumab treatment in NF2-related schwannomatosis. Otol Neurotol. 2026;47(2):370–7. doi:10.1097/MAO.0000000000004759. PMID:41250253.
60. Nakhate V, Ly I, Muzikansky A, Rapalino O, Allen JC, Blakeley JO, et al. Effect of bevacizumab on non-target intracranial meningiomas and non-vestibular schwannomas in NF2-related schwannomatosis: NF104. J Neurooncol. 2025;173(3):751–7. doi:10.1007/s11060-025-05020-1. PMID:40434540.
61. Hudelist B, King AT, Marinelli JP, Roland JT Jr, Pathmanaban O, Raza-Knight S, et al. Salvage microsurgery after failed bevacizumab treatment for NF2-related schwannomatosis vestibular schwannoma: a multicentric retrospective study. Neurosurgery. 2026;98(5):1085–94. doi:10.1227/neu.0000000000003728. PMID:40956079.
62. Slusarz KM, Merker VL, Muzikansky A, Francis SA, Plotkin SR. Long-term toxicity of bevacizumab therapy in neurofibromatosis 2 patients. Cancer Chemother Pharmacol. 2014;73(6):1197–204. doi:10.1007/s00280-014-2456-2. PMID:24710627.
63. Wood CG, Kozanno L, Stemmer-Rachamimov A, Lee DJ, Barker FG 2nd, Carter BS, et al. Increased hemorrhage during excision of bevacizumab-treated NF2-related vestibular schwannomas. Laryngoscope. 2025;135(11):4355–63. doi:10.1002/lary.32342. PMID:40511882.
64. Plotkin SR, Blakeley JO. Expanding therapeutic options for people with NF2-related schwannomatosis: encouraging results with brigatinib. Neuro Oncol. 2024;26(11):1949–50. doi:10.1093/neuonc/noae137. PMID:39389170.
65. Takeda Pharmaceuticals America, Inc. Alunbrig (brigatinib) tablets: US prescribing information. NDA 208772; version effective 12 Jan 2026; initial authorisation on 28 Apr 2017 (Drugs@FDA). DailyMed. Accessed on 7 Oct 2026.
66. European Medicines Agency. Alunbrig (brigatinib): EPAR – product information (summary of product characteristics), section 4.2; marketing authorisation on 22 Nov 2018. PDF. Accessed on 26 Oct 2026.
67. ClinicalTrials.gov. Registry records accessed through API v2 on 7 Oct 2026: NCT04374305 (INTUITT-NF2, recruiting, updated on 2 Sep 2026), NCT04283669 (crizotinib, completed, results published in the registry), NCT03095248 (selumetinib, terminated), NCT05130866 (REC-2282, terminated), NCT03079999 (aspirin, active), NCT04163419 (tanezumab, status unknown), NCT05684692 (STARFISH, recruiting), NCT05521048 (intralesional doxycycline in cutaneous schwannomas, active), NCT07707947, NCT07708285, NCT07713745, NCT07131722 and NCT06834438 (not yet recruiting), NCT01767792 (high-dose bevacizumab, completed).
68. Garcia MR, Hagiwara M, Yaffe A, Mitchell C, Akshintala S, Nicolaides T, et al. Phase II study of axitinib in patients with NF2-related schwannomatosis and progressive vestibular schwannomas. Neurooncol Adv. 2025;7(1):vdaf083. doi:10.1093/noajnl/vdaf083. PMID:40575410.
69. Zhao F, Li SW, Zhang S, Li P, Zhao C, Zhao XB, et al. Phase II trial of icotinib in adult patients with neurofibromatosis type 2 and progressive vestibular schwannoma. J Neurosurg. 2023;138(6):1680–7. doi:10.3171/2022.9.JNS22699. PMID:36272122.
70. Goutagny S, Raymond E, Esposito-Farese M, Trunet S, Mawrin C, Bernardeschi D, et al. Phase II study of mTORC1 inhibition by everolimus in neurofibromatosis type 2 patients with growing vestibular schwannomas. J Neurooncol. 2015;122(2):313–20. doi:10.1007/s11060-014-1710-0. PMID:25567352.
71. Tamura R, Yamanobe Y, Fujioka M, Morimoto Y, Fukumura M, Nakaya M, et al. Phase I/II study of a vascular endothelial growth factor receptor vaccine in patients with NF2-related schwannomatosis. J Clin Oncol. 2024;42(21):2578–87. doi:10.1200/JCO.23.02376. PMID:38776485.
72. Jabban S, Frome S, Anstett K, Hatley M, Yohay K, Roland JT Jr, et al. Response of vestibular schwannomas in NF2-related schwannomatosis (NF2-SWN) to losartan: a case series. Otol Neurotol. 2026 [published online on 2 Oct 2026]. doi:10.1097/MAO.0000000000005087. PMID:42826387.
73. Graillon T, Sanson M, Campello C, Idbaih A, Peyre M, Peyrière H, et al. Everolimus and octreotide for patients with recurrent meningioma: results from the phase II CEVOREM trial. Clin Cancer Res. 2020;26(3):552–7. doi:10.1158/1078-0432.CCR-19-2109. PMID:31969329.
74. Zalaquett NG, Khandalavala KR, Marinelli JP, Lohse CM, Chelf CJ, Carlson ML. Medical therapy for NF2-related schwannomatosis: a systematic review of current evidence. Otol Neurotol. 2026;47(8):e1012–20. doi:10.1097/MAO.0000000000004968. PMID:42276040.
75. European Medicines Agency. Alunbrig (brigatinib): EPAR, EMEA/H/C/004248, revision 10; CHMP opinion on 19 Sep 2018; authorisation issued on 22 Nov 2018; holder Takeda Pharma A/S. EPAR. Accessed on 7 Oct 2026.
76. INFARMED, I.P. Neurontin 300 mg capsules (gabapentin), Viatris Healthcare: summary of product characteristics (Infomed, med_id 6057); first authorisation on 27 Sep 1996. Infomed. Accessed on 26 Oct 2026.
77. Viatris Specialty LLC. Neurontin (gabapentin) capsules, tablets and oral solution: US prescribing information. NDA 020235/020882/021129; version effective 12 May 2026. openFDA. Accessed on 26 Oct 2026.
78. European Medicines Agency. Lyrica (pregabalin): EPAR – product information (summary of product characteristics), section 4.2; marketing authorisation on 6 Jul 2004. PDF. Accessed on 26 Oct 2026.
79. Viatris Specialty LLC. Lyrica (pregabalin) capsules and oral solution: US prescribing information. NDA 021446/022488; version effective 15 Apr 2025. openFDA. Accessed on 26 Oct 2026.
80. INFARMED, I.P. ADT 25 mg film-coated tablets (amitriptyline), Generis Farmacêutica: summary of product characteristics (Infomed, med_id 137). Infomed. Accessed on 26 Oct 2026.
81. Preferred Pharmaceuticals Inc. Amitriptyline hydrochloride tablets: US prescribing information. ANDA 217411; version effective 23 Jan 2026. openFDA. Accessed on 26 Oct 2026.
82. Viatris Specialty LLC. Effexor XR (venlafaxine extended-release) capsules: US prescribing information. NDA 020699; version effective 15 Aug 2023. openFDA. Accessed on 26 Oct 2026.
83. Moisset X, Bouhassira D, Avez Couturier J, Alchaar H, Conradi S, Delmotte MH, et al. Pharmacological and non-pharmacological treatments for neuropathic pain: systematic review and French recommendations. Rev Neurol (Paris). 2020;176(5):325–52. doi:10.1016/j.neurol.2020.01.361. PMID:32276788.
84. Finnerup NB, Attal N, Haroutounian S, McNicol E, Baron R, Dworkin RH, et al. Pharmacotherapy for neuropathic pain in adults: a systematic review and meta-analysis. Lancet Neurol. 2015;14(2):162–73. doi:10.1016/S1474-4422(14)70251-0. PMID:25575710. PMCID:PMC4493167.
85. Da JLW, Merker VL, Jordan JT, Ly KI, Muzikansky A, Parsons M, et al. Design of a randomized, placebo-controlled, phase 2 study evaluating the safety and efficacy of tanezumab for treatment of schwannomatosis-related pain. Contemp Clin Trials. 2022;121:106900. doi:10.1016/j.cct.2022.106900. PMID:36038003.
86. Farschtschi SC, Kumps C, Milagre TH, Makrythanasis P, Van Tongerloo A, Denayer E, et al. ERN GENTURIS guideline on counselling on reproductive options for individuals with a cancer predisposition syndrome (including genturis). Eur J Hum Genet. 2026;34(3):307–13. doi:10.1038/s41431-025-02007-4. PMID:41530368.
87. Douwes JPJ, Tops AL, Dibbets-Schneider P, Verbist BM, Fuentealba Bassaletti CA, Jansen JC, et al. [89Zr]bevacizumab PET/CT imaging of vestibular schwannomas for the prediction of bevacizumab treatment effect in patients with symptomatic NF2-related schwannomatosis: a study protocol for a phase II single centre, prospective, feasibility trial. BMJ Open. 2026;16(4):e114812. doi:10.1136/bmjopen-2025-114812. PMID:41951260.
88. Campbell S, Beaney K, Hosking J, Sorrell L, Allgar V, Aspinall PJ, et al. Repurposing anti-retroviral drugs to treat NF2-related tumours: a protocol for a phase 0 trial (RETREAT). BMJ Open. 2026;16(6):e109719. doi:10.1136/bmjopen-2025-109719. PMID:42303397.
89. Li Q, Liu G. Neurofibromatosis type 2 misdiagnosed as amblyopia — a case report and literature review. Front Med (Lausanne). 2025;12:1556494. doi:10.3389/fmed.2025.1556494. PMID:40852360.
90. Hexter A, Jones A, Joe H, Heap L, Smith MJ, Wallace AJ, et al. Clinical and molecular predictors of mortality in neurofibromatosis 2: a UK national analysis of 1192 patients. J Med Genet. 2015;52(10):699–705. doi:10.1136/jmedgenet-2015-103290. PMID:26275417.