During the last few years, we witnessed important new findings and developments that have substantial impact on NF1 and related disorder diagnostics. These include the identification of LZTR1 pathogenic variants as cause of café au lait spots and, hence, a relevant differential diagnosis in suspected NF1/Legius syndrome patients; the development of new sequencing techniques, including long-read sequencing, that allow the identification of previously undetected NF1 pathogenic variants; the elucidation of novel patho-mechanisms underlying variants in regulatory regions; and the first internationally developed criteria for the classification of NF1 and SPRED1 variants. This presentation will provide an overview of these new developments that will be addressed also in other presentations of the session. Strategies and challenges to incorporate these new developments into a clinical diagnostics pipeline will be discussed.

Mosaicism represents one of the major challenges in the molecular diagnosis of schwannomatosis, particularly when pathogenic variants are present at very low variant allele frequencies. This talk addresses the biological and technical factors that complicate the detection of low-level mosaicism, as well as the implications for patients with suspected disease. By reviewing recent advances in highly sensitive sequencing approaches for NF2 and related peripheral nerve sheath tumor genes, this presentation, together with the other talks in this session, will demonstrate how increased analytical sensitivity can uncover pathogenic mosaic variants that remain undetected by conventional genetic testing.

Background:
Accurate interpretation of genetic variants remains a significant challenge in clinical genomics, particularly for inherited tumor predisposition syndromes. Although the 2015 ACMG/AMP guidelines established a standardized framework for variant classification, inter-laboratory discordance persists, highlighting the need for disease-specific refinements and expert curation.
Methods:
The ClinGen Neurofibromatosis and Schwannomatosis (NF-SWN) Variant Curation Expert Panel (VCEP) was established to improve the consistency and accuracy of variant interpretation through disease-specific specifications. The panel brings together an international group of clinicians, researchers, genetic counselors, and diagnostic laboratories to share evidence, resolve classification discrepancies, submit expert-reviewed classifications to ClinVar, and develop gene-specific variant curation specifications available for public use globally. Current efforts include rule specification development for NF1, SPRED1, SMARCB1, NF2, and LZTR1.
Results:
This presentation reviews the rationale, structure, and progress of the NF-SWN VCEP and highlights major updates to proposed NF1 and SPRED1 classification criteria. Key changes include implementation of a quantitative points-based framework, expanded use of RNA evidence, refined computational thresholds, phenotype-specific definitions for case-level evidence, and updated guidance for segregation and de novo data. Through SPRED1 variant case studies, this presentation illustrates how structured evaluation of clinical, segregation, population, and computational evidence can lead to more consistent pathogenic and benign classifications.
Conclusion:
The presentation also emphasizes the critical role of clinicians in variant interpretation through detailed phenotyping, family history documentation, communication of additional testing results, and collaboration with diagnostic laboratories. These contributions are essential for reducing variants of uncertain significance and improving the clinical utility of genetic testing in neurofibromatosis and schwannomatosis care.

Bayes' theorem allows to calculate the posterior probability given the prior probability, and an event that modifies this prior probability. We can apply Bayes' theorem to calculate the posterior probability that a child with more than 5 café-au-lait macules and a pathogenic variant in one of the following genes (NF1, SPRD1 or LZTR1) has neurofibromatosis type 1, Legius syndrome or LZTR1-associated café-au-lait macules. To calculate the posterior probability, we need the sensitivity and specificity of the molecular tests. The ACMG classification guidelines for genetic variant interpretation also use a Bayesian system to score genetic variants as benign, likely benign, unknow significance, likely pathogenic and pathogenic. A calculation to asses the strength of evidence (odds for pathogenicity) of a de novo missense variant in the NF1 gene will be presented and discussed.

Background: Café-au-lait macules (CALMS) are one of the distinctive and most precocious signs associated with neurofibromatosis type 1 (NF1) and Legius syndrome. Their observation in newborns or young children often lead to molecular explorations in order to get an early confirmation of the diagnosis. In 2024, two publications suggested the association of CALMS with heterozygous loss-of-function variants in LZTR1, a gene previously associated with schwannomatosis and Noonan syndrome. Here we explore this correlation in the French cohort.
Methods: The study included patients addressed to our laboratory for a suspicion of NF1 and diagnosed between October 2021 and May 2026 (approximately 2800 patients). All patients had undergone a next-generation sequencing including the NF1 (neurofibromatosis type 1), SPRED1 (Legius), NF2, SMARCB1 and LZTR1 genes (schwannomatoses).
Results: We identified 32 patients with a heterozygous loss-of-function variant in LZTR1, in the absence of such variants in NF1 and SPRED1. These variants were identified in about 2.2% of patients with isolated pigmentary lesions (CALMS +/- frecklings), compared to 0.48% in an internal control cohort. The variant was inherited in 94% cases (16/17), half of the time from an asymptomatic parent. No history of schwannomatosis nor Noonan syndrome was reported in patients or relatives.
Conclusion: Our study corroborates previous ones and show a clear over-representation of LZTR1 heterozygous pathogenic variants in individuals with CALMS. The apparent incomplete penetrance of both CALMS and schwannomatosis associated with these variants complicates genetic counselling and will require more comprehensive follow-up of affected patients.

Background: Although multi-step gDNA- and mRNA-based NF1 mutation analysis reaches detection rates of up to 94% in familial cases, genetically unresolved cases remain, likely due to structural rearrangements, regulatory, and low-level mosaic variants. PacBio high-fidelity (HiFi) long-read sequencing, enabling simultaneous detection of single-nucleotide variants and structural variants in coding, non-coding and regulatory regions promise to close this diagnostic gap.
Material and Methods: To leverage the HiFi -sequencing advantages in a cost-effective manner, we designed a custom-panel long-read sequencing approach (cp-lrseq, TWIST) targeting 44 cancer genes and comprehensively covering the low copy repeat-flanked NF1 locus. We evaluate this approach with challenging benchmarking samples, a cohort of 231 genetically unresolved suspected NF1 patients, NF1 copy number variants of unknown significance (VUS) and selected routine diagnostic cases.
Results: Benchmarking successfully identified a deep intronic variant, resolved the nature and breakpoints at nucleotide resolution of an atypical, type 1 and a type 2 NF1-microdeletion, and due to its high coverage (average 600x) detected two low-level mosaic variants.
In the so far 82 unsolved cases, cp-lrSeq uncovered two translocations and a likely pathogenic promoter variant. We identified a de-novo SVA element insertion and resolved two previously uncharacterized NF1 duplications, enabling their reclassification as pathogenic and likely benign, respectively.
Conclusion: Our cost-effective cp-lrseq approach improved the diagnostic yield by identifying structural and regulatory variants previously missed by comprehensive testing in 3.6% of the patients and enabling the reclassification of VUS. Capturing also low-level mosaic variants, this approach may serve as a single assay in routine diagnostics.

Background
Detection and interpretation of NF1 and NF2 variants is challenging in mosaic cases where low-frequency pathogenic variants may escape detection or are dismissed as sequencing artefacts. We evaluated a high-resolution targeted next-generation sequencing (tNGS) approach combined with nanopore sequencing and adaptive sampling-enrichment to validate suspected low-level mosaic variants.

Methods
Suspected mosaic variants, detected with high-coverage NGS (>1000 reads), were assessed in a validation strategy, including negative controls, positive controls with variant detection in additional tissues and suspected mosaics. Long amplicons were generated to enhance target specificity and to facilitate discrimination from circulating tumor DNA fragments. Adaptive sampling-based nanopore sequencing resulted in ultra-deep coverage of target regions through real-time read selection. Sequence data were visualized and inspected in the Integrative Genomics Viewer.

Results
Fourteen variants were evaluated, ten NF1 variants (eight suspected mosaics and two positive controls) and four NF2 variants (three suspected mosaics and one positive control). Among the NF1 cases, two suspected mosaic variants were likely to be artefacts, four classified as likely heterozygous variants with variant allele frequencies (VAFs) between 20–41%, and three as mosaic variants with VAFs of 1%, 1%, and 9%. For NF2, two variants were identified as mosaic (2% and 10% VAF, including one positive control) and one as likely heterozygous (25% VAF). VAFs were concordant between tNGS and adaptive sampling-based nanopore sequencing.

Conclusions
This high-resolution targeted sequencing workflow reliably distinguishes true low-level mosaic variants from artefacts and improves the detection of clinically relevant NF1 and NF2 variants that may be missed by a conventional diagnostic approach.

Background
Each year, approximately 100 new adult patients are clinically diagnosed with (mosaic) Neurofibromatosis type 1 (NF1) or Schwannomatosis (SWN) at our NF1 Expertise Center. A subset presents with benign peripheral nerve sheath tumors (bPNSTs) in the absence of other typical NF1 or SWN features, such as café-au-lait macules, cutaneous neurofibromas, skinfold freckling, and Lisch nodules, and without germline or mosaic variants in the typical NF1- or SWN-associated genes. A recent study reported somatic pathogenic variants in ERBB2 in bPNSTs. In the current study, we analyzed whether ERBB2 mosaicism could explain the phenotype in the group of patients with bPNSTs who lack other features of NF1 or SWN.

Methods
We performed molecular analyses of bPNSTs from 11 patients clinically suspected of having mosaic NF1 or SWN, who tested negative for germline or mosaic variants in NF1, NF2, SMARCB1, and LZTR1. In total, 19 tumors were analyzed by targeted next generation sequencing using a gene panel that included ERBB2.

Results
Pathogenic variants NM_004448.4(ERBB2):c.2305G>T, p.D769Y or NM_004448.4(ERBB2):c.2329G>T, p.V777L were identified in 15 lesions from 7 patients. In 4 of these patients, multiple tumors were analyzed, with each patient’s lesions harboring the same pathogenic ERBB2 variant.

Conclusion
ERBB2 mosaicism may explain the clinical symptoms observed in a subset of patients with multiple bPNSTs who lack other typical features of NF1 or SWN. These findings may have potential therapeutic implications, as a recently reported patient demonstrated clinical improvement to ERBB2-targeted therapy.