Precision medicine in Neurofibromatosis type 1: deciphering the functional effects of variants of uncertain significance in the NF1 gene in molecularly unsolved cases to improve differential diagnosis

Neurofibromatosis type 1 is a genetic tumour-predisposing disease with marked genotypic and phenotypic variability. Clinical overlap with other Rasopathies and variants of uncertain significance (VUS) make differential diagnosis challenging, especially in incomplete or atypical phenotypes. Here, we address NF1 genotype-phenotype complexity, to refine its differential diagnosis among RASopathies and improve diagnostic yield through functional studies. We reviewed data from 280 probands referred for NF1 differential diagnosis by RASopathies NGS panel and NF1 MLPA analyses. In unsolved cases functional analyses employed CRISPR/Cas9-engineered NF1 cell models, NF1 expression and RAS pathway testing. Genetic testing confirmed NF1 in 82% of clinically diagnosed patients and in 27% of probands undiagnosed by clinical criteria alone, mostly due to younger age (p=0.001). Overall, 155 probands harbored pathogenic/likely pathogenic NF1 variants, including 17 (11%) upgraded from VUS due to their de novo nature. Disease severity correlated with variant type, with truncating/splicing variants showing more complications than missense variants (p=0.02). NF1 VUS were identified in 12 probands (4%), while variants in non-NF1 RASopathy genes in 16 (6%), increasing diagnostic yield by 18% among probands with pigmentary-only features. Among them, we identified LZTR1 as implicated in 9% of pigmentary phenotypes. Functional assays on skin-derived fibroblasts clarified a splicing-altering NF1 VUS (NM_001042492.3:c.6226G>T) without RAS pathway effect, indicating alternative pathogenic mechanisms. We showed an EGF-inducible RAS pathway in RPE-1 and Nthy cells and NF1 H1348P mutant models were generated using Nthy and hiPSC. To dissect factor contributing to disease expressivity, NF1 Y489C iPSC were differentiated into Schwann cells, revealing an unbalanced expression of alternative NF1 transcripts, which may contribute to phenotype modulation (p=0.02). Our findings emphasize careful differential diagnosis with pigmentary-only manifestations, stressing LZTR1 role, highlight the relevance of familial and functional analyses for incomplete NF1 phenotypes, and establish suitable cell models for investigating NF1 VUS and factors modulating disease expressivity.