When truncation is not loss of function: neo-tail architecture as a determinant of pathogenicity in NMD-escaping frameshift variants

The interpretation of truncating variants in clinical genomics has been dominated by a loss-of-function (LoF) paradigm. Nonsense and frameshift variants are typically assumed to cause disease through nonsense-mediated mRNA decay (NMD) or unstable protein products, leading to haploinsufficiency. This framework underpins ACMG/AMP guideline recommendations, notably the PVS1 criterion and has proven highly effective for many Mendelian disorders [ 1 , 2 ]. However, accumulating evidence suggests that a distinct subset of truncating variants behaves fundamentally differently. Variants occurring within terminal exons or the distal portion of penultimate exons frequently escape NMD, generating stable transcripts that are translated into mutant proteins [ 3 ]. In these circumstances, pathogenicity may arise not from loss of gene dosage, but from the properties of the mutant protein itself.