NM_024675.3:c.1684+1G>A is a canonical +1 donor splice site variant in intron 4 of PALB2, a gene for which loss of function is an established mechanism for hereditary breast, ovarian, and pancreatic cancer predisposition (ClinGen PALB2 VCEP v1.2.0).

This variant is absent from gnomAD v2.1 and v4.1 population databases, satisfying the PALB2 VCEP PM2_Supporting threshold of ≤0.000333% allele frequency (PM2_Supporting).

RNA analysis performed by Lopez-Perolio et al. (2019, PMID:30890586) confirmed aberrant splicing for c.1684+1G>A, supporting classification as a null variant under the PALB2 VCEP PVS1 decision tree at Very Strong strength (PVS1).

The variant introduces a premature termination codon upstream of p.Tyr1183 and is predicted to be NMD-prone; combined with RNA-confirmed splice defect, PM5_Supporting is applied per PALB2 VCEP rules (PM5_Supporting).

No benign criteria are met: the variant is not present at significant population frequency (BA1/BS1 not met), SpliceAI predicts a strong splice impact (BP4 not met), the variant is not at a position eligible for BP7, and no evidence of normal splicing or lack of segregation is available.

Applying the ACMG/AMP combination rules (Richards et al. 2015) as adopted by the PALB2 VCEP: 1 Pathogenic Very Strong (PVS1) + ≥2 Pathogenic Supporting (PM2_Supporting, PM5_Supporting) yields a classification of Pathogenic under Rule 4.

Note: The ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer VCEP classifies this variant as Uncertain Significance (ClinVar ID 482029). The discrepancy between the automated adjudication (Pathogenic) and the expert panel classification (VUS) may reflect additional VCEP-specific PVS1 decision tree nuances, PS1 splicing table consultation, or clinical judgment not fully captured in the available evidence materials. Human review by a PALB2 VCEP curator is recommended.