NM_139276.2:c.1940A>T (p.Asn647Ile) in STAT3 was assessed using generic ACMG/AMP 2015 criteria (PMID:25741868). No CSPEC/VCEP framework is available for STAT3.¹ This missense variant is located in the SH2 domain (residues 583–688), a well-established mutational hotspot where numerous pathogenic variants have been identified in both germline hyper-IgE syndrome and somatic lymphoproliferative disorders, satisfying PM1 at moderate strength.² The variant is absent from gnomAD v2.1 (0/251,490 alleles) and extremely rare in gnomAD v4.1 (2/1,614,212 alleles; AF=1.24×10⁻⁶), satisfying PM2 at supporting strength.³ In silico predictors do not reach consensus for pathogenicity (REVEL 0.359; BayesDel −0.126; SpliceAI 0.00), and do not reach consensus for benign impact either — PP3 and BP4 are not met.⁴ ClinVar classification is conflicting: two clinical laboratories report Uncertain Significance and one reports Pathogenic. No expert panel review is available. PP5 is not met due to conflicting interpretations, and BP6 is not met as no submitter classifies the variant as Benign.⁵ Several criteria (PS3, PS4) could not be fully assessed because variant-specific evidence from cited publications could not be confirmed without full-text access. PMID:27345172 ('Distinct mutations at the same positions of STAT3 cause either loss or gain of function') is of particular relevance for functional evidence but has no abstract or full-text available in the evidence packet. With 1 moderate criterion (PM1) and 1 supporting criterion (PM2) met, and no benign criteria met, the variant does not reach the Likely Pathogenic threshold (requires at least 1 strong + 1–2 moderate, or 3 moderate, or 2 moderate + 2 supporting). The variant is classified as a Variant of Uncertain Significance (VUS) according to generic ACMG/AMP 2015 combination rules.⁶