NM_005228.4:c.2146A>G

EGFR · NP_005219.2:p.(Lys716Glu) · NM_005228.4

GRCh37: chr7:55241698 A>G · GRCh38: chr7:55174005 A>G

Gene: EGFR Transcript: NM_005228.4

Final call

VUS

PM2 supporting

Variant details

Gene

EGFR

Transcript

NM_005228.4

Protein

NP_005219.2:p.(Lys716Glu)

gnomAD AF

ClinVar

Uncertain significance

OncoKB

Unknown Oncogenic Effect

Classification rationale

Interpretation summary

Generated evidence synthesis

NM_005228.4:c.2146A>G (p.Lys716Glu) is a missense variant in EGFR exon 18, encoding part of the kinase domain ATP-binding pocket. The variant is absent from gnomAD v2.1, v4.1, and gnomAD-Canada (PM2_supporting).

In silico predictions are indeterminate: REVEL score 0.426 and BayesDel score 0.075 fall between benign and pathogenic thresholds; SpliceAI predicts no splicing impact (max delta 0.00). Neither PP3 nor BP4 is met.

The variant has been reported once in a somatic context in a patient with lung adenocarcinoma enrolled in a study of non-classical EGFR mutations; p.Lys716Glu was classified as a resistant mutation to afatinib (progressive disease, PFS 1.1 months) (PMID:28676220). This observation does not meet PS3 or PS4 thresholds in a germline variant interpretation context.

No experimental functional data, no segregation data, no de novo reports, and no pathogenic ClinVar classifications exist for this variant. ClinVar reports Uncertain significance from two clinical laboratories.

With a single supporting-level criterion (PM2_supporting) and no other criteria met in either the pathogenic or benign direction, this variant is classified as a Variant of Uncertain Significance (VUS) under generic ACMG/AMP 2015 combination rules.

Final determination: Generic ACMG/AMP 2015 fallback rules do not meet benign, likely benign, likely pathogenic, or pathogenic combination thresholds, so the variant is classified as Variant of Uncertain Significance.

Criteria assessment

ACMG/AMP criteria review

Criteria shown when status is available

All criteria require review: For research and educational purposes only.

Criterion	Status	Rationale	Evidence used
PVS1	N/A	NM_005228.4:c.2146A>G is a missense variant (p.Lys716Glu). PVS1 is reserved for null variants (nonsense, frameshift, or canonical ±1,2 splice site) where loss of function is the established disease mechanism. This substitution does not fall into any null-variant bucket under the ClinGen SVI PVS1 decision tree (PMC6185798).	pvs1_generic_framework
PS1	Not met	No known pathogenic variant at NM_005228.4 nucleotide position c.2146 with a different nucleotide change leading to the same amino acid substitution (p.Lys716Glu) has been identified in ClinVar or the literature. PS1 requires a different nucleotide change at the same codon producing the same missense change with established pathogenicity.
PS2	Not met	No de novo occurrence has been documented for NM_005228.4:c.2146A>G in the available literature or case materials. No germline parent-of-origin or trio sequencing data are available.
PS3	Not met	No experimental functional data (biochemical, cellular, or animal model) exists for p.Lys716Glu. The variant was observed in one patient in PMID:28676220 as a somatic EGFR mutation in lung adenocarcinoma, where it was classified as resistant to afatinib based on clinical outcome (progressive disease, PFS 1.1 months). This constitutes a clinical drug-response observation, not experimental functional evidence. No systematic range characterization (tiling screen, saturation mutagenesis, truncation series) includes this residue. Domain-level inference from exon 18 mutation patterns does not qualify for PS3.	PMID:28676220
PS4	Not met	The variant has been observed in one patient in PMID:28676220 in a somatic NSCLC context and once in COSMIC (COSV104369001). No germline case-control data exist. PS4 requires statistically significant enrichment in affected individuals versus controls, which cannot be established from a single somatic observation.	PMID:28676220
PS5	Not met	ClinVar classifies this variant as Uncertain significance (2 clinical laboratories, variation ID 2681884). No reputable source has classified it as pathogenic. PS5 requires a reputable source to have reported the variant as pathogenic without the primary evidence being available for independent review.	clinvar
PM1	Not met	The variant p.Lys716Glu is located in EGFR exon 18, which encodes part of the ATP-binding pocket within the kinase domain — a critical functional domain. However, residue K716 is not a statistically significant mutational hotspot per cancerhotspots.org, and there is no evidence of pathogenic variant clustering at this specific residue. Domain membership alone, without hotspot-level residue significance, is insufficient for PM1.
PM2	Met	NM_005228.4:c.2146A>G is absent from gnomAD v2.1, gnomAD v4.1, and gnomAD-Canada v1.0. Allele frequency is 0% across all population databases, meeting the PM2 threshold for variants absent or present at extremely low frequency (<0.1%) in large population cohorts.	gnomad_v2 gnomad_v4 gnomad_canada
PM5	N/A	No pathogenic comparator variants at the same amino acid residue (p.Lys716) were identified in ClinVar. The PM5 candidate search returned zero eligible comparators, so same-residue pathogenic comparator assessment cannot be performed.
PM6	Not met	No de novo germline observation has been documented for this variant. The single observation in PMID:28676220 is a somatic finding in a 59-year-old male with lung adenocarcinoma and does not constitute a de novo germline event with confirmed maternity/paternity.	PMID:28676220
PP1	Not met	No segregation data are available for this variant. No family studies or co-segregation analyses have been reported in the literature.
PP2	Not met	HCI prior probability is not available for EGFR (gene not supported by the HCI prior dataset). PP2 requires a gene-specific prior probability of pathogenicity for missense variants, which cannot be assessed without HCI or equivalent data. The variant is a missense change in a disease gene with a low rate of benign missense variation, but without the quantified prior probability, PP2 cannot be applied.
PP3	Not met	In silico predictions are indeterminate: REVEL score is 0.426 (below the 0.5 threshold for deleterious prediction), BayesDel score is 0.075 (below typical deleterious thresholds), and SpliceAI predicts no splicing impact (max delta = 0.00). No individual tool or combination of tools provides supporting evidence for a deleterious effect.	revel bayesdel spliceai
PP4	Not met	No patient phenotype information is available for this variant in a germline context. PP4 requires the patient's phenotype or family history to be highly specific for the disease associated with the gene. The single somatic observation in PMID:28676220 describes a lung adenocarcinoma patient but provides no germline phenotype specificity.	PMID:28676220
PP5	Not met	No reputable source has classified this variant as pathogenic. ClinVar reports Uncertain significance from 2 clinical laboratories. The PMIDs associated with the ClinVar submission (28676220, 35274704, 25394175, 26467025) were reviewed; none provides variant-specific evidence for pathogenicity in a germline context. PMID:25394175 is a general cancer genetics referral guideline, not a variant-specific report.	clinvar PMID:25394175
BA1	Not met	The variant is absent from all gnomAD population databases (v2.1, v4.1, Canada). Allele frequency is 0%, far below the BA1 threshold of >1% in any population.	gnomad_v2 gnomad_v4 gnomad_canada
BS1	Not met	The variant is absent from all gnomAD population databases. Allele frequency is 0%, far below the BS1 threshold of >0.3% in any population.	gnomad_v2 gnomad_v4 gnomad_canada
BS2	Not met	The variant has not been observed in healthy adult controls. It is absent from population databases including gnomAD. No evidence of observation in a healthy individual exists.
BS3	Not met	No functional studies demonstrating a benign effect for p.Lys716Glu exist. The single clinical observation in PMID:28676220 reports drug resistance to afatinib in a somatic NSCLC context, which is a clinical outcome observation and does not constitute experimental functional evidence for or against pathogenicity.	PMID:28676220
BS4	Not met	No segregation data are available for this variant. BS4 requires lack of segregation with disease in affected family members, which cannot be assessed without family-level data.
BP1	Not met	NM_005228.4:c.2146A>G is a missense variant, not a truncating variant. BP1 applies when a truncating variant is observed in a gene where the primary disease mechanism is missense (gain-of-function). This criterion is for truncating variants in genes with alternative mechanisms; a missense variant does not qualify for BP1.
BP2	Not met	No co-occurrence data with a known pathogenic variant in EGFR are available. BP2 requires observation in trans (for recessive disorders) or in cis (for dominant disorders) with a pathogenic variant.
BP4	Not met	In silico predictions are indeterminate: REVEL score 0.426 and BayesDel score 0.075 fall between benign and pathogenic thresholds. SpliceAI predicts no splicing impact (max delta = 0.00). No tool provides strong evidence for a benign effect. BP4 requires multiple lines of computational evidence to suggest no impact on the gene or gene product, which is not established from these indeterminate scores.	revel bayesdel spliceai
BP5	Not met	No evidence that this variant is found in a case with an alternate molecular basis for disease. BP5 requires that the variant is observed in an individual with a different molecular cause for the phenotype, which is not documented.
BP6	Not met	No reputable source has classified this variant as benign. ClinVar reports Uncertain significance. BP6 requires a reputable source to classify the variant as benign or likely benign with the evidence available for independent review.	clinvar
BP7	Not met	NM_005228.4:c.2146A>G is a missense variant (p.Lys716Glu), not a synonymous variant. BP7 applies only to synonymous variants where splicing algorithms predict no impact on the splice consensus sequence.

Disclaimer: The content and results provided by LYFE Sciences are for research and educational purposes only and must not be used as a substitute for professional medical judgment, diagnosis, or treatment. Always consult a qualified healthcare professional before making any clinical decisions.