ERBB2 c.2033G>A, p.Arg678Gln

Reasoning

PS3 is met at supporting strength. PMID:30449325 directly tested ERBB2 p.Arg678Gln (R678Q) in a saturation-mutagenesis screen and orthogonal BaF3 survival/signaling and MCF10A 3D acinar assays, with wild-type HER2 controls and anti-HER2 inhibition, and reported activating effects including enrichment with wild-type HER2, elevated phosphorylated HER2, and abnormal acinar growth. This is variant-specific, biologically relevant, and supported by multiple assay systems, but the evidence is not fully consistent: PMID:23220880 reported little effect across phosphorylation/signaling, 3D culture, soft agar, and xenograft assays, and PMID:24971884 stated that a gain-of-function consequence could not be demonstrated despite increased phosphorylation in some assays. The discordance prevents a stronger calibration.

Rule

Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.

Evidence considered

• PMID:30449325: R678Q was directly tested in HER2 transmembrane/juxtamembrane saturation mutagenesis, BaF3 IL-3-withdrawal survival and signaling assays, and MCF10A 3D Matrigel acinar assays; it was highly enriched with wild-type HER2, associated with elevated phosphorylated HER2, and induced large multi-acinar structures.
• PMID:30449325: Wild-type HER2 controls and anti-HER2 antibody inhibition experiments provided biological relevance and assay controls for the activating interpretation.
• PMID:23220880: R678Q showed little effect on phosphorylation/signaling, formed wild-type-like Matrigel structures, produced fewer soft-agar colonies than wild type, and had xenograft growth indistinguishable from wild type.
• PMID:24971884: R678Q showed enhanced ErbB2 phosphorylation in selected assays, but the authors stated that they and others could not demonstrate a gain-of-function consequence.

Still missing

• Resolve the discordance among orthogonal functional studies using a standardized, independently replicated assay panel with prespecified wild-type and known-activating controls.

Sources

Reasoning

Cancer Hotspots reports that ERBB2 R678 lies in a statistically significant hotspot. Independent variant-specific literature places R678Q in the ERBB2 transmembrane/juxtamembrane region and reports recurrent activating behavior, supporting a critical functional region. The hotspot extraction is flagged as partially parsed and should be visually confirmed.

Rule

Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.

Evidence considered

• Cancer Hotspots reports: "This variant lies in a statistically significant hotspot." The record is for ERBB2 R678.
• PMID:24971884 states that R678Q is located near the transmembrane domain.
• PMID:30449325 explicitly studies R678Q among recurrent HER2 juxtamembrane-domain mutations and reports activating functional effects in cellular assays.

Still missing

• The Cancer Hotspots result should be visually confirmed because its structured extraction was partial.

Sources

Reasoning

PM2 is met at supporting strength. The generic SVI threshold supplied for this case is AF <=0.0001. The variant is absent from gnomAD v2.1 all-comers and its reported non-cancer subset, absent from the reported gnomAD v3.1 non-cancer subset, and present only once in gnomAD v4.1 all-comers (total AF=6.19525715892941e-07; 1/1614138 alleles; homozygotes=0). This frequency is below 0.0001 and supports an extremely rare population allele. No VCEP-specific population source or threshold supersedes the generic rule.

Rule

PM2 (supporting): <= 0.0001 — SVI recommendation to downgrade PM2 to Supporting (ClinGen SVI 2020) (PMID:25741868) ... Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.

Evidence considered

• gnomAD v2.1 all-comers: absent; reported gnomAD v2.1 non-cancer subset: absent.
• Reported gnomAD v3.1 non-cancer subset: absent.
• gnomAD v4.1 all-comers: total AF=6.19525715892941e-07, 1/1614138 alleles, homozygotes=0; exome AC=1/1461876 and genome AC=0.
• Published generic calibration threshold supplied for this case: PM2 <=0.0001, described as the SVI recommendation to downgrade PM2 to Supporting (ClinGen SVI 2020; threshold attribution supplied with PMID:25741868).

Sources

Reasoning

No reviewed source provides a case-control comparison for ERBB2 NM_004448.4:c.2033G>A (p.Arg678Gln) in affected versus unaffected individuals. PMID:30449325 reports the variant in 192 tumors from a 111,176-tumor somatic cohort, but does not report an unaffected-control comparison or a germline disease case-control analysis. The other reviewed variant-specific papers report functional assays or a statement that four cases had been documented, without controls. gnomAD v4.1 reports 1/1,614,138 alleles overall (AF 6.19526e-07) and 1/1,180,028 European non-Finnish alleles (AF 8.47438e-07), but no affected-case frequency is available; therefore an OR and 95% CI cannot be computed. The ACMG/AMP PS4 threshold is OR or RR >5.0 with a confidence interval excluding 1.0, and no such estimate exists here.

Rule

The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. ... Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0.

Evidence considered

• PMID:30449325, targeted-exome analysis of 111,176 tumors: R678Q was observed in 192 tumors; the extracted full text reports no unaffected or population control comparison for this exact variant.
• PMID:24971884, four previously documented R678Q cases are mentioned, but no affected-versus-unaffected control counts or frequency comparison are provided.
• PMID:23220880, R678Q is studied in somatic functional models and a patient tumor context, without case-control prevalence data.
• PMID:29420467 and PMID:29967253, reviewed for variant-specific prevalence evidence; neither reports R678Q case-control data, and PMID:29967253 does not explicitly mention R678Q.
• gnomAD v4.1: 1/1,614,138 alleles overall (AF 6.19526e-07); European non-Finnish 1/1,180,028 alleles (AF 8.47438e-07). No affected-case frequency is available, so OR and 95% CI were not calculable.

Still missing

• Affected-case and matched-control counts or frequencies for this exact variant are unavailable, so OR and 95% CI cannot be calculated.

Sources

Reasoning

The variant is a missense substitution, NM_004448.4:c.2033G>A (p.Arg678Gln), so the generic missense path uses REVEL only; SpliceAI is not used for PP3/BP4 in this situation. REVEL is 0.417, below the generic PP3 supporting threshold of >= 0.644. BayesDel is not used because no generic ACMG-strength calibration is established.

Rule

Missense variant -> REVEL only. Its SpliceAI score says nothing about whether the amino-acid substitution is tolerated, so 'no predicted splice impact' is never BP4 support for a missense variant. ... PP3 (supporting): >= 0.644 — ClinGen SVI REVEL calibration (Pejaver et al. 2022, Am J Hum Genet) (PMID:36413997)

Evidence considered

• Local REVEL predictor lookup: REVEL score 0.417 for NM_004448.4:c.2033G>A (p.Arg678Gln).
• Generic ClinGen SVI REVEL calibration: PP3 (supporting) requires >= 0.644; Pejaver et al. 2022, Am J Hum Genet (PMID:36413997).
• SpliceAI max delta 0.012 was not used because this is a missense variant and the governing generic rule assigns missense variants to the REVEL-only path.
• BayesDel score -0.0912836 was not used because no generic ACMG-strength calibration is established.

Sources

Reasoning

The exact variant is listed in ClinVar as Likely oncogenic with one submission, criteria provided, single submitter, one review star, and somatic origin. This is not an expert-panel Pathogenic or Likely pathogenic classification. OncoKB labels the variant Oncogenic, but that is not an exact-variant ClinVar expert-panel classification and cannot trigger generic PP5. The reviewed papers provide functional or somatic oncology observations, not the required reputable-source classification for PP5.

Rule

Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.

Evidence considered

• ClinVar variation 376344: Likely oncogenic, criteria provided/single submitter, one star, one submission, somatic origin; no expert-panel classification is present.
• OncoKB: variant-specific Oncogenic label and gain-of-function context, but this is not an exact-variant ClinVar expert-panel assertion and is not used to apply PP5.
• PMID:30449325: reports activating oncogenic functional effects for R678Q, but does not constitute an exact-variant ClinVar expert-panel classification.

Sources

Reasoning

BA1 is not met. The generic stand-alone threshold supplied for this case is allele frequency >=0.05. The highest reported all-comers gnomAD v4.1 frequency is 6.19525715892941e-07 (1/1614138 alleles), and the highest observed ancestry-specific point estimate is 8.47437518431766e-07 in European (non-Finnish) individuals (1/1180028 alleles), both far below 0.05. gnomAD v2.1 all-comers and its reported non-cancer subset are absent. The gnomAD v4.1 grpmax FAF is not reported; the low observed frequencies do not support stand-alone benign evidence.

Rule

BA1 (stand-alone): >= 0.05 — ACMG/AMP 2015 stand-alone benign allele frequency (PMID:25741868) ... Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.

Evidence considered

• gnomAD v4.1 all-comers: total AF=6.19525715892941e-07, 1/1614138 alleles, homozygotes=0; European (non-Finnish) point estimate=8.47437518431766e-07, 1/1180028 alleles.
• gnomAD v2.1 all-comers and reported non-cancer subset: absent.
• Published generic calibration threshold supplied for this case: BA1 >=0.05, attributed to ACMG/AMP 2015 (PMID:25741868).

Sources

Reasoning

BS1 is not met on the available population evidence. The generic default threshold supplied for this case is >=0.01, while the gnomAD v4.1 all-comers frequency is 6.19525715892941e-07 and the highest ancestry-specific point estimate is 8.47437518431766e-07, both far below 1%. The variant is absent from gnomAD v2.1 all-comers and the reported non-cancer subset, and from the reported gnomAD v3.1 non-cancer subset. No elevated frequency is present to justify BS1 or a maximum-credible-frequency calculation. The gnomAD v4.1 grpmax FAF is not reported, and the ancestry-specific point estimate is not used as a substitute for FAF.

Rule

BS1 (strong): >= 0.01 — Generic SVI default; per-gene derivation preferred (Whiffin et al. 2017) (PMID:28518168) ... Allele frequency is greater than expected for the disorder.

Evidence considered

• gnomAD v4.1 all-comers: total AF=6.19525715892941e-07, 1/1614138 alleles; highest ancestry-specific point estimate=8.47437518431766e-07 in European (non-Finnish) individuals, 1/1180028 alleles; grpmax FAF is not reported.
• gnomAD v2.1 all-comers and reported non-cancer subset: absent; reported gnomAD v3.1 non-cancer subset: absent.
• Published generic calibration threshold supplied for this case: BS1 >=0.01; per-gene maximum-credible-frequency derivation is preferred (Whiffin et al. 2017, PMID:28518168).

Sources

Reasoning

BS3 is not met because the available functional evidence does not consistently show no damaging effect. PMID:23220880 reported no functional effect for R678Q across phosphorylation/signaling, 3D Matrigel, soft agar, and NIH3T3 xenograft assays, and PMID:24971884 reported inability to demonstrate gain of function, but PMID:30449325 directly demonstrated activating effects in multiple orthogonal assays with wild-type controls and pharmacologic response. The contradictory variant-specific results preclude a conclusion that well-established studies show no damaging effect on protein function.

Rule

Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.

Evidence considered

• PMID:23220880: R678Q showed little effect on phosphorylation/signaling and no enhancement of Matrigel, soft-agar, or xenograft growth relative to wild-type HER2.
• PMID:24971884: The authors reported that they and others were unable to demonstrate a gain-of-function consequence for R678Q, despite enhanced phosphorylation in selected assays.
• PMID:30449325: R678Q was enriched with wild-type HER2 and produced elevated phosphorylated HER2, BaF3 survival signaling, and abnormal MCF10A acinar structures, contradicting a uniform no-damaging-effect interpretation.

Still missing

• Resolve the discordance among orthogonal functional studies using a standardized, independently replicated assay panel with prespecified wild-type and known-activating controls.

Sources

Reasoning

The variant is a missense substitution, NM_004448.4:c.2033G>A (p.Arg678Gln), so the generic missense path uses REVEL only; SpliceAI is not used for PP3/BP4 in this situation. REVEL is 0.417, above the generic BP4 supporting threshold of <= 0.29, so BP4 is not met. BayesDel is not used because no generic ACMG-strength calibration is established.

Rule

Missense variant -> REVEL only. Its SpliceAI score says nothing about whether the amino-acid substitution is tolerated, so 'no predicted splice impact' is never BP4 support for a missense variant. ... BP4 (supporting): <= 0.29 — ClinGen SVI REVEL calibration (Pejaver et al. 2022, Am J Hum Genet) (PMID:36413997)

Evidence considered

• Local REVEL predictor lookup: REVEL score 0.417 for NM_004448.4:c.2033G>A (p.Arg678Gln).
• Generic ClinGen SVI REVEL calibration: BP4 (supporting) requires <= 0.29; Pejaver et al. 2022, Am J Hum Genet (PMID:36413997).
• SpliceAI max delta 0.012 was not used because this is a missense variant and the governing generic rule assigns missense variants to the REVEL-only path.
• BayesDel score -0.0912836 was not used because no generic ACMG-strength calibration is established.

Sources

Reasoning

ClinVar contains one somatic single-submitter Likely oncogenic assertion with criteria provided and one review star, not an exact-variant expert-panel Benign or Likely benign classification. No qualifying reputable expert-panel benign classification is present, so BP6 is not met.

Rule

Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.

Evidence considered

• ClinVar variation 376344: Likely oncogenic, criteria provided/single submitter, one star, one submission, somatic origin; no expert-panel Benign or Likely benign classification is present.

Sources

Reasoning

The variant is p.Arg678Gln, but the supplied evidence does not document a previously established pathogenic variant producing the same amino-acid change through a different nucleotide change. The available reports describe somatic oncogenic or functional observations and are not established pathogenic classifications for PS1.

Rule

Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.

Evidence considered

• The case variant is NM_004448.4:c.2033G>A, NP_004439.2:p.(Arg678Gln).
• PMID:23220880 and PMID:24971884 report p.R678Q in somatic cancer or cell-line contexts, while ClinVar records the variant as likely oncogenic with a single somatic submission; none establishes PS1-level pathogenicity for the same amino-acid change.

Still missing

• A clinically established pathogenic variant with the same p.Arg678Gln amino-acid change, but a different nucleotide change, is needed.

Sources

Reasoning

The available case evidence does not report an affected proband with a documented negative family history together with confirmed maternity and paternity, nor does it provide parental genotypes. The variant-specific literature reviewed concerns somatic ERBB2 R678Q in tumors and does not establish a germline de novo observation. Therefore, the required PS2 evidence cannot be assessed.

Rule

De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.

Evidence considered

• No case-level proband, family-history, maternity, paternity, or parental-testing evidence is present in the available case records.
• The variant-specific publications available for review describe somatic ERBB2 R678Q in cancer models or tumors, not a germline de novo patient observation.

Still missing

• Affected proband phenotype and diagnosis are not documented.
• Family history is not documented.
• Maternal and paternal relationships are not confirmed.
• Parental variant testing results are not reported.

Reasoning

PM3 cannot be assessed because the available evidence does not document a recessive disease context, an affected proband carrying NM_004448.4:c.2033G>A with a pathogenic variant in the same gene, or parental/offspring testing establishing trans phase.

Rule

For recessive disorders, detected in trans with a pathogenic variant ... This requires testing of parents (or offspring) to determine phase.

Evidence considered

• The case evidence documents NM_004448.4:c.2033G>A (NP_004439.2:p.Arg678Gln) but contains no affected-proband observation with a second pathogenic ERBB2 variant and no parental or offspring phase result.
• The available variant-specific literature concerns somatic tumor or cell-line studies and does not establish a germline affected-proband trans observation for this criterion.

Still missing

• A defined recessive ERBB2 disease context applicable to this variant.
• An affected proband carrying NM_004448.4:c.2033G>A and a second pathogenic ERBB2 variant.
• Parental or offspring testing establishing that the variants are in trans.

Reasoning

The supplied evidence reports alternate missense substitutions R678P and R678W at the same residue, but does not establish either alternate change as pathogenic under the generic ACMG/AMP rule. Somatic recurrence or activating functional evidence alone does not establish the required PM5 comparator.

Rule

Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.

Evidence considered

• PMID:30449325 reports alternate same-residue substitutions R678P and R678W, including activating functional evidence for R678W, but does not establish either as a pathogenic ACMG/AMP comparator.
• PMID:24971884 discusses R678Q and prior cases, not a different pathogenic Arg678 substitution; OncoKB provides variant-specific oncogenic context but no qualifying PM5 comparator in the supplied evidence.

Still missing

• A different missense variant at residue Arg678 with an established pathogenic determination is needed.

Sources

Reasoning

No case-level evidence identifies this variant as assumed de novo, and no affected proband, family history, or parental testing information is provided. The available variant-specific publications concern somatic ERBB2 R678Q and do not support an assumed germline de novo event. Therefore, PM6 cannot be assessed.

Rule

Assumed de novo, but without confirmation of paternity and maternity.

Evidence considered

• No case-level de novo assertion or parental-testing result is present in the available case records.
• The available variant-specific publications describe somatic ERBB2 R678Q observations and do not establish assumed germline de novo status.

Still missing

• A germline affected proband observation is not documented.
• An assumed de novo interpretation is not documented.
• Parental variant testing results are not reported.

Reasoning

No affected family members, informative meioses, variant-positive relatives, or cosegregation results are reported. The available literature concerns somatic ERBB2 R678Q in cancer and does not provide family segregation data for a disease caused by this variant. Therefore, PP1 cannot be assessed.

Rule

Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.

Evidence considered

• No family pedigree, affected-relative phenotypes, relative genotypes, or segregation results are present in the available case records.
• The available variant-specific publications report somatic cancer observations and do not provide familial cosegregation data.

Still missing

• The relevant disease and affected family members are not documented.
• Variant genotypes of relatives are not reported.
• Informative meioses and segregation outcomes are not reported.

Reasoning

The supplied evidence does not establish both PP2 requirements for ERBB2: a low rate of benign missense variation and missense variants as a common disease mechanism in the relevant clinical context. Somatic activating studies and one very rare population observation do not provide the required gene-level assessment.

Rule

Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.

Evidence considered

• gnomAD v4.1 reports this variant at total AF 6.19526e-07, with 1 allele among 1,614,138 alleles; this is variant-level rarity and does not establish the PP2 gene-level benign-missense requirement.
• PMID:30449325 demonstrates recurrent activating ERBB2 somatic missense variants, but does not establish the generic PP2 gene-level clinical mechanism requirement.

Still missing

• A gene-level analysis establishing low benign missense variation is missing.
• A gene-level clinical analysis establishing missense variants as a common disease mechanism is missing.

Sources

Reasoning

The supplied case evidence does not provide the patient's phenotype, family history, or a defined germline disease context that can be evaluated for high specificity. The reviewed ERBB2 literature concerns somatic tumors and functional studies rather than a phenotype in this case. PP4 therefore cannot be assessed.

Rule

Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.

Evidence considered

• The case evidence identifies ERBB2 p.Arg678Gln and reviewed publications describe somatic cancer cohorts or functional assays, but no patient phenotype or family-history data are supplied for this case.

Still missing

• Patient phenotype and family history are missing.
• A specific germline disease diagnosis with a single genetic etiology has not been established for this case.

Reasoning

BS2 cannot be assessed from the available evidence. gnomAD v4.1 contains one observed allele and zero homozygotes, but the record does not establish that the carrier was a healthy adult, provide phenotype or age information, or establish the disease's required full penetrance and early-age presentation for this variant. The available ClinVar record is a single somatic likely-oncogenic submission and is not evidence of a healthy adult germline observation.

Rule

Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.

Evidence considered

• gnomAD v4.1: total AC=1, AN=1614138, total AF=6.19525715892941e-07, homozygotes=0; no individual-level age or phenotype data are provided.
• ClinVar variation 376344: single-submitter, one-star, somatic likely-oncogenic record; it does not establish a healthy adult germline observation.

Still missing

• A verified healthy adult carrier observation with phenotype and age information is unavailable.
• Disease-specific inheritance, penetrance, and age-of-onset requirements for applying BS2 are not established in the available evidence.

Sources

Reasoning

The available evidence does not identify affected family members who were tested for the variant, and it provides no genotype-phenotype discordance or non-segregation result. Consequently, lack of segregation cannot be established and BS4 cannot be assessed.

Rule

Lack of segregation in affected members of a family.

Evidence considered

• No tested affected relatives or non-segregation observations are present in the available case records.
• The available variant-specific publications do not report a family with affected members who lack the variant.

Still missing

• Affected family members are not documented.
• Variant testing results for affected relatives are not reported.
• Phenotype-genotype discordance and possible phenocopies cannot be evaluated.

Reasoning

The supplied evidence does not establish that ERBB2 disease is caused primarily by truncating variants. Recent evidence describes a germline HER2 deficiency disorder, while the variant-specific literature concerns activating somatic missense variants; neither source provides the systematic disease-mechanism distribution required for BP1.

Rule

Missense variant in a gene for which primarily truncating variants are known to cause disease.

Evidence considered

• PMID:42060361 describes a HER2 deficiency developmental disorder but does not establish that truncating variants are the primary disease mechanism.
• PMID:30449325 concerns recurrent activating ERBB2 missense variants in tumors and does not support the BP1 gene-level premise.

Still missing

• A systematic clinical spectrum or curated disease-mechanism source documenting predominantly truncating ERBB2 disease variants is missing.

Sources

Reasoning

BP2 cannot be assessed because the available evidence does not document this variant in trans with a pathogenic variant in a fully penetrant dominant disorder or in cis with a pathogenic variant in any inheritance pattern; no phase-resolved affected-proband or family observation is provided.

Rule

Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.

Evidence considered

• No phase-resolved cis or trans observation with a pathogenic variant is documented in the case evidence.
• Reported co-occurrence in somatic tumor literature is not sufficient to establish cis or trans phase, a pathogenic partner classification, or the inherited dominant-disorder context required for BP2.

Still missing

• A documented pathogenic variant occurring with NM_004448.4:c.2033G>A in the same individual.
• Confirmed cis or trans phase between the two variants.
• For a trans observation, evidence that the relevant disorder is fully penetrant and dominant.

Reasoning

The available evidence documents ERBB2 p.Arg678Gln in somatic cancers and functional studies, but does not document this variant in a patient with an alternate molecular basis for the disease being evaluated. Somatic tumor occurrence alone is not evidence of the BP5 case scenario. No alternate molecular diagnosis for a specific case is supplied, so BP5 cannot be assessed.

Rule

Variant found in a case with an alternate molecular basis for disease.

Evidence considered

• PMID:23220880 and PMID:24971884 describe R678Q in somatic cancer models or tumor-derived contexts, but do not establish an alternate molecular basis for a clinical disease case.
• PMID:30449325 reports R678Q among somatic tumors and functional assays, without the BP5 alternate-molecular-basis case scenario.
• COSMIC records somatic occurrence (COSV54062926, n=145) but does not provide a clinical case with an alternate molecular basis for the disease.

Still missing

• A clinical case with an established alternate molecular basis for disease is not provided.

Sources