AXIN2 c.1168A>G, p.(Ser390Gly)

Reasoning

No applicable AXIN2-specific VCEP or CSPEC specification was identified. gnomAD v4.1 reports two homozygotes for the variant (total AF 0.00103457; homozygotes = 2), including two homozygotes in the European (non-Finnish) subgroup. This supports the generic BS2 observation of a variant in ostensibly population-representative adults for an autosomal-dominant condition. However, gnomAD does not provide individual phenotype or age-of-onset confirmation, and AXIN2-associated polyposis may have variable or incomplete penetrance; therefore the evidence is applied with moderate confidence and requires human review of whether the available homozygotes can be treated as unaffected adults for this disease context.

Rule

Generic ACMG/AMP BS2: strong benign evidence when a variant is observed in a healthy adult individual for a fully penetrant dominant disorder; phenotype and penetrance must be considered.

Evidence considered

• gnomAD v4.1 reports two homozygotes overall and two homozygotes in the European (non-Finnish) subgroup.
• The gnomAD record is population data rather than a phenotype-confirmed healthy-adult cohort, so the BS2 interpretation is conditional on the usual population-database assumption and needs human review.
• The supplied case evidence identifies AXIN2-associated disease context but does not establish penetrance or clinical status of the homozygous gnomAD individuals.

Still missing

• Individual phenotype and age-of-onset data for the two gnomAD homozygotes are unavailable.
• The penetrance and disease consequences of AXIN2 homozygosity are not established in the supplied evidence.

Sources

Reasoning

BP1 is met at supporting strength. The variant is a missense substitution, NM_004655.4:c.1168A>G (NP_004646.3:p.(Ser390Gly)), and the available AXIN2 gene-context assessment supports loss of function as the germline disease mechanism. In the absence of an applicable AXIN2-specific VCEP rule limiting BP1, this supports the generic BP1 concept that missense variation is less consistent with the established disease mechanism than truncating loss-of-function variation.

Rule

BP1 (supporting) applies to a missense variant in a gene for which loss of function, rather than missense variation, is the established disease mechanism.

Evidence considered

• The case identifies NM_004655.4:c.1168A>G as the missense change NP_004646.3:p.(Ser390Gly).
• The AXIN2 gene-context assessment supports loss of function as the germline disease mechanism and no AXIN2-specific VCEP rule is available.

Sources

Reasoning

The variant is a missense change, p.(Ser390Gly), so BP4 is applicable. REVEL is 0.286, meeting the supplied generic BP4 supporting threshold of <=0.29 (Pejaver et al. 2022, PMID:36413997). The maximum SpliceAI delta score is 0.004, also meeting the supplied generic BP4 supporting threshold of <=0.1 (Jaganathan et al. 2019, PMID:30661751). The concordant calibrated predictions support BP4 at supporting strength; evidence is not double-counted to assign more than one BP4 strength. BayesDel is not used because no generic ACMG-strength calibration is established and no governing VCEP cutoff is available.

Rule

Generic ACMG/ClinGen SVI computational evidence: BP4 supporting requires SpliceAI max delta <=0.1 or REVEL <=0.29. No calibrated threshold is available for BayesDel.

Evidence considered

• Missense consequence: NM_004655.4:c.1168A>G, NP_004646.3:p.(Ser390Gly).
• REVEL score is 0.286; the supplied generic calibration gives BP4 supporting at <=0.29, from Pejaver et al. 2022 (PMID:36413997).
• SpliceAI maximum delta score is 0.004; the supplied generic calibration gives BP4 supporting at <=0.1, from Jaganathan et al. 2019 (PMID:30661751).
• BayesDel score is -0.266688, but no generic ACMG-strength calibration is established and no governing VCEP cutoff is available, so it is not used for BP4.

Sources

Reasoning

PS1 is not met. The exact variant NM_004655.4:c.1168A>G, NP_004646.3:p.(Ser390Gly), was reported in one targeted sequencing study and classified as class 3 (variant of uncertain significance), not as pathogenic or likely pathogenic. ClinVar shows conflicting single-submitter classifications without an expert-panel pathogenic or likely pathogenic assertion. No established pathogenic same-amino-acid substitution was identified in the reviewed evidence.

Rule

PS1 requires a previously established pathogenic same-amino-acid change caused by a different nucleotide change; that evidence is absent here.

Evidence considered

• PMID:27300758 reports AXIN2 c.1168A>G, p.(S390G) in patient sLS-44 and classifies it as class 3, a variant of uncertain significance.
• ClinVar variation 127934 has conflicting single-submitter classifications, including likely benign, uncertain significance, and benign, with no expert-panel pathogenic assertion.

Still missing

• No validated pathogenic or likely pathogenic comparator producing the same amino-acid change was identified.

Sources

Reasoning

No case-control enrichment, odds ratio, likelihood ratio, or statistically supported excess of the exact AXIN2 NM_004655.4:c.1168A>G (p.Ser390Gly) variant in affected individuals is reported. The exact variant was observed in one suspected Lynch syndrome patient in PMID:27300758, but that single observation is not PS4 enrichment evidence, and the 1,231-case/93-control cohort in PMID:28944238 does not explicitly identify this variant.

Evidence considered

• PMID:27300758 reports the exact variant in one patient (sLS-44) but provides no case-control comparison or enrichment statistic.
• PMID:28944238 studied 1,231 colorectal cancer cases and 93 controls, but the supplied extraction does not explicitly report this exact variant or an exact-variant association statistic.

Still missing

• No exact-variant case-control counts or enrichment statistic were found.

Sources

Reasoning

PM1 is not met. No applicable AXIN2 VCEP or gene-specific domain table is available, so the authoritative domain-table pathway cannot identify an approved critical functional domain. The case evidence also reports no statistically significant hotspot for AXIN2 residue 390. Gene-level WNT/tumor-suppressor context alone does not establish that Ser390 lies in a critical functional domain or mutational hotspot suitable for PM1.

Rule

PM1 requires a well-established critical functional domain or mutational hotspot, without benign variation, appropriate for the gene and disease mechanism; neither is established for Ser390.

Evidence considered

• vcep_materials contains no AXIN2 VCEP specification and its domain_tables object is empty, so no approved critical domain covers residue 390.
• The hotspot assessment reports that AXIN2 S390 does not lie in a statistically significant hotspot.

Still missing

• No applicable AXIN2 VCEP domain table or validated gene-specific critical-domain specification is available.

Sources

Reasoning

No applicable AXIN2-specific VCEP or CSPEC specification was identified, so the generic all-comers gnomAD v4.1 source was used. The variant is present at total AF 0.00103457 and at a maximum subpopulation AF of 0.00134658, exceeding the supplied generic PM2 threshold of <=0.0001. It is also present in gnomAD v2.1 all-comers, gnomAD v2.1 non-cancer exomes (AF 0.000681124), and gnomAD v3.1 non-cancer genomes (AF 0.00071642). Therefore PM2 is not met; the variant is not absent or sufficiently rare under the supplied PM2 calibration.

Rule

Generic ClinGen SVI PM2: supporting evidence when population allele frequency is <=0.0001; governing VCEP specifications would take precedence.

Evidence considered

• No governing AXIN2-specific VCEP or CSPEC specification was available; generic ACMG/AMP defaults therefore apply.
• gnomAD v4.1 reports total AF 0.00103457 (1670/1614190 alleles), with maximum subpopulation AF 0.00134658.
• The supplied generic PM2 calibration is <=0.0001; the observed gnomAD v4.1 frequency is above this threshold.
• The variant is also present in gnomAD v2.1 non-cancer exomes at AF 0.000681124 and gnomAD v3.1 non-cancer genomes at AF 0.00071642, so restricting to non-cancer subsets does not restore PM2.

Sources

Reasoning

PP2 is not met. The available AXIN2 gene-context assessment supports loss of function as the established germline disease mechanism, rather than an established mechanism dominated by pathogenic missense variation. No applicable VCEP specifies PP2 for AXIN2, and the reviewed gene-level sequencing evidence does not establish that pathogenic missense variants are common while benign missense variation is rare.

Rule

PP2 requires a gene in which missense variants are a common disease mechanism and benign missense variation is uncommon; that AXIN2 pattern is not established.

Evidence considered

• The AXIN2 gene-context assessment supports loss of function as the germline disease mechanism.
• PMID:28944238 reports targeted sequencing of AXIN2 in 1,231 colorectal cancer cases and 93 controls, but does not establish a pathogenic-missense-dominant mechanism or rare benign missense variation for AXIN2.

Still missing

• No gene-specific quantitative missense-enrichment or benign-missense constraint analysis supporting PP2 is available.

Sources

Reasoning

The variant is a missense change, p.(Ser390Gly), so PP3 is applicable. The maximum SpliceAI delta score is 0.004, below the generic PP3 supporting threshold of 0.2 and moderate threshold of 0.5 (Jaganathan et al. 2019, PMID:30661751). REVEL is 0.286, below the generic PP3 supporting threshold of 0.644 (Pejaver et al. 2022, PMID:36413997). BayesDel is not usable for PP3 because no generic ACMG-strength calibration is established and no governing VCEP cutoff is available. These calibrated computational results do not meet PP3.

Rule

Generic ACMG/ClinGen SVI computational evidence: PP3 supporting requires SpliceAI max delta >=0.2 or REVEL >=0.644; moderate requires SpliceAI max delta >=0.5 or REVEL >=0.773. No calibrated threshold is available for BayesDel.

Evidence considered

• Missense consequence: NM_004655.4:c.1168A>G, NP_004646.3:p.(Ser390Gly).
• SpliceAI maximum delta score is 0.004; the supplied generic calibration gives PP3 supporting at >=0.2 and moderate at >=0.5, from Jaganathan et al. 2019 (PMID:30661751).
• REVEL score is 0.286; the supplied generic calibration gives PP3 supporting at >=0.644, from Pejaver et al. 2022 (PMID:36413997).
• BayesDel score is -0.266688, but no generic ACMG-strength calibration is established and no governing VCEP cutoff is available, so it is not available for PP3.

Sources

Reasoning

The reported phenotype is colorectal cancer at age 41 with loss of MSH2/MSH6 staining and high microsatellite instability in a suspected Lynch syndrome patient. This is not a highly specific AXIN2 phenotype, and the paper does not document the AXIN2-associated polyposis/oligodontia pattern or a phenotype uniquely attributable to this variant.

Evidence considered

• PMID:27300758 identifies the carrier as having colorectal cancer at age 41, loss of MSH2/MSH6 staining, and high microsatellite instability, with the variant classified as uncertain significance.
• The extracted report does not describe a highly specific AXIN2-associated phenotype such as characteristic polyposis with oligodontia or ectodermal findings.

Still missing

• No documented AXIN2-specific polyposis or oligodontia phenotype for the variant carrier.

Sources

Reasoning

The exact-variant ClinVar record has no expert-panel submission. Its reported Likely benign, Benign, and Uncertain significance classifications are from ordinary laboratory or other non-expert submissions, which cannot trigger PP5 under the specified rule; no exact-variant expert-panel Pathogenic or Likely pathogenic classification is present.

Evidence considered

• ClinVar variation 127934 is an exact match for NM_004655.4:c.1168A>G, but the expert-panel submission set is empty.
• The available classifications are ordinary laboratory or non-expert assertions and are excluded from PP5.

Sources

Reasoning

No applicable AXIN2-specific VCEP or CSPEC specification was identified, so the generic all-comers gnomAD v4.1 source was used. The variant's total allele frequency is 0.00103457 (1670/1614190), with a maximum reported population frequency of 0.00134658 in Europeans (non-Finnish), both far below the generic BA1 stand-alone threshold of >=0.05 from ACMG/AMP 2015 (PMID:25741868). The gnomAD v2.1 all-comers and non-cancer exome results are concordant and also far below 0.05; gnomAD v3.1 non-cancer genomes are likewise below the threshold. Therefore BA1 is not met.

Rule

Generic ACMG/AMP BA1: stand-alone benign when population allele frequency is >=0.05; governing VCEP specifications would take precedence.

Evidence considered

• No governing AXIN2-specific VCEP or CSPEC specification was available; generic ACMG/AMP defaults therefore apply.
• gnomAD v4.1 reports total AF 0.00103457 (1670/1614190 alleles), highest subpopulation AF 0.00134658 in Europeans (non-Finnish), and two homozygotes.
• The supplied generic BA1 calibration is >=0.05 from ACMG/AMP 2015 (PMID:25741868); the observed maximum AF is well below this threshold.
• The supplied gnomAD v2.1 all-comers, gnomAD v2.1 non-cancer exomes, and gnomAD v3.1 non-cancer genomes are also far below the BA1 threshold, indicating no threshold artifact from the population subset choice.

Sources

Reasoning

No applicable AXIN2-specific VCEP or CSPEC specification was identified, so the generic all-comers gnomAD v4.1 source was used. The variant's total allele frequency is 0.00103457, and the highest reported population frequency is 0.00134658 in Europeans (non-Finnish), both below the supplied generic BS1 threshold of >=0.01. The gnomAD v2.1 all-comers and non-cancer exome data and gnomAD v3.1 non-cancer genome data are similarly below 0.01. Thus the population frequency does not support BS1.

Rule

Generic ACMG/AMP BS1: strong benign evidence when population allele frequency is >=0.01; governing VCEP specifications would take precedence.

Evidence considered

• No governing AXIN2-specific VCEP or CSPEC specification was available; generic ACMG/AMP defaults therefore apply.
• gnomAD v4.1 reports total AF 0.00103457 and highest subpopulation AF 0.00134658 in Europeans (non-Finnish).
• The supplied generic BS1 calibration is >=0.01; the observed maximum AF is below this threshold.
• The supplied non-cancer subset results are concordant with the all-comers results and do not alter the BS1 conclusion.

Sources

Reasoning

The exact-variant ClinVar record has no expert-panel submission. Although ordinary submissions report Likely benign or Benign classifications, aggregate ClinVar labels, laboratory submissions, and non-expert assertions are explicitly excluded from BP6; no exact-variant expert-panel Benign or Likely benign classification is present.

Evidence considered

• ClinVar variation 127934 is an exact match and includes ordinary-submission Likely benign and Benign labels, but its expert-panel submission set is empty.
• Non-expert and aggregate ClinVar classifications are excluded from BP6 under the specified rule.

Sources

Reasoning

No proband-parent testing data are available for NM_004655.4:c.1168A>G (NP_004646.3:p.(Ser390Gly)). The exact-variant report in patient sLS-44 does not provide parental genotypes, maternity/paternity confirmation, or evidence that the variant was absent from both parents; therefore de novo status cannot be assessed.

Rule

Generic ACMG/AMP PS2 requires a confirmed de novo occurrence in the patient with maternity and paternity confirmed; no applicable AXIN2-specific specification was found.

Evidence considered

• The exact variant was reported in patient sLS-44, but the report does not provide parental genotypes or a confirmed de novo finding.
• No proband-parent testing, maternity/paternity confirmation, or de novo observation is present in the case evidence.

Still missing

• Parental genotypes for the proband are unavailable.
• Confirmed maternity and paternity are unavailable.
• The parental testing assay and coverage are unavailable.

Sources

Reasoning

No validated functional assay result for AXIN2 NM_004655.4:c.1168A>G (p.Ser390Gly) was identified. The exact variant was reported in a targeted germline sequencing study, but that study used in-silico prediction and database consultation and explicitly did not provide a variant-specific functional assay, assay controls, or a biologically relevant experimental readout. Therefore, PS3 cannot be applied.

Evidence considered

• The exact AXIN2 c.1168A>G (p.Ser390Gly) variant was identified in patient sLS-44 in a targeted germline sequencing study, but the study reported no variant-specific functional assay or experimental readout.
• The study's methodology relied on targeted sequencing, in-silico protein-prediction programs, and LOVD/ClinVar consultation rather than a validated functional assay.

Still missing

• A validated variant-specific functional assay with appropriate positive and negative controls and a biologically relevant readout is needed to assess PS3.

Sources

Reasoning

No applicable AXIN2-specific VCEP specification was identified, so generic ACMG/AMP logic was used. The exact variant NM_004655.4:c.1168A>G (p.Ser390Gly) was reported in one suspected Lynch syndrome patient, but the available report provides no second pathogenic variant, biallelic genotype, phase, trans configuration, or affected-proband observation establishing recessive disease. AXIN2 inheritance information is also not specified in the case materials. PM3 therefore cannot be assessed from the available evidence.

Evidence considered

• PMID:27300758 reports the exact germline AXIN2 c.1168A>G (p.Ser390Gly) variant in patient sLS-44, but gives no second pathogenic variant, phase, trans status, or biallelic disease observation.
• The case materials identify no AXIN2-specific VCEP framework and provide no inheritance-mode specification for applying PM3.

Still missing

• Documented affected-proband biallelic genotype
• A second pathogenic or likely pathogenic AXIN2 variant
• Phase information showing the variants are in trans
• Applicable AXIN2 inheritance mode

Sources

Reasoning

PM5 was not assessed to a reportable conclusion. The PM5 candidate artifact contains no same-residue comparator variants and explicitly states that classic PM5 semantics could not be confirmed safely. The available literature documents the tested p.Ser390Gly variant as a VUS but does not establish a different pathogenic missense change at Ser390. A complete residue-level pathogenic comparator search would be needed before applying or excluding PM5.

Rule

PM5 requires a pathogenic missense change at the same residue caused by a different nucleotide change; the required validated comparator evidence is unavailable.

Evidence considered

• pm5_candidates reports zero same-residue candidates and instructs that PM5 should not be applied from that artifact alone because classic PM5 semantics were not confirmed.
• PMID:27300758 reports the tested p.Ser390Gly variant as class 3/VUS but does not provide a pathogenic alternate missense comparator at Ser390.

Still missing

• A validated search for pathogenic or likely pathogenic alternate missense variants at AXIN2 Ser390 is still required.

Sources

Reasoning

No credible de novo observation without confirmed maternity and paternity is reported for NM_004655.4:c.1168A>G (NP_004646.3:p.(Ser390Gly)). The exact-variant report in patient sLS-44 lacks parental genotype data, so PM6 cannot be assessed.

Rule

Generic ACMG/AMP PM6 applies to a presumed de novo occurrence when maternity and paternity are not confirmed; no applicable AXIN2-specific specification was found.

Evidence considered

• The exact variant was identified in one patient, but the publication does not describe it as de novo or provide parental testing results.
• No case-level evidence establishes de novo origin when maternity and paternity are unconfirmed.

Still missing

• A documented de novo observation is unavailable.
• Parental genotype results are unavailable.
• The extent of maternity and paternity confirmation is unavailable.

Sources

Reasoning

No affected or unaffected relatives with genotype and phenotype data are reported for NM_004655.4:c.1168A>G (NP_004646.3:p.(Ser390Gly)). The exact-variant report describes a single patient and provides no family segregation, informative meioses, or phenotype concordance across relatives; PP1 cannot be assessed.

Rule

Generic ACMG/AMP PP1 requires cosegregation with disease in affected family members; no applicable AXIN2-specific specification or segregation data was found.

Evidence considered

• The exact variant was reported in a single patient, with no reported relatives, familial testing, segregation pattern, or informative meioses.
• No phenotype-genotype concordance across relatives is available for evaluation.

Still missing

• Genotypes of affected relatives are unavailable.
• Genotypes of unaffected relatives are unavailable.
• Number of informative meioses and phenotype concordance are unavailable.

Sources

Reasoning

No validated functional assay result demonstrating preserved AXIN2 function for NM_004655.4:c.1168A>G (p.Ser390Gly) was identified. The exact variant was reported in a targeted germline sequencing study, but no variant-specific functional assay, assay controls, or biologically relevant normal-function readout was provided. Therefore, BS3 cannot be applied.

Evidence considered

• The exact AXIN2 c.1168A>G (p.Ser390Gly) variant was reported in a clinical sequencing study without a variant-specific functional assay or normal-function readout.
• The available study assessed the variant using sequencing, in-silico prediction, and database consultation; these observations do not establish preserved biological function for BS3.

Still missing

• A validated variant-specific functional assay with appropriate controls and a biologically relevant normal-function readout is needed to assess BS3.

Sources

Reasoning

No informative unaffected relatives have been genotyped for NM_004655.4:c.1168A>G (NP_004646.3:p.(Ser390Gly)). The exact-variant report provides no family structure, phenotype assessment of relatives, or genotype results demonstrating non-segregation; BS4 cannot be assessed.

Rule

Generic ACMG/AMP BS4 requires lack of segregation in affected family members or presence in clinically unaffected relatives when penetrance is expected; no applicable AXIN2-specific specification or informative family data was found.

Evidence considered

• The exact variant report contains no testing of unaffected relatives and no family-based non-segregation analysis.
• No phenotype-discordant relatives with reliable genotype data are available to evaluate BS4.

Still missing

• Genotypes of clinically unaffected relatives are unavailable.
• Reliable phenotype assessments of tested relatives are unavailable.
• Informative non-segregating meioses are unavailable.

Sources

Reasoning

No applicable AXIN2-specific VCEP specification was identified, so generic ACMG/AMP logic was used. The exact variant was reported in one patient, but the available evidence contains no documented second pathogenic variant in cis or trans, no co-occurrence analysis, and no unaffected-individual observation that could establish the BP2 configuration. BP2 therefore cannot be assessed from the available evidence.

Evidence considered

• PMID:27300758 reports the exact AXIN2 c.1168A>G (p.Ser390Gly) variant but does not report co-occurring variants, phase, cis/trans configuration, or an unaffected-individual observation.
• No AXIN2-specific VCEP rule modifying BP2 was identified in the case materials.

Still missing

• Co-occurrence data from informative individuals
• A documented second pathogenic or likely pathogenic variant
• Phase information establishing cis or trans configuration
• Relevant unaffected-individual observation, if available

Sources

Reasoning

The suspected Lynch syndrome case has MSH2/MSH6 loss and high microsatellite instability, but the supplied report does not establish a pathogenic alternate molecular etiology, such as an identified pathogenic MSH2 or MSH6 variant, that explains the phenotype independently of AXIN2 c.1168A>G. The BP5 condition therefore cannot be confirmed.

Evidence considered

• PMID:27300758 reports loss of MSH2/MSH6 staining and high microsatellite instability in the patient, but the extracted evidence does not identify a pathogenic alternate MSH2/MSH6 variant or otherwise establish an independent molecular cause.

Still missing

• An identified pathogenic alternate variant or other confirmed molecular etiology explaining the phenotype is not reported.

Sources