PIK3CA c.3140A>G, p.His1047Arg

Reasoning

PS3_Supporting is met. The exact PIK3CA p.H1047R variant was tested in multiple functional systems. PMID:16322248 reported 2.5- to 3.5-fold higher PI3K activity than wild type in MCF-10A cells, persistent Akt phosphorylation after growth-factor withdrawal, EGF-independent proliferation, anchorage-independent colony formation, anoikis resistance, abnormal Matrigel acinar structures, and inhibitor-sensitive activity. PMID:15016963 reported greater lipid kinase activity than wild type in an NIH3T3 assay. PMID:17376864 reported transformation, increased lipid kinase activity, and constitutive Akt/TOR signaling. PMID:16432179 reported H1047R-induced tumors in chickens, with 4 of 5 animals developing tumors, increased Akt phosphorylation, and suppression by RAD001. PMID:19029981 demonstrated H1047R-driven lung adenocarcinoma in an inducible mouse model and regression after doxycycline withdrawal, showing dependence on continued H1047R expression. These studies provide concordant damaging functional evidence for the exact variant. The VCEP permits supporting PS3 from an eligible publication; evidence above supporting cannot be assigned because the available extracts do not document the VCEP-required positive and negative validation controls and replicate details needed for moderate or strong evidence.

VCEP rule

Follow recommendations set forth by the SVI in conjunction with specifications added by the BMVCEP for quality metrics and minimum validation controls required (PMID: 31892348). ... Any publication within the spreadsheet can be used as evidence for a supporting level of evidence (PS3). ... Any paper must have validation controls (positive and negative) in order to be used as evidence for a level above supporting.

Evidence considered

• ClinGen Brain Malformations VCEP Supplemental Document 1, Functional Assay Validation: any eligible publication may support PS3; evidence above supporting requires positive and negative validation controls independent of PS3/BS3.
• PMID:16322248, Results, Figure 2A: exact H1047R showed PI3K activity 2.5- to 3.5-fold higher than WT in MCF-10A cells, with downstream growth and transformation phenotypes.
• PMID:15016963, Results, Fig. S2: expression of mutant H1047R p110-alpha conferred more lipid kinase activity than wild-type protein in NIH3T3 cells.
• PMID:17376864, Results/Table 1 and Fig. 2: H1047R showed transformation, increased lipid kinase activity, and constitutive Akt/TOR signaling in chicken embryo fibroblast assays.
• PMID:16432179, Results: 4 of 5 animals treated with H1047R developed tumors, with increased Akt signaling and RAD001-sensitive tumor growth.
• PMID:19029981, Results: inducible H1047R expression produced lung adenocarcinomas, and doxycycline withdrawal caused rapid tumor regression, demonstrating dependence on continued H1047R expression.

Still missing

• The available records do not document the positive and negative validation controls required by the VCEP for PS3 above supporting.
• The available records do not document the replicate details required by the VCEP for PS3 above supporting.
• No exact-variant entry was found in clingen_brainmalform_acmg_specifications_v1_1.pdf.txt when searched for c.3140A>G, p.His1047Arg, and p.H1047R; the criterion was evaluated from the file's governing PS3 rule and the case's functional publications.

Sources

Reasoning

The governing VCEP requires PM2 before PS4 and assigns phenotype points using Table 2A, with PS4_Moderate at 1.5-3.25 points. PM2 is supported: the variant is absent from gnomAD v4.1 and occurs once in gnomAD v2.1 (1/248,274 alleles; zero homozygotes). Table 2A assigns 0.25 points per tumor occurrence and permits COSMIC; the COSMIC record reports 3,986 occurrences, but the VCEP caps counting at 9 tumor samples per variant, yielding 9 × 0.25 = 2.25 points. The explicit inequality is 1.5 <= 2.25 <= 3.25? yes, so PS4_Moderate is met. The reviewed cancer papers document exact H1047R tumor observations, but they do not report unaffected or population controls, so no odds ratio or 95% confidence interval could be computed; the VCEP's pre-assigned tumor-occurrence point path was used instead.

VCEP rule

Points are assigned for phenotype according to (Table 2A). Phenotype criteria can only be used if the variant is absent from controls (PM2). Strength of evidence is determined by points according to (Table 2B). PS4_Moderate = 1.5-3.25 points ... Presence of this variant in a tumor sample (databases such as COSMIC can be used) ... A maximum of 9 tumor samples can be counted for an individual variant keeping with the idea that tumor data cannot be utilized for any strength above supporting.

Evidence considered

• ClinGen Brain Malformations VCEP specification, Table 2A: 'Presence of this variant in a tumor sample (databases such as COSMIC can be used)' = 0.25 points; 'A maximum of 9 tumor samples can be counted for an individual variant'.
• COSMIC record for PIK3CA c.3140A>G/p.H1047R: COSV55873195, somatic count n = 3986; VCEP cap applied as 9 counted samples × 0.25 = 2.25 points.
• PM2 prerequisite: gnomAD v4.1 absent; gnomAD v2.1 has 1 allele among 248274 alleles and 0 homozygotes.
• PMID:15016963: exact H1047R was tested and reported as an activating mutant; the paper does not provide variant-specific controls.
• PMID:15520168, Table 1: A3140G/H1047R was identified in 13 primary cancers; no unaffected or population-control comparison was reported.
• PMID:15608678, Table 1: A3140G/H1047R was identified in 27 tumors; no unaffected or population-control comparison was reported.
• Explicit threshold check: 1.5 <= 2.25 <= 3.25? yes.

Still missing

• The reviewed case-series papers do not provide unaffected or population-control counts, so a case-control odds ratio and 95% confidence interval could not be calculated.

Sources

Reasoning

The governing VCEP applies PM2 at Supporting strength for a variant absent or rare in an ethnically matched control cohort, with a one-person maximum to account for sequencing and calling artifacts. The variant is present in gnomAD v2.1 exomes at 1/248274 alleles with 0 homozygotes; the highest population count is also one allele, and it is absent from gnomAD v4.1. This satisfies the VCEP's one-person maximum and supports PM2 at Supporting strength.

VCEP rule

Absent/rare from controls in an ethnically-matched cohort population sample ( ≤1).

Evidence considered

• ClinGen Brain Malformations VCEP specification: PM2 is Supporting for an absent/rare variant in an ethnically matched cohort population sample, with one person maximum; the detailed recommendation states that this accounts for possible sequencing and calling errors.
• gnomAD v2.1 exomes: 1/248274 alleles, AF 4.027807986337675e-06, 0 homozygotes; European (non-Finnish) maximum 1/112236 alleles, AF 8.909797213015431e-06, 0 homozygotes.
• gnomAD v4.1: variant absent.

Still missing

• The governing file was searched for c.3140A>G, p.His1047Arg, and p.H1047R; no exact variant-specific table entry was found in clingen_brainmalform_acmg_specifications_v1_1.pdf.

Sources

Reasoning

Expert panel ClinGen Brain Malformations Variant Curation Expert Panel classified as Pathogenic.

VCEP rule

Reputable source recently reports variant as pathogenic but the evidence is not available to the laboratory to perform an independent evaluation ... Brain Malformation Recommendation: Not Applicable

Evidence considered

• ClinGen Brain Malformations VCEP specification, PP5 row: the criterion is listed as Not Applicable.
• ClinVar expert panel classification

Sources

Reasoning

Residue 1047 is within the authoritative PIK3CA kinase-domain interval AA 797-1068 in Table 4. However, the VCEP specifically excludes residues subject to recurrent gain-of-function mutations from PM1_Supporting and directs those recurrent residues to PS4. H1047R is repeatedly reported in the available extracts, including 13 tumors in PMID:15520168 and 27 tumors in PMID:15608678, so the VCEP exception applies despite domain membership.

VCEP rule

Brain Malformation Recommendation: A supporting point may be awarded if the variant of interest is located in an exon within one of the below approved functional domains Table 4. Please note that specific residues subject to recurrent gain-of-function mutations are not covered by this criterion; these are instead accounted for by PS4.

Evidence considered

• ClinGen Brain Malformations VCEP Table 4 lists PIK3CA kinase domains at AA 322-483 and AA 797-1068; residue 1047 falls within AA 797-1068.
• ClinGen Brain Malformations VCEP PM1_Supporting rule states that a supporting point may be awarded for an approved functional-domain exon, but specific residues subject to recurrent gain-of-function mutations are not covered and are instead accounted for by PS4.
• PMID:15520168, Results and Discussion/Table 1: A3140G/H1047R was identified in 13 primary cancers and described as the most common mutation in that study.
• PMID:15608678, Results/Discussion/Table 1: A3140G/H1047R was identified in 27 tumors and described as the most common PIK3CA mutation in breast and gastric cancers.

Sources

Reasoning

The governing VCEP BA1 threshold is an allele frequency greater than 0.0926%. The highest documented gnomAD v2.1 population frequency is 8.909797213015431e-06 in the European (non-Finnish) population, equivalent to 0.00089098%, and the overall frequency is 4.027807986337675e-06 (0.00040278%); gnomAD v4.1 reports the variant as absent. These frequencies are below the VCEP BA1 threshold, so BA1 is not met.

VCEP rule

An allele frequency (>0.0926%) was approved.

Evidence considered

• ClinGen Brain Malformations VCEP specification, benign criteria: BA1 approved allele-frequency threshold is >0.0926%.
• gnomAD v2.1 exomes: 1/248274 alleles, overall AF 4.027807986337675e-06 (0.00040278%), 0 homozygotes; maximum reported population AF 8.909797213015431e-06 (0.00089098%) in European (non-Finnish).
• gnomAD v4.1: variant absent.

Still missing

• The governing file was searched for c.3140A>G, p.His1047Arg, and p.H1047R; no exact variant-specific table entry was found in clingen_brainmalform_acmg_specifications_v1_1.pdf.

Sources

Reasoning

The governing VCEP BS1 threshold is an allele frequency greater than 0.0185%. The highest documented gnomAD v2.1 population frequency is 8.909797213015431e-06, equivalent to 0.00089098%, and gnomAD v4.1 reports the variant as absent. The observed frequency is below 0.0185%, so BS1 is not met.

VCEP rule

An allele frequency (>0.0185%) was approved. (Supplemental Table 3).

Evidence considered

• ClinGen Brain Malformations VCEP specification, benign criteria: BS1 approved allele-frequency threshold is >0.0185%.
• gnomAD v2.1 exomes: overall AF 4.027807986337675e-06 (0.00040278%) and maximum reported population AF 8.909797213015431e-06 (0.00089098%) in European (non-Finnish).
• gnomAD v4.1: variant absent.

Still missing

• The governing file was searched for c.3140A>G, p.His1047Arg, and p.H1047R; no exact variant-specific table entry was found in clingen_brainmalform_acmg_specifications_v1_1.pdf.

Sources

Reasoning

The governing VCEP requires at least three homozygotes in gnomAD or at least three heterozygous observations in well-phenotyped family members for BS2. gnomAD v2.1 reports zero homozygotes, gnomAD v4.1 reports the variant as absent, and the available case evidence contains no qualifying well-phenotyped family observations. Therefore BS2 is not met.

VCEP rule

Award BS2 if ≥3 homozygotes present in gnomAD or ≥3 heterozygous in well phenotyped family members.

Evidence considered

• ClinGen Brain Malformations VCEP specification: BS2 requires ≥3 homozygotes in gnomAD or ≥3 heterozygous observations in well-phenotyped family members.
• gnomAD v2.1 exomes: homozygotes = 0.
• gnomAD v4.1: variant absent; no homozygous observations are reported.
• No qualifying well-phenotyped family-member observations are documented in the available case evidence.

Still missing

• The governing file was searched for c.3140A>G, p.His1047Arg, and p.H1047R; no exact variant-specific table entry was found in clingen_brainmalform_acmg_specifications_v1_1.pdf.
• No qualifying well-phenotyped family-member observations were available in the reviewed case evidence.

Sources

Reasoning

BS3 is not met. The available exact-variant functional studies do not show absence of a damaging effect; instead, they consistently show increased PI3K activity and oncogenic or tumor-forming phenotypes for H1047R. PMID:16322248 reported 2.5- to 3.5-fold increased PI3K activity over WT, while PMID:15016963 reported greater lipid kinase activity than WT. PMID:16432179 and PMID:19029981 demonstrated H1047R-dependent tumor formation in animal models. Therefore the evidence is directionally incompatible with the BS3 requirement for functional studies showing no damaging effect.

VCEP rule

Follow recommendations set forth by the SVI in conjunction with specifications added by the Brain Malformation Group for quality metrics and minimum validation controls required.

Evidence considered

• PMID:16322248, Results, Figure 2A: H1047R PI3K activity was 2.5- to 3.5-fold higher than WT, with additional proliferative and transformation phenotypes.
• PMID:15016963, Results, Fig. S2: mutant H1047R p110-alpha had more lipid kinase activity than wild-type protein.
• PMID:16432179, Results: H1047R caused tumors in 4 of 5 treated animals and increased Akt signaling.
• PMID:19029981, Results: continued H1047R expression was required to maintain established lung tumors, because doxycycline withdrawal caused rapid tumor regression.

Still missing

• No exact-variant entry was found in clingen_brainmalform_acmg_specifications_v1_1.pdf.txt when searched for c.3140A>G, p.His1047Arg, and p.H1047R; the criterion was evaluated from the file's governing BS3 rule and the case's functional publications.

Sources

Reasoning

The governing VCEP rule requires the same amino-acid change as a previously established pathogenic variant regardless of nucleotide change, and specifies that the previously established variant must meet BMVCEP pathogenic criteria independently. The consulted VCEP file contains no entry for c.3140A>G, p.His1047Arg, or p.H1047R, and the available literature records the same H1047R change rather than a different nucleotide change producing His1047Arg. Therefore, the required alternate-nucleotide same-amino-acid evidence is not available for assessment.

VCEP rule

Same amino acid change as a previously established pathogenic variant regardless of nucleotide change ... Brain Malformation Recommendation: Previously established variant must meet criteria as pathogenic per BMVCEP criteria independent of this point.

Evidence considered

• ClinGen Brain Malformations VCEP specification, PS1 rule: same amino acid change as a previously established pathogenic variant regardless of nucleotide change, with the prior variant required to meet BMVCEP pathogenic criteria independently.
• The VCEP full-text file was searched for c.3140A>G, p.His1047Arg, and p.H1047R; no variant-specific entry was found.
• Available paper extracts report the exact H1047R variant, not an alternate nucleotide change producing His1047Arg.

Still missing

• No c.3140A>G entry was found in clingen_brainmalform_acmg_specifications_v1_1.pdf.txt.
• No p.His1047Arg entry was found in clingen_brainmalform_acmg_specifications_v1_1.pdf.txt.
• No p.H1047R entry was found in clingen_brainmalform_acmg_specifications_v1_1.pdf.txt.
• A separately established pathogenic PIK3CA variant with a different nucleotide change but the same His1047Arg amino-acid change was not identified.

Sources

Reasoning

The VCEP requires PS2_Strong only when both criteria are fulfilled: the variant is present at a detectable allele fraction in an affected proband and absent from parental samples with confirmed maternity and paternity, and it is present at a detectable allele fraction in affected tissue but absent from or lower in another tissue. The case evidence includes an exact-match ClinVar submission annotated with origin 'de novo', but does not document confirmed maternity and paternity, detectable allele fractions, affected tissue, or a comparison tissue. Therefore PS2 cannot be assessed from the available evidence. The governing VCEP file was searched for c.3140A>G, p.His1047Arg, and p.H1047R; no variant-specific entry was found.

VCEP rule

Award the PS2_Strong point if Criteria 1 AND Criteria 2 are fulfilled. ... Criteria 1: if the variant is present at a detectable allelic fraction in a proband with the disease but is absent from parental samples with confirmed maternity and paternity. Criteria 2: can be awarded if the variant is present at a detectable allele fraction in an affected tissue sample but is absent from or detected at a lower allelic fraction in another tissue (e.g. if present in 5% of brain tissue but absent from the blood or skin this point can be awarded). ... Award the PS2_Moderate point if Criteria 1 is fulfilled, OR if parents are not available but Criteria 2 is fulfilled.

Evidence considered

• The exact-match ClinVar audit contains a submission annotated with origin 'de novo', but provides no confirmed maternity or paternity, allele-fraction measurements, affected-tissue result, or comparison-tissue result.
• The VCEP specification states that PS2_Strong requires both parental absence with confirmed maternity and paternity and the affected-tissue versus another-tissue allele-fraction criterion; its PS2_Moderate rule permits either the parental criterion or the tissue criterion when parents are unavailable.

Still missing

• No documented confirmed maternity and paternity or parental allele-fraction results.
• No documented affected-tissue and comparison-tissue allele fractions.
• No entry for c.3140A>G, p.His1047Arg, or p.H1047R was found in clingen_brainmalform_acmg_specifications_v1_1.pdf.txt.

Sources

Reasoning

The VCEP PM5 rule requires a different missense change at the same amino-acid residue that has been determined to be pathogenic, independently meeting BMVCEP pathogenic criteria. The collected comparator search found five candidates overall but zero same-residue candidates, and no alternate missense change at His1047 with an independently established BMVCEP pathogenic classification is available. The criterion therefore cannot be evaluated from the available evidence.

VCEP rule

Missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before ... Brain Malformation Recommendation: Previously established variant must meet criteria as pathogenic per BMVCEP criteria independent of this point.

Evidence considered

• ClinGen Brain Malformations VCEP PM5 rule requires a different missense change at the same amino-acid residue previously determined pathogenic, with the prior variant meeting BMVCEP pathogenic criteria independently.
• PM5 comparator collection: total candidates seen 5; same-residue candidates 0; expert-panel pathogenic/likely pathogenic same-residue candidates 0; non-expert-majority pathogenic/likely pathogenic same-residue candidates 0.

Still missing

• A qualifying alternate missense comparator at residue His1047 with independent BMVCEP pathogenic determination is unavailable.

Sources

Reasoning

The VCEP permits PP2 for PIK3CA only when the missense constraint z-score computed in ExAC/gnomAD is greater than 3.09. The available case evidence does not provide a PIK3CA ExAC/gnomAD missense-constraint z-score, so the VCEP threshold cannot be evaluated.

VCEP rule

Brain Malformation Recommendation: Missense constraint computed in ExAC/gnomAD was utilized. Award PP2 if the z-score > 3.09. (applicable to MTOR, PIK3CA and AKT3 but not PIK3R2)

Evidence considered

• ClinGen Brain Malformations VCEP PP2 rule: use missense constraint computed in ExAC/gnomAD and award PP2 if the z-score is greater than 3.09; PIK3CA is explicitly included.

Still missing

• PIK3CA missense-constraint z-score from ExAC/gnomAD is not available.

Sources

Reasoning

The VCEP rule permits BP2 when the variant is observed in cis or trans with a known pathogenic variant in the same gene. PMID:15608678 reports that A3140G (H1047R) and G1633A (E545K) occurred in the same invasive ductal breast carcinoma, but the extracted report does not establish whether the variants were in cis or trans, nor does it provide direct phase testing. Therefore the available evidence does not establish the VCEP's required allelic relationship, and BP2 is not assessed.

VCEP rule

Observed in cis or trans with a known pathogenic variant in the same gene.

Evidence considered

• ClinGen Brain Malformations VCEP specification v1.1, BP2: "Observed in cis or trans with a known pathogenic variant in the same gene."
• PMID:15608678, Results paragraph 2 and Table 1: the paper reports one invasive ductal breast carcinoma with both G1633A (E545K) and A3140G (H1047R), but does not establish cis or trans phase.
• The VCEP file was searched under c.3140A>G, p.His1047Arg, and p.H1047R; no variant-specific BP2 table entry was present.

Still missing

• Phase between H1047R and E545K is not reported; direct cis/trans testing or an equivalent allelic-resolution observation is needed.
• The governing VCEP file was searched for c.3140A>G, p.His1047Arg, and p.H1047R; no variant-specific BP2 pre-assigned code or entry was found.

Sources

Reasoning

The VCEP leaves BP5 unchanged and requires the variant to be observed in a case with an alternate molecular basis for disease in a different gene. The available evidence documents PIK3CA H1047R in tumors and affected cases but does not establish an alternate molecular diagnosis in a different gene for any such case; therefore BP5 cannot be evaluated from the available evidence.

VCEP rule

Variant found in a case with an alternate molecular basis for disease ... Brain Malformation Recommendation: Observed in a case with an alternate molecular basis for disease in a different gene

Evidence considered

• ClinGen Brain Malformations VCEP specification, BP5: 'Observed in a case with an alternate molecular basis for disease in a different gene.'
• PMID:15608678 reports one tumor containing both PIK3CA H1047R and PIK3CA E545K, which is a second variant in the same gene, not an alternate molecular basis in a different gene; this does not satisfy BP5.
• PMID:15520168 reports H1047R in primary cancers without an alternate molecular diagnosis in a different gene.

Still missing

• A case-level report linking this exact variant to an alternate molecular basis for disease in a different gene was not available.

Sources